Min Max Geekhttps://www.minmaxgeek.com/Mon, 06 Apr 2026 18:24:33 +0000en-USSite-Server v@build.version@ (http://www.squarespace.com)Build Log: Skype DeskDIYAVDPMon, 27 May 2024 02:30:00 +0000https://www.minmaxgeek.com/blog/skype-desk6020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:665481738c96df1a4cdb5cdfI originally built Skype Station for hardwood floors and as such, I could get away with a fairly shallow (front to back) footprint. At this house though, the upstairs office has carpet, can lead to some fleeting moments of excitement when you bump into the desk and the desk rocks backwards.

After build the LED Slat Wall last weekend, I opted to convert the existing desk into a side table of sorts. In the interim, I pulled out one of the ubiquitous plastic banquet tables and pressed it into service as a computer desk. Even when directly clamping onto steel under-frame, over the course of the week, I could see the plastic top deforming under the clamping force of the monitor-arms.

Guess I’m building a desk then.

Quick Links

Goals and Objectives

By far, the primary objective was to get a desk that addressed three things:

  • It needed to be sturdy and stable — I was planning on mounting monitors to the desk, I didn’t want to deal with a tipsy, unbalanced desk. To make things harder, remember that everything is on carpet as well

  • I wanted the desk to fit into the rest of the aesthetics of the room

  • I wanted to substantially reduce the cable clutter behind the desk

As a bonus, I wanted this to be a quick build from materials I had on hand.

[Back to Top]

Rough Design

I originally planned on making a desk similar to the existing one: square ‘wheels’ out of 2x4s connected with a few stretchers and with a plywood topper on top. After doing the slat wall though, I was really on a streak for nice, hard, crisp edges so I ultimately decided against that. I wanted something that was more ‘skeleton’ like.

Rough plan for the desk

Since I’m paranoid about tipping, I debated adding an anti-tip leg - something to catch the desk and forcibly keep it from tipping backwards. I wasn’t sure if I was going to need it or not, but I kept this as a backup plan.

Plan for attaching an anti-tip leg that gets bolted to catch the desk if it tilts backwards

[Back to Top]

Step 1 - Making the top

A little while back I had bought a [damaged] sheet of melamine (with red-oak veneer on the opposing side); the sheet I bought was supposed to be 5/8 but with some of the post-damage swelling, it was closer to 3/4 in thickness and there was a definite bow to it. I knew I wanted to laminate two layers of this together to make the top — but which faces to sandwich together? I originally wanted to have melamine exposed on one side and red oak on the other — and make the desk top symmetrical so we have the flexibility to flip the top if desired but I don’t have any melamine glue so that plan was dashed.

The four different options to laminate two sheets together

I definitely didn’t want to rely on mechanical fasteners so this decision ended up being straightforward — the only option available to me was to glue the two wood-surfaces together. My next problem though was that the sheet had a substantial bow in it — I would have to cut slits into the sheet so that it could lay flat. I marked off slits on both surfaces (making sure to offset them so that a slit would not be across from another slit) and after cutting a bunch of slits, the sheet laid flat for easy gluing.

View fullsize Before
 Before View fullsize After
 After

To trim out the perimeter, I debated going with walnut for the incredible contrast but I didn’t have the right lengths handy (and wasn’t in the mood to rip and join smaller strips together). I kept this simple and just used plywood. I cut the edging 1/4” wider than needed and very carefully glued the edging to be perfectly flush with the top side; I had planned to come back and flush-trim the bottom edge, but ultimately didn’t since I just didn’t trust the router to not make a disaster of the melamine. I might revisit that down the road as it’s easy enough to do.

Squaring out the panel
 Squaring out the panel Applying glue to the edge
 Applying glue to the edge More edge trim
 More edge trim Risers
 Risers Masking for stain
 Masking for stain Splash of color
 Splash of color

[Back to Top]

Step 2 - Making the legs

Based on my existing desk, I wanted to hit a top-side height of roughly 29-inches, working backwards, I figured out I needed to cut 27-1/4” legs give or take (the carpet does allow for some extra give). To ensure things were as rigid as possible, I made the legs from plywood glued at 90-degrees. I ripped the two necessary strips so that, after joining, they would be equally sized. To keep everything as consistent as possible, I gang-cut all the strips where possible — and the off-cuts made for perfect spacers to position the legs on the top!

View fullsize Gang cut where possible
 Gang cut where possible View fullsize Determine width to rip leg
 Determine width to rip leg View fullsize Glue and pins
 Glue and pins View fullsize All four legs
 All four legs View fullsize Laminating the front-back brace
 Laminating the front-back brace View fullsize Getting ready for glue-up
 Getting ready for glue-up View fullsize Sanity checking for square
 Sanity checking for square View fullsize Double checking for square
 Double checking for square View fullsize Screwd on both axes
 Screwd on both axes View fullsize Visualizing the leg placement
 Visualizing the leg placement View fullsize Making the long stretchers
 Making the long stretchers View fullsize Adding tape to the bottom of the legs
 Adding tape to the bottom of the legs View fullsize Everything stained, now to dry overnight
 Everything stained, now to dry overnight View fullsize Setting the height of the bottom brace
 Setting the height of the bottom brace View fullsize Glue, pin and screw
 Glue, pin and screw

[Back to Top]

Step 3 - Connecting the legs and top

With the individual legs done, the next step was to connect the two sets of legs with a pair of left-right stretchers and then somehow, connect the table top to the legs. Note that the connection between the two pairs of legs does not need to be super strong — the weight is being loaded down through the legs directly. The stretcher is there to minimize any sagging in the table top and to keep the two sets of legs physically connected to make it easier to handle. I opted to screw-only here, in case I ever wanted to take this apart.

View fullsize Pre-drilling into endgrain of plywood
 Pre-drilling into endgrain of plywood View fullsize Clamping everything in place for drilling
 Clamping everything in place for drilling View fullsize Attaching a back brace
 Attaching a back brace View fullsize Making my own angle brace
 Making my own angle brace View fullsize The connection to the table top
 The connection to the table top View fullsize Super clean
 Super clean View fullsize Extending the hole into the angle
 Extending the hole into the angle View fullsize Clearing room for threaded insert
 Clearing room for threaded insert View fullsize Coming together
 Coming together View fullsize The skeleton assembled
 The skeleton assembled View fullsize With the top in place
 With the top in place

[Back to Top]

Step 4 - Installation

Being comprised of two separate parts made this super easy to bring inside and install in place. As part of the desk upgrade, I put a trailer bed mat down to be a chair mat: figuring out which way to spin the mat and the underlying plywood mat (which I took from my original office desk build) took us the better part of an hour. In the end, we didn’t need to cut anything (which was great, I wasn’t looking forward to trying to cut the mat). The last major thing to do was figure out where to place the monitor arms — which then determines where to drill the grommet hole.

View fullsize Switching to passthrough mode
 Switching to passthrough mode View fullsize Placing the legs
 Placing the legs View fullsize Placing the table top
 Placing the table top View fullsize Figuring out where to place the monitor stands
 Figuring out where to place the monitor stands View fullsize Sacrificial piece to prevent blowout
 Sacrificial piece to prevent blowout View fullsize Precise drilling
 Precise drilling View fullsize Sanity check
 Sanity check View fullsize Fairly clean hole
 Fairly clean hole View fullsize Monitor arms installed
 Monitor arms installed View fullsize And leveled
 And leveled View fullsize Installed
 Installed

[Back to Top]

Step 5 - Positioning

In order to get the positioning, we needed to work backwards a bit

  1. The camera was centered against the center of the wall

  2. The monitor was centered on that same centerline

  3. The came centerline would focus on the user’s left eye

    • The floor mat is then adjusted to be centered relative to the user

    • The desk is shifted to be appropriate for the user’s relative position

  4. The second monitor is then place at a slight angle

  5. The elevation of the tripod and monitors is adjusted to frame the user

  6. The microphone is placed out of frame

View fullsize Positioning the user
 Positioning the user View fullsize The view from the camera
 The view from the camera

[Back to Top]

Milestone

This was a fairly quick project where things mostly fell into place and I’m super grateful for that. I had actually originally planned on doing this desk as part of the LED Slat Wall build last week (albeit as a simpler version) but the wiring woes chewed up all my time.

So far, I don’t think I need to add any kind of anti-tip leg (although it’s not hard to retrofit that down the road if needed).

Desk all done for now

[Back to Top]

Appendix

  • Monitor arms: Charmount Single monitor arm gas-spring mount (max load, 19.8lb, 2.3” grommet capacity, 3.3” clamp capacity). Since the grommet is just an M8 bolt, you could theoretically replace this with whatever custom length you need (if you needed more than 2.3” capacity). I specifically went with two single-mounts rather than a single double-mount as this gives me better flexibility on placement and better weight/force distribution

  • Mat: 4x6x5/8 trailer bed mat. Alternatively, you could also go with a 4x6x3/4 horse stall mat

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Build Log: Skype DeskInterim Skype Station UpdateAVTechDPMon, 06 Mar 2023 12:00:00 +0000https://www.minmaxgeek.com/blog/skype-station-update-2023-16020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:64050bee33a42373b63389a6I have some plans for upgrading Skype Station this year — just waiting on some parts and time to work on it. I plan to split the current setup into two, much more specialized variants. For one of those specialized variants, I wanted to build it around a very intriguing camera setup; luckily I was able to test drive this camera and see where it falls in the line-up of gear I have on hand.

Quick Links

Specialized variants

Objectively speaking, Skype Station is pretty slick: it’s easy to use, fairly reliable, performant and delivers a great experience all around. Over time, I want to improve on what Skype Station can do — better image quality, easier to use etc. One thing Skype Station isn’t good at, is being portable.

We have a fairly nice backyard and I think my wife would like to work outside, amidst all the gross fresh air and nature. I prototyped some ideas for getting decent audio in the backyard — but I’m waiting on the cost of Raspberry Pi’s to come down in order to work on making it a bit more robust.

But what about video? I had previously performed some ad-hoc tests using a bridge camera (funny enough, it was the camera I originally built Skype Station around) but the ease of setup and use was a weak point. Using a mobile phone or a webcam addresses the ease of setup but now I have to deal with battery life and abysmal image quality.

Last year, Insta360 announced a very exciting webcam that threw a ton of buzzwords at us: 4K, gimbal, AI, etc. I was definitely interested — but I’ve also been ‘burned’ by Insta360 before…

Memory lane with Insta360

I own the Insta360 ONE camera which is neat. A single device that lets you take a spherical photo? That’s awesome (and totally handy for doing house-walkthroughs). My problem was with the software - or at least the Android experience. You couldn’t really do anything meaningful with the camera without physically tethering it it. Essentially, it turns your phone into a glorified Bluetooth shutter button for the otherwise fairly awesome gear.

Wirelessly tethering to the camera was a mess too - think olden days of wondering if the damn thing will pair or not.

So I wasn’t keen on being a guinea-pig for this camera. By sheer luck, a colleague of mine was excited enough to go for it and I let me try it out.

The contenders

I have a fairly decent camera line-up to test ranging from full-frame all the way down to the ubiquitous webcam of the 2000s.

 

The lineup of cameras for my test

 

For the tests to be meaningful and allow each camera to perform at its best:

  • Each camera is tested with the best available video settings including hand-tuned color-grading for each camera

  • The camera is mounted into the same teleprompter setup with minor positioning adjustments as needed to keep the framing as consistent as possible (to account for different focal lengths)

  • A mannequin head mounted on a tripod was used as a subject to ensure minimize/eliminate any influence subject-positioning might have

  • Where possible/feasible, I locked-off as many settings as possible (i.e., white-balance)

  • Lighting was controlled for the environment and consistent from camera to camera. For some of the cameras, I needed to move the lighting ever so slightly further away in order to keep the lights out of frame

The test environment

Note

One thing to note though: although I did tune each camera to try and give the best results, the ZV-E10 has a not-insignificant advantage since it’s dedicated to Skype Station and has had the most time spend fine-tuning.

[Back to Top]

Results

Sony A7 III

For full-frame cameras, I have both the A7iii and the A7R iv available; they are roughly equivalent in the headline feature: both will provide 4K with 8-bit color. I opted to actually test the A7 iii here because [a] the price point is lower, making it a more useful data point and [b] the A7 iii has an AA filter, which will give it a slight advantage on video. Will it matter? Likely not for scenarios like this.

Shot specifics:

  • The video stream is 4K

  • I used the Tamron 28-75 f/2.8 at 22mm

  • Aperture was locked off at f/2.8

  • Shutter speed was locked off at 1/50s

  • White balance was locked off at 5500K

In practice, the A7Riv may have a usability advantage as it has realtime eye-autofocus in video but for the purposes of this test, the camera was pre-focused as best as reasonably available.

Sony A7 iii — I guess I could have moved that light ever so slightly more out of frame huh? ;)

[Back to Top]

Sony ZV-E10

This camera is an APS-C, vlogger-centric camera that I bought as an upgrade camera to the original Skype Station; as much as I would have wanted 10-bit, the cheapest Sony option at the time would have been the A7S iii which was prohibitively expensive to use as a webcam lol. Even now, the cheapest option would be the A7 iv which is still pushing the limits on what I would consider reasonable as a webcam.

I have this camera and lens permanently allocated for Skype Station - having a prime lens on here allows me to have an even lower aperture than achievable with my full-frame camera (with the lenses available to me at this time).

Shot specifics:

  • The video stream is 4K

  • I used the Sigma 30mm f/1.8. Other than having a faster aperture, one nice advantage of prime-lenses is reliability: I never have to worry about whether the lens has shifted focal length ever

  • Aperture was locked off at f/1.8

  • Shutter speed was locked off at 1/50s

  • White balance was locked off at 5500K

Even though this setup has a smaller sensor than the A7iii, I’m able to leverage the amazing Sigma prime to really drive the shot here. Since my full-frame glass is ‘only’ f2/.8, technically speaking, this setup has a lower aperture (f/2.1).

Sony ZV-E10 — this is the baseline camera & lens allocated for Skype Station

[Back to Top]

Sony RX10

This was the camera that started it all — the original camera I used for Skype Station. This is an old camera (2013!) but covers all the basics: constant f/2.8 aperture, a relatively large 1.0” sensor, eye-AF and even supports modern Sony hot-shoe accessories. After seeing the modern cameras, you can totally tell this is a substantial downgrade but if you take into account it’s age, this is stellar. This camera has a maximum video resolution of 1080, so keep that in mind.

Shot specifics:

  • The video stream is 1080

  • This camera has a built in lens with a constant aperture of f/2.8 I shot this at 18mm (about 48mm equivalent)

  • Aperture was locked off at f/2.8

  • Shutter speed was locked off at 1/50s

  • White balance was locked off at 5500K

Sony RX10 - the camera that started it all

Initially this image looks noisier than the previous but the specific choice of using ‘bricks’ as a backdrop helps a lot and lets some of the noise be ‘explained away’. If you remember that this camera is limited to 1080, this is actually very workable quality.

[Back to Top]

OnePlus 9 Pro

I got new phones a little while back so this was a good opportunity to see just how far smartphones have come. Using a plugin called DroidCam, we can feed a camera stream via the network into OBS. This is actually really cool because this is a super easy way to get into multi-cam streaming/recording. From what I can tell, the stream is limited to 1080 (coming into OBS) but the camera on the phone is capable of 4K so I suspect that through the process of super-sampling, noise is reduced substantially.

Shot specifics:

  • The video stream is super-sampled 1080

  • The phone has multiple lenses, from what I can tell, this is shot using the primary lens which is 1/1.43”, 23mm (equivalent)

  • Aperture is f/1.8 (but with crop factor, this comes out to f/16.7)

  • There’s no option to lock off shutter speed

  • There is an option to lock off white-balance but there’s no numbers — just a scale that you drag, so I opted to leave this on ‘auto’

OnePlus 9 Pro - crazy what we can do with smartphones

So right off the bat, one huge takeaway here is that smartphones are awesome in terms the image quality they can put out - everything is crisp and sharp and there’s virtually no noise. This comes at the cost of

  1. Limited adjustability. Maybe this is a OnePlus limitation, maybe it’s an app limitation etc.

  2. Smartphone video tends to trend towards overdoing the sharpness filters though; in this specific context, I think it’s fine however, smartphones also overdo the saturation levels: this is the most saturated result by a huge margin

  3. Battery life, heat etc.,

For short bursts, phones make an excellent supplemental camera if you can deal with the extra saturation. Specific to Skype Station, one thing to note is that a lot of phones have their cameras in the corners of the phone — which makes it tricky to physically place the phone behind a small teleprompter as it’s really easy to hit the sides of the teleprompter.

[Back to Top]

Logitech C920

What might be the original premium webcam, this was the webcam to get in the early 2010s and sadly, that’s about when Logitech’s drive to innovate and push out better products went out to lunch. They launched the Brio 4K a little while back but that feels like a case of too little, too late. Now that we’ve had a taste for what is possible, the C920 is categorically bad.

Shot specifics:

  • The video stream is 1080

  • Information on specifics of this webcam are exceedingly difficult to find

  • From what I can guess, this has a 1/3” sensor and uses a 28mm equivalent focal length

  • There’s almost nothing we can set, everything is ‘auto’

Logitech C920 - hard to think, this was the best we could do at one point

A decade ago, this webcam was hot stuff but I don’t think it has aged well. I think Logitech got comfortable with ‘well, it’s better than integrated webcam, what else are you going to do’ and when COVID and Elgato’s Camlink carved up a whole new market segment.

Logitech has released a couple minor follow-up webcams recently, but I can’t imagine they’ve substantially improved anything - certainly not for the ~$100 they are asking.

  • C920

    • C920S: (lol) Adds a privacy cover (something you could just add anyways)

    • C920E: (lol) Microphone is disabled by default, also adds a privacy cover

    • C920X: (lol) Adds a 3-month license for Xplit

  • C922: Adds a nearly pointless 60fps recording — only at 720P

    • C922X: (lol) Adds a 3-month license for Xplit

    • C922E: (lol) Microphone is disabled by default, also adds a privacy cover

  • C930: widens the field of view to 90° (up from 78°); whether this is useful is dependent on your specific setup

[Back to Top]

Insta360 Link 4K

And finally we get to the camera I set out to test, the Insta360 Link. While it’s not truly unique — there is a similar, slightly cheaper OBSBOT Tiny PTZ 4K but I would argue the Link is more compelling because of the [ever so slightly] bigger sensor and smaller footprint - the Insta360 Link is positively tiny, even when compared to the OBSBOT which is already quite small. Of course, what I’m here to investigate is the image quality.

Shot specifics:

  • With HDR, the stream is limited to 1080. Without HDR, the stream is 4K

  • This camera has a built in lens of 26mm (equivalent)

  • Aperture was locked off at f/1.8 (f/9.7 equivalent)

  • There’s no control for the shutter speed

  • White balance was locked off at 5500K

View fullsize With HDR (1080)
 With HDR (1080) View fullsize Without HDR (4K)
 Without HDR (4K)

It’s a bit unfortunate that HDR comes at the cost of a resolution drop (you also lose high-framerate recording but that’s less of an issue in this application). If all I knew was the 1080-footage-with-HDR, I think I would be satisfied but now knowing how sharp it can be without HDR? In my opinion, you can claw back a bit of the colors with a bit of grading and looking at them back to back, the HDR footage starts to look bad.

 
 

One small gotcha…

When run through OBS, the video feed introduces a noticeable amount of lag. Oddly enough, I noticed that there was substantially more lag if the video feed was configured as H264 rather than MJPEG — and here I thought encoding tech had gone past the days when this was even something to consider. Pardon the following grainy captures — there’s only so much I can do with animated gifs ;)

Configured as H264

The lag here is quite noticeable and will certainly be jarring for people viewing it.

Configured as MJPEG

There is still a bit of lag but it is substantially reduced.

The process of doing the stream capture and then converting to an animated gif does tend to exaggerate the motion a bit (which is helpful in this case when I’m trying to identify lag). In the real world, (where frames are not being decimated to create animations), there is virtually no lag when configured as MJPEG. A bit disappointing that this limitation cropped up given that I thought almost anything with more computational horsepower than a glass of water could handle H264 now. So for best results with the Insta360 Link, it appears we should:

  • Set HDR off

  • Use 4K resolution

  • Bring colors back with a quick grade

  • Configure the feed to use MJPEG

[Back to Top]

Conclusions?

The image quality from the Insta360 Link (non-HDR) looks great. Compared back to back with the ZV-E10, for sharpness, it definitely punches well above its weight. It’s only when you start to look for differences:

  • The lamp in the background is sharper (because the ZV-E10 has a much shallower depth of field)

  • Looking back and forth between the photos, you might notice bit of the mannequins face might have some noise (the sensor on the ZV-E10 is much bigger so there will be less noise)

Categorically, the ZV-E10 is better — but the fact that the Link passes the quick-glance test is incredible. The Link isn’t exactly cheap ($400) but it’s in a great position where anything cheaper is substantially worse and anything meaningfully better will cost a lot more. Worth noting is the possibility of using a smartphone as a portable-camera although this depends on your specific phone (as you start looking at older and older phones, the capabilities may fall off dramatically) and you’ll need to figure a way to handle the heat and drain on the battery.

I initially set out to see about finding a camera option for a more portable Skype station — and the Insta360 looks like the camera that fulfills that niche.

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Interim Skype Station UpdateBuild Log: Entertainment HutchDIYAVDPWed, 22 Sep 2021 10:45:00 +0000https://www.minmaxgeek.com/blog/entertainment-hutch6020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:614954505b5212576724291aThis is our first time being in a place that might justify having a second TV (not just for the sake of having a second TV) and because we’ve only ever really been in places where one TV made sense, we didn’t have much by way of furniture, accessories or organization for the second TV. Being the secondary setup, it was it’s mostly just me piling crap haphazardly in the general direction of the TV - my wife’s been pretty patient about this being a total mess! Since this is a secondary setup and intended to be a quick in-between-projects build, I didn’t want to make it too fancy or expensive.

Quick Links

Goals and Objectives

This entertainment hutch is intended to be a quick project that let me use up some material I had lying around and to make a home for a bunch of crap in our living room. In particular, I wanted to have a somewhat dedicated space for our robo-vac to park so we’re not tripping over the base station all the time. In our space, we happen to have a cable passthrough already installed in the wall for a TV so I wanted to leverage that to keep things as tidy as I could. I also wanted to have a cubby for my subwoofer on one side, a UPS on the other and to try and keep the cables manageable — all without buying a receiver.

[Back to Top]

Step 1 - Left Cubby

The left cubby was the easy one: I had already built a similar enclosure when making my second wall-desk. In fact, a long time ago, I had stumbled upon the very same set of speakers at a pawn shop for $60 so I had an extra set kicking around. Well, I had this second set at work, but we all know what happened to going into the office. Having the same set at home made it easy to mock out and size the cubby in person.

View fullsize Laying everything out
 Laying everything out View fullsize Working with materials on hand
 Working with materials on hand View fullsize Notching out for the baseboard
 Notching out for the baseboard View fullsize More scraps
 More scraps View fullsize Cutting notches for the braces
 Cutting notches for the braces View fullsize Cutting out an opening for the subwoofer
 Cutting out an opening for the subwoofer

[Back to Top]

Step 2 - Right Cubby

The right cubby was a bit complicated: I originally started with just a side panel and a way to mount the UPS high up: this way, the bottom portion of the panel would be ‘clear’ so I didn’t have to make the entire thing even wider to accommodate the robovac, but then I thought that might not be very rigid, not to mention super tippy. So I added another panel to make a box out of it but also had to cut out a large opening so that I could actually install/remove/use the UPS. But by making this its own ‘box’, I had the space below the UPS which would come in handy later.

View fullsize Laminated 'shelf'
 Laminated 'shelf' View fullsize Cutting triangle braces
 Cutting triangle braces View fullsize Assembled
 Assembled View fullsize Cutting the access hatch, part 1
 Cutting the access hatch, part 1 View fullsize Cutting the access hatch, part 2
 Cutting the access hatch, part 2 View fullsize Cutting the access hatch, part 3
 Cutting the access hatch, part 3 View fullsize Trying to figure spacing on the UPS
 Trying to figure spacing on the UPS View fullsize Good fit
 Good fit View fullsize Laying out the shelf
 Laying out the shelf View fullsize Laying everything out
 Laying everything out View fullsize Pocket holes to the rescue
 Pocket holes to the rescue View fullsize In the house
 In the house

[Back to Top]

Step 3 - Bringing it Together

I thought the hard parts were behind me, funny enough, getting everything wired up in clean way was nearly just as much work!

View fullsize Gawd it's a mess
 Gawd it's a mess View fullsize Rough-wired in
 Rough-wired in View fullsize Cord management
 Cord management View fullsize Installed
 Installed View fullsize Mostly assembled
 Mostly assembled View fullsize Making a door
 Making a door View fullsize Cutting a catch
 Cutting a catch View fullsize Like this
 Like this View fullsize Washers to the rescue
 Washers to the rescue View fullsize Voila!
 Voila! View fullsize Friction fit shelf
 Friction fit shelf View fullsize Very clean
 Very clean

[Back to Top]

Milestone

When I went back to the office, I grabbed my speakers — the living room TV setup is good for now

Done for now

[Back to Top]

Mixups, Surprises and Lessons Learned

By far the hardest challenge was being constrained to working with that initial off-cut of plywood as the primary material for this build:

  • The width of the off-cut dictated how deep the overall shelf could be

  • However, the size of the UPS dictates how big the opening for it needs to be

  • Which ultimately means the amount of material left after cutting out the opening was fairly small

On the front side, this doesn’t have a huge impact - the plywood off-cut I used as the shelf support only needs to support a screw… but on the back side (where I biased my cutout), this determines the front-to-back placement of the cord cover … and I could totally use another inch or two to shove all the cords!

View fullsize Limited space behind the cord cover
 Limited space behind the cord cover View fullsize Finicky front panel
 Finicky front panel View fullsize Cable cut outs could be bigger
 Cable cut outs could be bigger

The remaining gripes are mostly fit and finish:

  • The front cover was a bit of an afterthought, I couldn’t really make a super clean door for it and, due to the way I had to cut a relief on the back side, it exaggerated the tendency for it to warp

  • I made the cable cutouts prior to assembling (and didn’t have all of the cables handy) so it was a bit of an eye opener when I started feeding cables into/out of the cubbies that I was totally maxing out the capacity of the meager pass-through (I actually bought thin HDMI cables to alleviate this)

 
 

On the A/V side, the big challenge I have is that, I have no way to remotely control the volume: instead of buying a receiver, I used a HDMI splitter with a built-in audio extractor - this lets me directly hook up my computer speakers (but my speakers don’t have a remote). My current cheapo TV does have an audio-out that I can connect the speakers to, but it is not linked to the TV’s volume setting (and I don’t have a headphone jack which would be tied to the TV’s volume control). I snaked a 3.5mm stereo cable up through the wall in preparation for a future TV upgrade that will have a headphone port.

At some point, I think I’ll want to replace my cheap little Android Kodi box with something more performant that won’t choke up parsing through my NAS or playing 4K HDR x265 content. Since this is intended to be a secondary setup, I don’t think I’ll go with a full blown HTPC as I did in the basement, but it’s definitely something I’ll revisit after I upgrade the TV.

[Back to Top]

Featured Gear

The bones of the setup is pretty plain:

  • Cheap 1080P TV we had before I upgraded the main TV setup downstairs

  • Google Chromecast Ultra and Amazon Fire Stick 4K

  • A cheapo Android Kodi Box (AMLogic S912 , 2GB RAM). Officially, this decodes 4K60 HDR10 in H265, but reality suggests otherwise

The bits that might be a bit more unique:

  • APC Back-UPS 550 which is strong enough to keep everything running through a power blip; I used a CableCreation right-angle outlet saver to reduce the footprint to the wall

  • Mounting Dream MD2617-CA fully articulating TV mount; I probably could have got away with a tilting mount, but I was familiar with the installation of this one specifically (since I just installed it for the main TV) and the extra articulation comes in handy if I ever want to do anything behind the TV (like run a stereo cable up)

  • Logitech Z-2300 speakers which are sadly, discontinued however, the modern spiritual successors (Z623/Z623) are a good stand-in.

  • A 3-source HDMI switcher with audio extraction (both 2.1 and 5.1) with a remote; I used thin HDMI cables (2’/6’) in conjunction with couplers to make HDMI extensions - this allowed me to put the Fire and Chromecast out of sight underneath the UPS

  • SofaBaton smart universal remote to try and keep the amount of clutter in the living room to a minimum

  • Our faithful robo-vac, the Eufy 30C; there are some newer models now but this model goes on sale regularly and does everything we need (for now!)

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Build Log: Entertainment HutchA Dive into Security Cameras, Part 1AVLifeDPTue, 15 Jun 2021 04:00:00 +0000https://www.minmaxgeek.com/blog/security-cams16020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:60c43da0a8a0ec164d3c62e5You should feel safe in your home and for some people, having security cameras provides a bit more peace of mind. For me, I’ve always been interested in footage management and how adding a single device (the camera) can be a super-simple ‘open the box and go’ affthe air to the other extreme of a multi-layered system with object tracking, thermal overlays and acoustic-triggering. Of course, finding a way to tame the chaos of a multitude of cameras recording nonstop and finding a sweet point in automation is its own flavor of fun (or frustration - it’s a very thin line!)

Quick Links

Getting Started

For consumers, there are three broad use cases of surveillance or security cameras

  1. Security: this can be something like a camera watching the driveway for people coming up to the house, or be more subtle like a nanny camera just to make sure the babysitter is on the up and up

  2. Package notification: I think this might be the bigger motivator: we all worry about porch pirates, so having a quick ping to the phone when a delivery comes is great

  3. Glorified baby monitors: this could literally be a baby monitor, an ‘is the pot boiling yet’ monitor, a ‘my kids are swimming in the pool is everything okay’ monitor etc.

At MinMax, of course, I’m looking to cover all of the above. And then some.

[Back to Top]

Goals, Objectives and Criteria

There are so many brands and models of cameras that we would be analyzing them forever; according to Synology’s camera compatibility list, they support over 7500 models from 140 vendors. We need to very quickly bring that number down to something that we can process through. I think, with five questions, we can sufficiently narrow down the options:

  1. How do you want to connect and power the camera: via power-over-ethernet (POE), WiFI (with dedicated power) or purely battery/solar powered?

    • Battery/solar powered gives you the flexibility to setup nearly anywhere (i.e., in a tree!); in addition to having to replace/recharge the batteries every so often, every aspect of the ecosystem will be battery-conscious, so you will generally be limited to motion-triggered capture and on-demand streaming only

    • POE cameras have constant power so you have the ability to do continuous recording reducing the reliance on motion or event detection; being constantly powered, they will typically give you the best video quality options. Having a wired connection allows for a very stable video stream. The downside is that you will need a direct connection between the camera and the video recording system

    • Wi-Fi cameras are somewhere in between: they are continuously powered but any power connection will do - their connection to the video recording system is wireless. The downside is that you’re at the mercy of a wireless connection. Video quality will be somewhere in between the battery and POE options

  2. What type of recording do you want to do: motion/activity triggered? Scheduled recording? Continuous (24x7) recording?

    • This choice is only available if you opted for a Wi-Fi or POE camera

    • The benefits are pretty obvious: you can capture everything and not have to worry about whether a motion/event triggering algorithm fires correctly or not

    • The downside is that continuous recording generates a lot of data (and cameras/recorders aren’t exactly concerned with optimized encoding). This adds another layer of complexity when you have to decide take into consideration the video codec used by the cameras. You’ll almost certainly be relying on some kind of centralized video recording system

  3. Are you looking for indoor or outdoor cameras? Do you have impact/vandalism considerations?

    • Indoor cameras will generally be a bit more svelte and compact as they don’t need to worry about water or snow

    • When looking at outdoor capable cameras, take into consideration the operating temperature if you live in and an environment where it gets particularly hot or cold

    • More so for public/commercial spaces, impact/vandal-resistant cameras are available - you’ll just need to pay to play!

  4. What kind of ‘smart home’ integration are you looking for? Or conversely, is there an ecosystem you are specifically looking for?

    • Having smart home integration can be pretty cool - you can have camera feeds automatically come up when events are triggered etc.

    • Of course, the downside is that this infrastructure presents a bigger cyber-security surface area to manage (if you care)

    • Note that you don’t necessarily have to go all-in or all-out — you can integrate individual cameras into your smart-home if desired

  5. Is there anything you want to do with the footage?

    • Do you want to get the footage offsite?

    • Do you want to do analytics on the footage (i.e. count people) or generate time-lapses?

    • Do you want to have a dedicated TV to display all of the camera feeds?

[Back to Top]

Step 1: Setting the Stage

We specifically wanted to look at continuous recording: if you plan for continuous and run into problems or limitations (likely disk space), you can always scale something back to make it work. Continuous is just easier to plan for - there will always be 24 hours in a day whereas motion-triggered recording will obvious vary based on how much activity is detected.

 
 

Powering POE Cameras

A POE camera receives data and power from via the built-in ethernet port. There are four ways to connect and power a POE camera

  1. Use a POE injector: you connect your POE injector to your existing network hardware (this provides the data link) and then you connect your POE camera to the POE injector. POE injector itself has a power cable and supplies the necessary power to the downstream device. Having a POE injector is great for testing!

  2. Use a POE switch: here, you just connect your POE camera directly to the switch like any other device. For non-POE devices, everything behaves as usual but POE-devices, will receive power from the switch as needed

  3. Use a POE NVR: essentially the same as connecting to a POE switch, but here, you connect the camera to a network video recorder (NVR) that has POE ports

  4. Use a DC power adapter: most POE cameras will also come with a DC input connector that allows you to connect to a DC power adapter (usually not included). In this case, you will have at least two cables running from your camera: one ethernet cable for data, running to your network infrastructure and a DC power cable to provide power.

Tip:

Even if you plan on using POE infrastructure, I would recommend getting a POE injector (preferred) or DC adapter (works as well). Having one will allow you to do testing while up on a ladder if you don’t necessarily have ethernet cables already in place. A POE injector is more useful since it will use a standard RJ45 connection so it will work also for additional POE devices down the road.

[Back to Top]

Centralizing Management

Going with POE cameras means that we need to figure out how to store all the footage and manage all of the cameras in a centralized manner. To do this we will need a network video record (NVR). Most NVRs that work with POE cameras will also have built in POE (meaning, you won’t need to buy a POE switch or buy a whole bunch of POE injectors).

What about Digital Video Recorders (DVR)?

A DVR is essentially the same thing as a NVR — except instead of connecting Wi-Fi or POE cameras, they allow you to connect older analog cameras that use a coaxial BNC connector. For my purposes (and I imagine, for most consumers/prosumers), I’ll be looking at digital (POE) cameras.

In our case, we have a Synology NAS unit that can act as a NVR (and I plan on adding a POE switch). Almost all NAS units will support having security cameras in one way or another, even if it is just via a FTP or network share for the video stream to pipe to although some brands offer a very rich feature-set. For some of the more popular brands

Your specific experience will vary based on the management software available as well as the hardware specifications of the specific NAS unit you’re looking at. Some NAS brands (QNAP, Synology) even offer more specialized NVR units.

A word on licenses:

One of the ways NAS brands ‘get you’ is with camera-licenses. Most NAS units will allow you to connect a whole bunch of NAS units (it varies by model, some will allow you to connect 40 or more cameras!) but the NAS will only include one or two camera licenses. If you want to connect more, then you’ll have to buy additional connection licenses. Thankfully these add-on licenses you buy are generally transferable.

The more specialized NVR NAS units from Synology and QNAP come with eight licenses out of the box which sounds great — except those licenses are embedded into that unit. This means that if you ever replace or upgrade that unit, you’ll either have to continue paying the premium for the more specialized NVR, or you’ll have to shell out for individual licenses (which thankfully, can be transferred)..

[Back to Top]

Step 2: What I wanted in a Camera

If you are going with a branded NVR unit (i.e., the same brand as the cameras), then this step won’t matter; if you’re using a NAS unit I would recommend getting a camera that is officially supported by your NAS unit. Consult your NAS IP camera compatibility list to be sure.

Before we dive into the possibilities, it’s important to figure out what we want from our camera. Every scenario will be different and you may even have different requirements for different cameras.

Note!

For audio recordings, please check your local laws on whether or not this is allowed! (you can orf course, always disable audio recording).

 
 

Regardless of whether you have a Synology NAS or not, their Supported IP Camera tool is one of the best that I have found because it allows you to specify what features you care about first (whereas all of the other compatibility tools assume you have a specific camera/brand in mind already).

 
 

[Back to Top]

Step 3: A look at Reolink

For one reason or another, I started my exploration with Reolink. What I liked the most actually, was navigating the product stack without having to hop through half a dozen layers and (for better or worse), not having a metric ass ton of cameras to sift through, clicking on endless ‘View More’ panels. It was a bonus that the cameras are reasonably affordable and relatively available. I also liked that the cameras were grouped by connectivity only (POE, Wi-Fi, Battery) and they were not grouped by form-factor (bullet, dome, etc.) but depending on what you’re looking for, it may not be an issue for you.

Although I’m primarily looking at POE cameras, I’ve included the Wi-Fi and Battery model breakdown as well with some general comments. Here’s the rough decision matrix for the three different families of Reolink cameras.

View fullsize Picking a POE cam
 Picking a POE cam View fullsize Picking a Wi-Fi cam
 Picking a Wi-Fi cam View fullsize Picking a Battery cam
 Picking a Battery cam

A few common terms:

  • IR-Cut: security cameras use IR illuminators to allow them to see at night; during the day when there is sufficient lighting, these IR illuminators can cause color distortion. Cameras with a an IR-cut filter can filter this out to improve the video signal

  • Sub-Stream: when you show the camera’s live-view, you are generally viewing the main video stream; but when you are looking at list of cameras with just the little thumbnails? That’s the sub-stream. Generally speaking, the main-stream is the one I am concerned with

 
 

POE Cameras

Since I’m specifically looking at cameras that are H265 and 4K, this limits me to the 8-series of cameras which are then broken down into two more decisions based on form-factor and to a lesser degree, pricing:

  • Bullet (810A vs 811A). The 811A is the functionally superior camera in terms of field-of-view and the potential flexibility of having zoom (even assuming that the 811A will likely work with Synology even without official support), but the 810A is likely the better camera for me. With the 811A, I would end up paying too much for features I outright don’t want (2-way audio, lights, siren)

  • Dome (820A vs 822A). Here, the 822A is the superior camera: better field of view, brighter fixed-aperture lens and with the flexibility for zoom

ABCDEHIJKLMNOPQRSTUVWXYZAAABACADAEAFAG
1CameraMain Video StreamSub-StreamViewAudioStorageEnvironmentalPOE
2ModelSynology SupportSpec SheetStyleSpecial Feature at a glanceSensorLensIR CutCodecResolutionFPSMax BitrateResolutionFPSMax BitrateH-FOVV-FOVIR DistanceMicSpeakerSD Card Recording ModesMax SDCard (GB)PlacementProtectionMinMaxWarranty (years)DetectionRTSPDrawPOE Spec
3RLC-410YSpecsBullet1/2.7"4.0mm f/2.0NoH2642560x1920258192640x4806512805830YesNoMotion64Indoor/OutdoorIP66-10C55C??MotionYes<10W802.3af
4RLC-423YSpecsPTZPan-Tilt-Zoom, Zoom (4x)1/2.7"2.7-12mm, f/1.6NoH2642560x1920308192640x480651231-8724-6560YesNoMotion64Indoor/OutdoorIP66-10C55C??MotionYes<20W802.3at
5RLC-520YSpecsDome1/2.7"4.0mm f/2.0YesH2642560x1920308192640x48015512805830YesNoMotion64Indoor/OutdoorIP66-10C55C2MotionYes<8W802.3af
6RLC-510AYSpecsBulletAI1/2.7"4.0mm, f/2.0YesH2642560x1920308192640x48015512804230YesNoMotion + Schedule256Indoor/OutdoorIP66-10C55C2Person, Car, MotionYes<12W802.3af
7RLC-520AYSpecsDomeAI1/2.7"4.0mm, f/2.0YesH2642560x1920308192640x48010512804230YesNoMotion + Schedule256Indoor/OutdoorIP66-10C55C2Person, Car, MotionYes<12W802.3af
8RLC-522YSpecsDomeZoom (3x)1/2.7"2.8-8mm, f/1.6YesH2642560x1920308192640x4801551248-9437-7130YesNoMotion64Indoor/OutdoorIP66-10C55C2MotionYes<10W802.3af
9RLC-511YSpecsBulletZoom (4x)1/2.7"2.7-12mm, f/1.6NoH2642560x1920308192640x480651231-9024-6730YesNoMotion64Indoor/OutdoorIP66-10C55C??MotionYes<10W802.3af
10RLC-810AYSpecsBulletAI1/2.49"4.0mm, f/2.0YesH2653840x2160258192640x48015512874430YesNoMotion + Schedule256Indoor/OutdoorIP66-10C55C2Person, Car, MotionYes<12W802.3af
11RLC-820AYSpecsDomeAI1/2.49"4.0mm, f/2.0YesH2653840x2160258192640x48015512874430YesNoMotion + Schedule256Indoor/OutdoorIP66-10C55C2Person, Car, MotionYes<12W802.3af
12RLC-822AYSpecsDomeAI, Zoom (3x)1/2.8"2.8-8mmm f/1.6YesH2653840x2160258192640x4801551250-9430-5330YesNoMotion + Schedule256Indoor/OutdoorIP66-10C55C2Person, Car, MotionYes<12W802.3af
13RLC-811ANSpecsBulletAI, Light, Siren, Two-Way Audio, Zoom (5x)1/2.8"2.7-13.5, f/1.6-3.3YesH2653840x2160258192640x4801551231-10517-5530YesYesMotion + Schedule256Indoor/OutdoorIP66-10C55C2Person, Car, MotionYes<12W802.3af

[Back to Top]

Wi-Fi Cameras

Compared to the POE cameras, there’s nothing that supports H265 or 4K. Given that we are trading the convenience of not having to pull network cables through, it makes the most sense to go for a PTZ to maximize on the strengths. Something to note: not supporting 4K might be understandable (due to the unpredictability of Wi-Fi), whether or not a camera supports H265 is a limitation of the processor on the camera.

  • PTZ (E1 Outdoor): realistically, this is the only choice. The other PTZ options make big compromises on bitrate, frame rate and have dimmer optics. This camera supports 2.4 and 5GHz WiFi bands as well as wired ethernet and comes with Person/Car detection as well

  • Bullet (510WA vs 511WA): like the 810A/811A comparison above, here the 511WA has the superior feature set, but if you’re going to go for the more svelt, battery conscious bullet form-factor, the 510WA is my pick since I’m not paying for features I don’t want

ABCDEHIJKLMNOPQRSTUVWXYZAAABACADAEAFAGAHAIAJAKALAMANAO
1CameraMain Video StreamSub-StreamViewAudio/AlertStorageEnvironmentalConnectivity
2ModelSynology SupportSpecsStyleSpecial Feature at a glanceSensorGlassIR CutCodecResolutionFPSMax BitrateResolutionFPSBitrateH-FOVV-FOVPan-HTilt-VIR Distance (m)Color NV Distance (m)MicSpeakerSpotlightAlarmSD Card Recording ModesMax SDCard (GB)PlacementProtectionMinMaxWarranty (years)DetectionEthernetWirelessBandsSecurityRTSPPower Draw
3RLC-410WYSpecsBullet1/2.7"4.0mm f/2.0NoH2642560x1440308192640x48015512804230YesNoNoNoMotion64Indoor/OutdoorIP66-10C55C?Motion10/100802.11n2.4/5WPA2-PSKYes<8W
4RLC-510WANSpecsBulletAI1/2.7"4.0mm f/2.0YesH2642560x1920308192640x48015512804230YesNoNoNoMotion/Schedule256Indoor/OutdoorIP66-10C55C2Person, Car, Motion10/100802.11n2.4/5WPA2-PSKYes<12W
5RLC-511WANSpecsBulletZoom (5X), Color Night Vision, 2-way audio1/2.7"2.7-13.5mm f/1.6-3.3YesH2642560x1920308192640x4801551227-9621-693030YesYesYesNoMotion/Schedule256Indoor/OutdoorIP66-10C55C2Person, Car, Motion10/100802.11n2.4/5WPA2-PSKYes<12W
6RLC-511WYSpecsBulletZoom (4X)1/2.7"2.7-12mm f/1.6NoH2642560x1920308192640x4801551231-9024-6730YesNoNoNoMotion64Indoor/OutdoorIP66-10C55C?Motion10/100802.11n2.4/5WPA2-PSKYes<10W
7E1 OutdoorNSpecsPTZPan, Tilt, Zoom (3X), Auto-tracking, Color Night Vision, 2-way audio1/2.7"2.8-8mm f/1.6YesH2642560x1920258192640x4801551248-9037-68355501212YesYesYesNoMotion/Schedule256Indoor/OutdoorWaterproof-10C55C2Person, Car, Motion10/100802.11n2.4/5WPA2-PSKYes<12W
8E1NSpecsPTZPTZ, 2-way audio1/2.7"4.0mm f/2.4NoH2642304x1296153072???82433555012YesYesNoNoMotion64Indoor Only-10C55?MotionNo802.11n2.4WPA2-PSKNo<6W
9E1 ProYSpecsPTZPTZ, 2-way audio1/2.7"4.0mm f/2.4YesH2642560x1440204096???87.5473555012YesYesNoNoMotion64Indoor Only-10C55?MotionNo802.11n2.4/5WPA2-PSKYes<6W
10E1 ZoomYSpecsPTZPTZ, 2-way audio1/2.7"2.8-8mm f/1.6YesH2642560x1920204096???48-9837-753555012YesYesNoYesMotion64Indoor Only-10C55?MotionNo802.11n2.4/5WPA2-PSKYes<6W
11LumusNSpecsPuck2-way audio, spotlight, siren, Color Night Vision1/2.8"2.8mm f/2.0YesH2641920x1080152048704x480157681005410YesYesYesYesMotion64Indoor/OutdoorIP65-10C552MotionNo802.11n2.4WPA2-PSKNo<6W

[Back to Top]

Battery Operated Cameras

With battery only cameras, we should really be targeting totally infeasible or awkward places to add a camera: trees, unpowered sheds etc. None of these cameras support continuous recording and only one supports H265 or resolution above 1080. For this class, battery life is the big consideration rather than image quality — which is evident by the lack of detailed information on sensor or lens information.

In certain applications (i.e., a private shed) it may be a viable option to run a small solar panel to a large power-bank to “plug in” these battery powered cameras. That may help to provide a substantial boost in effective runtime.

  • PTZ (Argus PT): although it may cost more power to run the PTZ components, this would allow us to truly maximize the flexibility of a battery camera. Note that any kind of motion triggering will only apply to the field of view the camera happens to be pointed at the given moment.

  • Puck (Argus 3 Pro): this is the do-it-all battery camera

ABCDEHIJKLMNTUVWXYZAAABACADAEAFAGAHAIAJAKALAMANAOAPAQAR
1CameraMain Video StreamViewAudio & AlertsStorageEnvironmentalConnectivityPower
2ModelSynology SupportSpecsStyleSpecial Feature at a glanceSensorGlassIR CutCodecResolutionFPSMax BitrateDiagonal FOVPan-HTilt-VIR Distance (m)Color NV Distance (m)MicSpeakerSpotlightAlarmSD Card Recording ModesMax SDCard (GB)Reolink CloudPlacementProtectionMinMaxWarranty (years)DetectionWirelessBandsSecurityRTSPUSB Power Req.Built-in CapacitySolar Capable?
3Argus 2NSpecsPuckStarlight sensor, 2-way audio, Digital Zoom (6X)1/2.7"'Fixed lens'StarlightH2641920x108015?13010YesYesNoYesMotion64YesIndoor/OutdoorIP65-10C55C2Motion802.11n2.4WPA2-PSKNo2A5200 mAhYes
4Argus 2ENSpecsPuck2-way audio?'Fixed lens'?H2641920x108015?12010YesYesNoYesMotion128YesIndoor/OutdoorIP65-10C55C2Motion802.11n2.4WPA2-PSKNo2A5200 mAhYes
5Argus ProNSpecsPuckStarlight sensor, 2-way audio, Digital zoom (6X)?'Fixed lens'StarlightH2641920x108015?13010YesYesNoNoMotion64YesIndoor/OutdoorIP65-10C55C2Motion802.11n2.4WPA2-PSKNo2A5200 mAhYes
6Argus EcoNSpecsBullet2-way audio?'Fixed lens'?H2641920x108015?10010YesYesNoNoMotion64YesIndoor/OutdoorIP65-10C55C2Motion802.11n2.4WPA2-PSKNo2A5200 mAhYes
7Argus PTNSpecsPTZPan, Tilt, Zoom, Starlight Sensor, Digital zoom (6X)?'Fixed lens'StarlightH2641920x108015?10535514010YesYesNoNoMotion64YesIndoor/OutdoorIP64-10C55C2Motion802.11n2.4WPA2-PSKNo2A6500 mAhYes
8Argus 3NSpecsPuckColor night vision, 2-way-audio?'Fixed lens'StarlightH2641920x108015?1201010YesYesYesYesMotion64YesIndoor/OutdoorIP65-10C55C2Motion802.11n2.4WPA2-PSKNo2A5200 mAhYes
9Argus 3 ProNSpecsPuckAI, Color night vision, spotlight, Digital zoom (16X)1/3"'Fixed lens'?H2652560x144015?1221010YesYesYesYesMotion128YesIndoor/OutdoorIP65-10C55C2Person, Car, Motion802.11n2.4/5WPA2-PSKNo2A6500 mAhYes

[Back to Top]

An observation about night vision

Most security cameras offer some kind of black-and-white night-vision mode (some even offer color-night-vision but I don’t have a model to test). The way they work is by using IR illuminators to light up the scene with ‘invisible’ light. There are two broad types of IR illuminators:

  • 850nm: much more common, brighter, generally better image, the LEDs will have some glow

  • 940nm: about half the illumination range, dimmer but the LEDs are effectively invisible

Just how visible is the red glow from the IR illuminators? Certainly not enough to overtly draw attention to themselves, but if you’re looking for a telltale glow and have an idea of where to look, you’ll spot it.

View fullsize IR LEDs On (8-bit)
 IR LEDs On (8-bit) View fullsize IR LEDs Off (8-bit)
 IR LEDs Off (8-bit) View fullsize IR LEDs On (10-bit)
 IR LEDs On (10-bit) View fullsize IR LEDs Off (10-bit)
 IR LEDs Off (10-bit)

There’s no need to worry about whether to go with 850nm or 940nm though — cameras need to be built from the ground up around 940nm illumination and I wasn’t able to find any well-documented, consumer (or prosumer, or commercial for that matter) oriented cameras built around 940nm. If you really don’t want the glow, turn off the IR LEDs and instead, add supplemental lighting (i.e. turn on your porch lights).

[Back to Top]

Step 4: Integrating with Synology (or another NAS)

So I ordered a single RLC-820A for testing purposes and I thought if I was happy with the camera, I would just order a whole bunch and hook them up to my NAS (after buying the licenses of course!). Something about this process felt a little too easy so I wanted to check a few other things.

There are three constraints that are interconnected

  1. Each NAS will have a limit on the number of cameras you are allowed to connect; most units will come with 2 licenses and additional licenses are available to purchase, up to the total connection limit.

  2. Based on the hardware/performance of each NAS, there is a budget (measured in fps) for a given resolution of camera that will be supported. The budget is split according to the video codec as well. Generally, for a given resolution, H265 will give you a bigger budget.

  3. Based on the available space available in the NAS (or at least the space available to surveillance station), there will be a limit to the number of days backlog you can store before having to overwrite older footage.

IJKLMNOPQRSTUVWXYZAAAB
1Populating NAS with Drives (TB)10
2
3DS1817+ with DX517DS1821+DS1621+DS1520+DS920+DS720+DS420+DS220+DS1817+ with DX517DS1821+DS1621+DS1520+DS920+DS720+DS420+DS220+
4Total Camera Connection Limit4040404040402525# Drive bays58654242
5H265Total 4K Framerate Limit480480480300300200200200RAID Configuration55555151
6Total 2.7K Framerate Limit760600600600600480250250Total capacity with 10TB drives4040304020102010
7H264Total 4K Framerate Limit160320320200200170120120
8Total 2.7K Framerate Limit280440440360360280240240
9Total 1080 Framerate Limit120012001200120012001200750750
10
11
12FRAME RATE BUDGET
Based on the framerate budget available per NAS, how many of each type of camera could we support (Red = #cameras exceeds the license-limit, Yellow = #cameras is exactly the license limit, Green = #cameras is under the license limit). Note that you can increase the number of supported cameras (subject to the license limit) by reducing the capture resolution.		DISK SPACE BUDGET
Assuming we allocate the entire available NAS budget to surveillance and we use the maximum number of cameras supported (the lesser of the license limit and the framerate limit), how many days of footage can we get. These numbers are just a rough example: the disk space requirements for high-activity daytime scnes will be substantially more than no-activity nighttime scenes. This will also change if we use different drive sizes or a different RAID level (or just allocate less space for surveillance) or if the cameras are not configured for continuous 24x7 recording. Red = less than a week of coverage, Yellow: 1-week coverage, Green = 1-month coverage
13
14POE CamerasDS1817+ with DX517DS1821+DS1621+DS1520+DS920+DS720+DS420+DS220+DS1817+ with DX517DS1821+DS1621+DS1520+DS920+DS720+DS420+DS220+
15# of supported RLC-410 cameras11171714141199With 11x RLC-410 cameras4429213517112713
16# of supported RLC-423 cameras914141212988With 9x RLC-423 cameras5435264020133015
17# of supported RLC-520 cameras914141212988With 9x RLC-520 cameras5435264020133015
18# of supported RLC-510A cameras914141212988With 9x RLC-510A cameras5435264020133015
19# of supported RLC-520A cameras914141212988With 9x RLC-520A cameras5435264020133015
20# of supported RLC-522 cameras914141212988With 9x RLC-522 cameras5435264020133015
21# of supported RLC-511 cameras914141212988With 9x RLC-511 cameras5435264020133015
22# of supported RLC-810A cameras1919191212888With 19x RLC-810A cameras2626194020153015
23# of supported RLC-820A cameras1919191212888With 19x RLC-820A cameras2626194020153015
24# of supported RLC-822A cameras1919191212888With 19x RLC-822A cameras2626194020153015
25# of supported RLC-811A cameras1919191212888With 19x RLC-811A cameras2626194020153015
26
27POE Cameras
28# of supported RLC-410W cameras914141212988With 9x RLC-410W cameras5454405427133015
29# of supported RLC-510WA cameras914141212988With 9x RLC-510WA cameras5454405427133015
30# of supported RLC-511WA cameras914141212988With 9x RLC-511WA cameras5454405427133015
31# of supported RLC-511W cameras914141212988With 9x RLC-511W cameras5454405427133015
32# of supported E1 Outdoor cameras11171714141199With 9x E1 Outdoor cameras5454405427132713
33# of supported E1 cameras1829292424181616With 18x E1 cameras7268516834184020
34# of supported E1 Pro cameras1422221818141212With 14x E1 Pro cameras6951385427174020
35# of supported E1 Zoom cameras1422221818141212With 14x E1 Zoom cameras6951385427174020
36# of supported Lumus cameras8080808080805050With 19x Lumus cameras1021027710251265126
37
38POE Cameras
39# of supported Argus 2 cameras8080808080805050Battery only cameras do not offer continuous recording.
40# of supported Argus 2E cameras8080808080805050
41# of supported Argus Pro cameras8080808080805050
42# of supported Argus Eco cameras8080808080805050
43# of supported Argus PT cameras8080808080805050
44# of supported Argus 3 cameras8080808080805050
45# of supported Argus 3 Pro cameras5040404040321616

While there is nothing you can do to exceed the total connection limit, there are three ways you can make better use of the frame-rate and disk-space budget:

  1. Reduce the capture frame-rate for the camera. This can be a bit of a balancing act: if you go too low, the footage will be begin to be too choppy to be useful

  2. Reduce the bitrate used. Again, we also want to do this in moderation otherwise the footage will be too noisy to be useful

  3. You can transition your cameras from continuous recording to motion-triggered, scheduled-continuous only or a combination

At the end of the day, you need to determine how much footage backlog you want to capture and retain. For some users (more commercial), this may also tie into how fast they can get footage off-site as well.

[Back to Top]

Making the most of the integration

Synology Surveillance Station (SSS) has a few neat features to explore

  1. Smart Timelapse: continuous recording ends up mostly being small blips of activity in a sea of ‘nothing happening’. With a time-lapse, you can get a ‘fast-fwd view’ of a period of time allowing you to compress overnight footage into a 2 minute clip that you can review with your morning coffee. The smart aspect comes in the way of allowing you dynamically slow down the footage when motion is detected and speed back up when it’s just a ‘sea of nothing happening’

  2. Edge Recording: the nice thing about network recording is everything just magically gets recorded to a share — but what happens if there is a network drop? With edge recording (and a supported camera with an SD card), the missing footage is automatically backfilled from the SD card to the NAS when the connection is restored. Sadly, my camera doesn’t support this feature — it seems like this is only available when you have pricier commercial/government oriented cameras. If this is a feature that is important to you, make sure to check beforehand

  3. LiveCam: one of the neat things about using a NAS as an NVR is that you have all your footage centralized into one location. But what if you don’t have coverage of a certain spot, or you want to take more detailed/specific footage after an event has happened? With LiveCam you can turn your phone into a temporary IP camera to go out in the field.

  4. Live Broadcast: I’m not sure if I would have a use for this but theoretically you could use a security camera to broadcast to YouTube. This feels like it’s a solution looking for a problem, but nonetheless, is pretty neat

[Back to Top]

Dedicated NAS?

I have a perfectly functional DS1817+ so this begs the question of why I might be looking at another NAS. There are three main drivers for this

  1. Reduce drive thrashing: I use my main NAS pretty extensively (both read and write) and sadly, I don’t have SSD caching. Surveillance is a constant sequential write operation which means, if I’m using my NAS for other things, there will be back and forth seeking around. Not an issue for one or two cameras, but perhaps more of an issue as I add more cameras

  2. Transition to surveillance drives: I use NAS ‘optimized’ drives in my NAS which is a fancy way of saying they are rated for continuous operation (from a vibration and endurance perspective). Surveillance optimized drives are also rated for continuous operation but also (supposedly) are also optimized for continuous sequential write access rather than ‘general purpose’ (once setup and initial testing are done, surveillance is 99% write, with only intermittent read)

  3. Segment the workflow: longer term, I’d like to explore automating a quick encode and pushing footage offsite and automated with life-cycle management; by having a separate device, I can keep everything related to this in one space. Of course, I this would allow me to leverage NAS to NAS backups as well

None of these are deal breakers so I’m not in a huge rush so this will be something to explore down the road.

[Back to Top]

Conclusions

My biggest takeaway is to make sure you thoroughly explore the compatibility of your camera with your NAS (if you’re going with that route) - I would have found out up front that Edge Recording wasn’t available on my camera and would have avoided the disappointment when I realized it’s effectively out of my price point.

Obviously, my next step is installing a few cameras and going from there. A tool like IPVM is super handy for this: it leverages Google Maps and Streetview to help you plan out your layout - you can even select specific camera models and it will dial in the field of view etc. Once some of the cameras are setup, I need to explore how to automate encodes and life-cycle management on the retained footage, stay tuned for a follow up!

[Back to Top]

]]>A Dive into Security Cameras, Part 1Guide: Video Encoding, Part 2 - Deep DiveAVDPMon, 08 Mar 2021 22:46:27 +0000https://www.minmaxgeek.com/blog/encoding-guide-26020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:603a9bd1a3f9d81e9721136aEach and every year, manufacturers in the consumer electronics space try to outdo each other, all promising to take our videos to the next level. For the most part, they all tend to deliver on these promises and their headlining features tend to mirror each other:

  • More resolution

  • More dynamic range

  • Wider color gamut, more colors

  • More frame-rate options and more extensive slow-motion

  • Better and more immersive audio

  • Log or raw video recording

  • Gimmicky features like 360° or 3D video

All of this is incredible - but it comes at the cost of larger and larger file sizes. Of course, we expect that as we take more videos, we’ll need more space to store all the footage but there are things we can do to mitigate the storage cost of our media.

I wrote this guide because nearly everything else I found was focused for streaming or recording footage — not what to do with footage that already exists with a goal of long-term archival in mind.

Quick Links

Planning

As I talked about in the previous post, every codec is a compromise between compression, quality, speed and ease of playback. Every scenario is different but since we have to start somewhere, I’ll describe my scenario and walk through the process I took.

At a high level, I’m looking for a way to reduce the storage requirements for my content without taking a substantial hit to image quality.

Any encoding or transcoding will result in image quality loss, but if I can keep the output image quality within a [subjective] margin of the input video, that is acceptable. I will of course consider advantageous gains favoring either image quality (i.e., 1% bigger file size, but 30% better image quality) or compression (1% worse image quality, but 30% smaller file size).

My scenario

  1. My content is generally high-quality

    • Currently, mostly encoded with high-bitrate H264; moving forward, more content will be captured with H265

    • Currently, mostly 1080P; moving forward, there will be a greater focus on higher resolutions

    • Currently almost all of my content is 8-bit and moving forward, there will be more 10-bit content

  2. I have a substantial workload

    • At my starting point, there was something in the range of 15,000 files to work through

    • Enough breathing room on storage that this can be a multi year project if needed however

    • Increasing the available storage capacity is not an option

  3. Reasonably performant hardware (for the time)

  4. Image quality is important: I accept the initial one-time hit that comes with any transcode, but it must pass the "‘eyeball test”

  5. Use cases

    • Encoded content must be accessible in the current time-frame

    • Content is entirely for my consumption so I don’t need to worry about whether or not others have decode acceleration or not; on the off chance I upload to something like YouTube - their servers will re-encode the video back again so that the masses can watch it

Candidate Codecs

  • H264 (CPU/GPU)

    • Since my current backlog is mostly H264, I won’t see any meaningful compression gains going with H264

    • Transcoding H265 content to H264 may be something that some people need to consider if their playback devices don’t have hardware acceleration to decode H265

  • H265 (CPU/GPU)

    • H265 promises up to 50% compression advantage at ‘the same’ image quality so this is a no-brainer to at least consider

    • Even for re-encoding the existing my H265 content at lower settings, I can leverage the advanced nature of the codec to go after large compression gains for relatively small quality losses

    • H265 encoding is noticeably slower than H264

    • Decode acceleration is generally widely available for H265

  • AV1 (CPU only)

    • AV1 promises up to 67% compression advantage versus H264 at ‘the same’ image quality

    • I’ve heard that AV1 can be quite a bit slower than even H265 encoding.

    • As of Q1-2021, there isn’t much in the way of decode acceleration for AV1 — only the latest generation of graphics cards have acceleration for AV1. If the compression levels are substantial enough, I may be willing to overlook this (since most of my playback devices have enough horsepower to decode AV1 without the need for dedicated acceleration)

For my specific use case, the front-runner will be H265 and it’s just a matter of doing some tests to see what settings make sense and whether using the GPU to do the heavy lifting is a viable option.

Testing Scenario

The ultimate goal is to identify a combination of encoder, codec and settings that substantially compresses my videos down without too much of a hit to image quality and without taking forever. In an ideal world, we would have infinite computing power and could just check all of the possible permutations…. Back in reality, using a representative clip to test as many configurations as you find feasible is a good stand-in. In my case, to get an ‘apples-to-apples‘ comparison, I started with a video trailer:

  • Available in all of the common resolutions: 480, 720, 1080, 2160

  • Nearly identical clip across all of the resolutions (3504-3660 frames) showing the same content

  • All of the clips were 8-bit, high-bitrate H264/AVC which is representative of the bulk of my content

I ran through a wide selection of transcode scenarios: I muxed the audio stream (aka ‘no processing’) to isolate this to just the video workload (generally audio encoding is super-fast anyways). For each clip

  • GPU — NVENC: H264 and H265 using 6 quality settings each

  • CPU — H264, H265 and AV1 using 10-30 quality settings each

The end goal is to objectively identify three properties:

  1. Compression performance: expressed as a % of how much smaller the output file was, compared to the original source - there are some cases where the output file can be bigger than the input file, so that’s something to watch out for

  2. Encode performance: expressed as an averaged frames/sec; each frame will take a different amount of time to encode but this will give us a starting point. Generally speaking, 24 fps is real-time (since most films are shot at that frame rate); if I can encode at 24 fps, this means a 30-minute clip will take about 30-minutes to encode.

  3. Calculated image quality: expressed as an overall average VMAF score, this attempts to quantify the perceived image quality of the entire clip. I actually collected PSNR, SSIM and VMAF data for each frame, in each clip, in case I want to dive deeper.

The ultimate goal is to find a combination of encoder, codec and settings that allows me to compress my footage down without too much of a hit to image quality and without taking 10x-realtime. I’m specifically looking with long-term archival in mind (rather than streaming or uploading to YouTube etc.)

 

Limitations

I use the same machine for all of the tests (see here for Q1-2021); unfortunately my current configuration does not give me access to AMF or QuickSync GPU encoding without fundamentally changing the machine. I suspect that:

  • NVENC is the most performant of the three major GPU options (it’s certainly the best-supported)

  • General conclusions I arrive at with NVENC will apply to AMF and QuickSync

I may revisit AMF/QuickSync (using fundamentally different hardware) as an addendum.

 

A word about Image Quality

I used the VMAF 4K data set when analyzing the encoded output files. Normally the 4K model is reserved for 4K content (duh) but with 4K (or higher resolution) displays becoming more and more prevalent, it made more sense to use the 4K model to compare all the results. After all, if you happen to be watching 1080P content, you’re not going to put your 4K TV away and go grab a 1080 display to watch it!

[Back to Top]

Data Collection

GPU Encoding - NVENC (H264 & H265)

NVENC is the hardware encoder that is included with nVidia-based graphics cards. Exactly what codecs and details supported by NVENC will vary depending on the specific generation of hardware you have: you can find the list of supported codecs for encode and decode acceleration here.

View fullsize H264_500x500_Squooshed.png
 View fullsize NVENC H264 Compression
 NVENC H264 Compression View fullsize NVENC H264 Performance
 NVENC H264 Performance View fullsize H265_500x500_Squooshed.png
 View fullsize NVENC H265 Compression
 NVENC H265 Compression View fullsize NVENC H265 Performance
 NVENC H265 Performance

Generally speaking, since our starting point is already H264, there’s no reason to opt for H264 if you have access to H265. There are two notable exceptions that apply to older/weaker playback devices:

  1. If you need to resize (usually, downscale) a video so that the weaker device can play it or,

  2. If your starting point is a more advanced codec like H265 or AV1 and the playback device struggles to decode it

In both of these cases, I might suggest that compression and (to a degree), image quality don’t matter since you’re likely going to be transcoding just so you can play the video ‘just this once’. Over time, as your playback device gets more powerful, you’ll be able to easily play the original video clip.

 
 

Data Deep Dive

With GPU encoding, there is always the risk that your output file ends up larger than your input file. If you’re transcoding from a more complex codec to a simpler one, that is to be expected but in this data set, I’m never doing that. It is easier to run into this scenario at lower resolutions — but if you crank the quality setting up, you can run into this, even at higher resolutions. It’s important to consider your priorities (compression, image quality, speed etc.) at the specific resolution of your content.

The following data deep dive is a collation of 48 encode runs capturing the following permutations:

  • 4 resolutions (480, 720, 1080, 2160)

  • 2 codecs (H264, H265)

  • 6 quality settings (15, 18, 21, 23, 27, 30) — (Note: the lower the number, the higher quality)

View fullsize NVENC - H264 vs H265
 NVENC - H264 vs H265 View fullsize Compression Performance: H264 vs H265 (GPU)
 Compression Performance: H264 vs H265 (GPU) View fullsize Encoding Performance: H264 vs H265 (GPU)
 Encoding Performance: H264 vs H265 (GPU) View fullsize Encoding Quality: H264 vs H265 (GPU)
 Encoding Quality: H264 vs H265 (GPU) View fullsize NVENC - H264 vs H265
 NVENC - H264 vs H265 View fullsize Compression Performance: H264 vs H265 (GPU)
 Compression Performance: H264 vs H265 (GPU) View fullsize Encoding Performance: H264 vs H265 (GPU)
 Encoding Performance: H264 vs H265 (GPU) View fullsize Encoding Quality: H264 vs H265 (GPU)
 Encoding Quality: H264 vs H265 (GPU) View fullsize NVENC - H264 vs H265
 NVENC - H264 vs H265 View fullsize Compression Performance: H264 vs H265 (GPU)
 Compression Performance: H264 vs H265 (GPU) View fullsize Encoding Performance: H264 vs H265 (GPU)
 Encoding Performance: H264 vs H265 (GPU) View fullsize Encoding Quality: H264 vs H265 (GPU)
 Encoding Quality: H264 vs H265 (GPU) View fullsize NVENC - H264 vs H265
 NVENC - H264 vs H265 View fullsize Compression Performance: H264 vs H265 (GPU)
 Compression Performance: H264 vs H265 (GPU) View fullsize Encoding Performance: H264 vs H265 (GPU)
 Encoding Performance: H264 vs H265 (GPU) View fullsize Encoding Quality: H264 vs H265 (GPU)
 Encoding Quality: H264 vs H265 (GPU)

Conclusions

  1. If you use NVENC for encoding, in general, it makes the most sense to leverage H265 on the higher side of the quality setting — there’s effectively no performance hit (system or encode) and H265 will let you push the quality a bit higher while still maintaining compression.

  2. If you are streaming or doing anything of that nature, going with a GPU encode is almost a predetermined fact just so that you don’t have the hit to system performance.

  3. GPU encoding is a viable option if you need a predictable encode time; for example, if you are compressing security camera footage to reduce the amount of time to upload footage offsite, you don’t want any time-savings from uploading smaller files to be lost by spending forever with an encode!

  4. If you have a relatively weak PC in general then, GPU encoding may be the only feasible option

[Back to Top]

CPU Encoding - H264, H265 and AV1

When encoding with the CPU, you can pick from almost any codec - the only unavailable codecs are those specifically made for GPU encoding. For my scenario, I was considering H264, H265 and AV1. H264, sometimes referred to as AVC is a fairly mature codec now; this means there are a ton of devices that ship with decode and encode acceleration. (As of 2021), H264 is the default codec of choice for the majority of phones, cameras, drones, dash-cams, doorbell-cams etc. H264 offers a great blend of compression, performance and quality even as ‘next generation’ codecs like H265 and AV1 make their way onto the stage. These more advanced codecs promise an optimistic 50% and 66% increase in compression while retaining the same image quality — at the cost of computational complexity.

In my case, the majority of my content is already encoded using H264, so at first glance it may not make sense to consider H264 as an option but there are different ways to encode with a codec and the data deep dive will show that we can still extract quite a bit of compression by changing a few things.

CPU encoding brings an overwhelming number of configuration options. Just like the GPU encoding process, there is a ‘quality index’ value (where lower values yield better quality) but there are also quality presets or profiles (Fast, Medium, Slow) which toggle a multitude of options at once. Generally speaking these options tweak how much time the encoder spends analyzing a scene with the idea that the more time the CPU spends, the better the image quality. There is also a fairly loose correlation with compression as well (longer analysis time = better compression). This first data set collates the metrics (compression & speed) by codec and the general ‘quality index’.

View fullsize H264_500x500_Squooshed.png
 View fullsize H264 @ CRF18 - Compression
 View fullsize H264 @ CRF18 - Performance
 View fullsize H264 @ CRF22 - Compression
 View fullsize H264 @ CRF22 - Performance
 View fullsize H264 @ CRF26 - Compression
 View fullsize H264 @ CRF26 - Performance
 View fullsize H265_500x500_Squooshed.png
 View fullsize H265 @ CRF18 - Compression
 View fullsize H265 @ CRF18 - Speed
 View fullsize H265 @ CRF22 - Compression
 View fullsize H265 @ CRF22 - Speed
 View fullsize H265 @ CRF26 - Compression
 View fullsize H265 @ CRF26 - Performance
 View fullsize AV1_500x500_Squooshed.png
 View fullsize AV1 @ CQ24 - Compression
 View fullsize AV1 @ CQ24 - Speed
  
 

Data Deep Dive

With CPU encoding, you have an overwhelming number of options to configure and while it’s possible to select a configuration where the output file ends up being bigger than the source file, that’s much less likely to happen as the entire process is much more efficient overall. As before, it makes the most sense to look at the metrics (compression, speed, quality) with a specific target resolution in mind.

The following data deep dive is a collection of 280 encode runs capturing the following permutations:

  • 4 resolutions (480, 720, 1080, 2160)

  • 3 codecs (H264, H265, AV1)

  • 3 quality levels (18, 22, 26) for H264/H265, only 1 quality level for AV1 (24) — (Note: the lower the number, the higher quality)

  • 10 quality sublevels (Ultrafast, Super Fast, Very Fast, etc.)

View fullsize 480_500x500_Squooshed.png
 View fullsize CPU Encoding Compression @ 480P
 CPU Encoding Compression @ 480P View fullsize CPU Encoding Speed @ 480P
 CPU Encoding Speed @ 480P View fullsize CPU Encoding Quality @ 480P
 CPU Encoding Quality @ 480P View fullsize 720_500x500_Squooshed.png
 View fullsize CPU Encoding Compression @ 720P
 CPU Encoding Compression @ 720P View fullsize CPU Encoding Speed @ 720P
 CPU Encoding Speed @ 720P View fullsize CPU Encoding Quality @ 720P
 CPU Encoding Quality @ 720P View fullsize 1080_500x500_Squooshed.png
 View fullsize CPU Encoding Compression @ 1080P
 CPU Encoding Compression @ 1080P View fullsize CPU Encoding Speed @ 1080P
 CPU Encoding Speed @ 1080P View fullsize CPU Encoding Quality @ 1080P
 CPU Encoding Quality @ 1080P View fullsize 2160_500x500_Squooshed.png
 View fullsize CPU Encoding Compression @ 2160P
 CPU Encoding Compression @ 2160P View fullsize CPU Encoding Speed @ 2160P
 CPU Encoding Speed @ 2160P View fullsize CPU Encoding Quality @ 2160P
 CPU Encoding Quality @ 2160P

Conclusions

  1. AV1 is a performance hot mess. I was originally going to run three different passes, but it just doesn’t scale to the number of available cores. I did try some other AV1 encoders (both Rav1-e and SVT-AV1) but I ran into some fussy issues with my workflow. As I cranked the quality dial on AV1, it became more appropriate to look at performance as frames-per-minute (!!).

  2. If you have the horsepower and the time to encode, CPU is generally a no-brainer: you get quality and compression. In particular, H265 @ Medium likely is the sweet spot for many people. At 1080 for example, you get approximately real-time encoding with very good compression

  3. The maturity of the H264 codec is commendable; again, H264 @ Medium (or Slow) might be a great sweet spot for a lot of people if you don’t have as much horsepower to power through H265

  4. H264 is surprisingly performant at lower quality settings, that it may be a reasonable stand-in for GPU encoding. If you can take the overall system performance hit (and likely a quality hit), you can certainly get speeds greatly exceeding GPU encoding.

  5. All of this performance is extremely sensitive to the hardware you have at your disposal. H264 and H265 both scale exceptionally well to the number of threads you have at your disposal

[Back to Top]

TL;DR: What did I end up going with?

While I do suggest that each user come up with their own criteria for success based on their overall objectives and use representative sample videos to do some initial testing, sometimes a recommendation can be a good starting point. At the time of writing, I’ve done through about 3000 encodes.

  • For almost all of my encode jobs, I use CPU encoding with H265 @ CRF22 with the Slow preset

 

Example:

--crf 22 --preset slow --profile main --rskip 0 --refine-mv 3 --hevc-aq --aq-motion --hist-threshold 0.01 --pmode --pme --colorprim bt2020 --colormatrix bt2020nc --transfer smpte2084 --range limited --opt-cu-delta-qp --rect --high-tier

 

  • I use the Matroska file container for all of my content

  • I don’t transcode the audio streams, I just copy/mux the existing stream into the output. This is reduces the amount of verification I have to do and simplifies my workflow - I don’t need to worry about changing the number of channels, bitrates etc.

  • I do make some minor tweaks to my baseline flow for 10-bit or animated (cartoon) content.

For real-world actual encode jobs, I generally see 50-90% compression with the overwhelming bulk being between 80-85% compression (typically, low resolution & low quality content doesn’t compress well).

[Back to Top]

Closing thoughts

Over the years I’ve made a few attempts at building a data-set like this but I’ve always given up well before completing because it just took so long and I didn’t have a compelling reason to continue on with the painful process of collecting and collating all of the data. A few things have changed over the years:

  1. I’ve got a lot more computing horsepower now (relative to the workload)

  2. Getting good sample videos across all of the resolutions is easier: movie trailers.

    • They come in a variety of resolutions ,

    • Are generally short,

    • Can be found (for each resolution) with the same starting codec,

    • Generally feature a variety of scenes (action, lighting etc.)

  3. I have motivation now: phones and cameras continue to produce bigger and better footage - my available disk space isn’t really growing that fast, so I need to make the most of it

 
 

I set out to find a codec/configuration for long-term footage archival, so: ‘encode once, store forever’. Any time-cost spent transcoding footage would have forever-gains in reduced disk footprint. This gives me a bit more tolerance for slower codecs and settings that might optimize for compression and/or image quality.

This project had two really big eye openers: the performance of H264 and AV1. H264 because of the ludicrously high performance with software encoders and AV1 for the ridiculously slow performance under all scenarios. I suppose the performance gains for H264 make sense — it’s been almost 10-years since I last explored this video compression stuff and the maturity of the codec and the relative hardware has come a long way.

AV1 performance though: I starts bad and gets worse as you climb the image quality ladder. Here are actual encode times for encoding a 2min26sec clip — 3660 frames @ 4K for each codec (using the highest quality settings for each)

  • NVENC H265 @ CQP15: 93 seconds (1 min 33 sec)

  • NVENC H264 @ CQP15: 99 seconds (1 min 39 sec)

  • AV1 @ CQ24 + Fastest: 1357 seconds (22 min 37 sec)

  • CPU H264 @ CRF18 + Placebo: 1510 seconds (25 min, 10 sec)

  • CPU H265 @ CRF18 + Placebo: 3158 seconds (52 min, 38 sec)

  • AV1 @ CQ24 + Slowest: 54124 seconds (15 hours, 2 min, 4 sec)

Yes. It took me over 15 hours to encode a 2min26sec clip. Even at its best settings (least image quality), it only slightly edges out H264 at maximum quality.

I don’t have enough exposure with AV1 to dig into this further (and certainly don’t have the willpower or motivation to learn more) but at first glance there was almost no thread scaling. H264 and H265 both saturated my processor to 100% load but AV1 spent the entire time at a paltry 20% load. Even if we take an overly optimistic penalty of 5x, the AV1 encode might have taken 3.5 hours - still 3X+ slower than H265.

If we take a closer look at the AV1 performance, specifically for higher resolutions (there’s no reason to consider a codec like AV1 for low-resolution content), we can see just how poor the performance is. Even the absolute fastest AV1 barely beats the absolute slowest H265!

AV1 @ 2160
AV1 @ 2160

That number, 0.065 — comes out to 234 frames per hour.

AV1 @ 1080
AV1 @ 1080

Even if we cut the resolution demands down to 1080 - a 75% reduction in workload, that number - 0.095 fps … that’s 342 frames per hour.

[Back to Top]

FAQ

  1. What encoding software do you use? I use StaxRip.

  2. How did you calculate the quality scores (VMAF)? You can use ffmpeg to calculate the scores directly, but I used a tool called FFMetrics in conjunction with a couple quick parsing utilities I threw together

  3. What about Handbrake? Handbrake is great: it’s free, super friendly and easy to use. In my personal experience, I’ve found that I don’t get full CPU saturation out of it (i.e. it runs slower). I’ve heard some rumblings about Handbrake being thread-limited — I don’t know if there is any truth to it, but since I have a working solution, I haven’t had the need to look into this further.

  4. What hardware do you run? At the time of this post, see these specs (Fury: Q1-2021)

  5. What about AMF? or Quick Sync? At the time of writing, I don’t have a modern AMD graphics card (one with hardware encoder for H265) so I can’t test that (although I’ve heard that the encoding experience isn’t so hot). Changing the CPU to an Intel one (to get Quick Sync) would fundamentally change the characteristics of this machine so the metrics would be even less comparable.

  6. What about codec XYZ? At this time, I don’t have a use for codecs outside of the candidates selected for this test. In the longer term, I may look at H266 or AV2 which both promise generational improvements over H265

  7. Will I revisit AV1? Unlikely unless spot testing show monumental performance gains.

  8. Are there any downsides to choosing H265? If your target playback device is a modern computer or phone, you should be fine; if your device is a cheap media box, then you may need to look at the specifications to see if the chip has decode acceleration (and in some cases, you’ll need to look further to check for resolution and color compatibility). Keep in mind that over time, more and more devices will come with H265 decode acceleration as well.

  9. Why matroska and not MP4? Matroska is an open container standard that works pretty well — if your player/device can’t support it, get a better player/device. On a [more] serious note, MP4 is probably fine - I just dislike it as a personal preference.

  10. What about 10-bit (or beyond) video? Unfortunately, I wasn’t able to find a set of trailers that had a span of resolution and bit-depth and I didn’t want to try and compare trailers for different movies (since different scenes may compress better or worse). In general, 10-bit videos will require setting --profile main10 --output-depth 10 (and so on and so forth for 12-bit etc.) and generally the encode tasks will hit a bit harder. I imagine that most of the same patterns covered by the data in this post should apply.

    [Back to Top]

 
 

Related

If this post went a bit too deep (but you’re still interested in video encoding), check out the ‘starter guide’ which breaks down the handful of terms you need to know and how all the pieces fit together.

Read More
 ]]>Guide: Video Encoding, Part 2 - Deep DiveGuide: Video Encoding, Part 1 - The BasicsAVDPThu, 04 Mar 2021 02:55:00 +0000https://www.minmaxgeek.com/blog/encoding-guide-16020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:603c70c4c029a67fb14b56e2Almost every time you pull out your smartphone to record a vlog, sit down in front of a webcam to stream, or roll out a cinema cam to record a music video or commercial, the footage you’re capturing is getting encoded and sometimes more than once. But what exactly does it mean when we talk about videos being encoded?

Video encoding can get really complicated in a hurry and the answer to “what’s best” will almost always be it depends. Before I can dive into the numbers and metrics, this post is to setup all the basics: the terminology and workflow of how the different pieces fit together.

Quick Links

Terminology

Encoding

Abstractly, this refers to the entire process of writing out a video file

  • Using a specific scheme (codec)

  • In a specific way (container)

  • With specific settings (frame-rate, color depth, sample rate, etc.)

  • With or without audio (the audio will have its own specific settings as well: sample rate, bit-depth, channels etc.)

  • With or without extra metadata (captions, keyword metadata, tags, embedded thumbnails etc.)

Transcoding

A portmanteau for ‘transform and encoding’, transcoding describes an encode where one or more properties of the video are being changed. Two really common reasons to transcode videos are

  1. Changing the video’s resolution to be appropriate for the playback device. This can be down scaling (4K source footage to play back on a 1080 screen) or up scaling (taking 1080 content and sizing it for 4K).

  2. For video compatibility: some phones or TVs won’t recognize certain file formats (containers) or encoding schemes (codecs), so by transcoding, we can change these attributes

 

Generally speaking, encoding and transcoding can be used interchangeably. Additionally, the term video compression is sometimes thrown into the mix as well since reducing the file size for videos is a big motivating factor for video encoding.

It’s important to note that just because you encode/compress a video, the resulting file is not guaranteed to be smaller. It’s quite possible to get output files that are substantially larger than the original video.

 

Codec

A portmanteau for coder/decoder, this represents how data in a file is stored. There are a few characteristics that different codecs apart:

  • Roughly how compressive it is (how small can we make our output file)

  • What kind of image quality can we get for a given setting

  • How easy is the encode process (how fast is the encode process)

  • How easy is it to decode

Picking the ‘best codec’ to use will be a balancing act which I’ll go into deeper in the next guide. Some examples of codecs are H264/AVC, VP9, AV1 and WVC1.

Container

A video file is often more than ‘just the video content’. A modern video file might contain a whole bunch of resources

  • The video stream data

  • Zero, one or more streams of data

  • Zero, one or more sets of subtitle or captioning data

  • Metadata like thumbnails or display tagging information

All of this data can be packaged up into a single file — the container. Containers are often described by their file extension: MKV, MOI, AVI and MP4 are some common examples.

 

In general, the container type does not dictate the codec you have to use. More often than not, you’ll want to consider the device that you intend to view the video: if it’s a basic media player box, that box may only recognize certain containers and codecs whereas a computer will generally be able to handle anything you throw at it. A service like YouTube will behave like a computer: almost all pairings of containers and codecs are supported.

  

Some examples of containers and codecs and how there doesn’t need to be any rhyme or reason with the pairing.

 

Encoders (and Decoders) & Acceleration

Fundamentally, a decoder is just something that allows a player to play a video and an encoder is something that allows a video to be encoded or written using a specific codec. Encoder and decoders can either be general purpose (software) or dedicated (hardware) implementations.

  • A software implementation will generally work ‘everywhere’ (that the software can run) and how well it works (it’s performance) will be vary depending on the amount of computing power you have at your disposal. Relatively speaking, a software solution will be much more taxing.

  • A hardware solution will generally highly performant but will require that you have a specific, physical thing. Hardware implementations are binary: either you have the required hardware component (or the right one), or you don’t. A hardware solution will always be faster than a software/CPU solution so the term hardware accelerated will be used to describe this. Often, this hardware component comes bundled with your video card so sometimes it’s described as using the GPU

To make things a bit more confusing, many times, people will refer to tools like Handbrake, StaxRip, Adobe Media Encoder, or ffmpeg as encoders. Technically speaking those are UI front ends that allow users to put together and run encode jobs. The actual encoders are usually hidden away: the user selects a codec and the application knows what encoder that it will use.

Multiplexers and Demultiplexers

I mentioned above that video files are just containers; they hold the audio and video streams, subtitles etc. Before we can begin to encode our videos, we need to be able to ‘open up the container’ — and that’s a demultiplexer/demuxer does. Conversely, at the end of an encode process, we need to ‘put everything back into a container’ — that’s what a multiplexer/muxer does.

Tip: If you just need to change the format of a file (i.e., so that an application or device can recognize it), all you need to do is demux the original file and mux it back into the file format you want. This is infinitely faster than going through an encode process.

[Back to Top]

How everything fits together

Sometimes it’s easier to see everything laid out in an example. This here is an example of the flow of transcoding a video file

  1. The incoming file is demuxed (broken into individual components)

  2. The audio and video streams are first sent to their respective decoders (blue dashed-line)

  3. The decoded stream is then sent to the appropriate encoder (green dashed-line)

  4. The encoder does it’s work on the streams (orange dashed-line)

  5. The two streams are then muxed back into the desired container (MKV)

[Back to Top]

Why encode or transcode?

Unless you’re encoding your videos using a lossless format (beyond the scope of this guide), every time you encode or transcode, you will lose quality. You will lose image quality even if:

  • You increase the quality setting of the codec or add more bitrate or,

  • Switch to a more advanced video codec or,

  • Switch to a more advanced encoder

The only time you won’t lose image quality is if you transcode to a lossless codec; even then the best you’ll do is the same image quality.

Future tech: there are some really interesting developments with respect to using AI to upscale video. When we encode an upscaled image, we’re no longer comparing apples to apples though (and every encode after that, you’ll still lose quality).

Maintaining maximum quality is not without drawbacks though: sheer disk space - just looking at some of the currently popular (and upcoming) recording modes and the ridiculous amount of data required to capture it all.

 
 

Without encoding, any time we press record, we would potentially be moments away from running out of storage! Thankfully virtually every camera has very efficient, built-in encoders. In ideal conditions, compression ratios in the 1:1000 range are possible. It’s because of these built-in hardware encoders that we are able to vlog at 4K using our smartphones!

Even with these astonishing compression ratios coming out of phones and cameras, if you capture a lot of video, the total storage footprint eventually becomes a problem. There are a bunch of obvious scenarios where you need to stay on top of your storage needs:

  • 24x7 security camera footage that you might need to hold on for a certain duration

  • Dash-cam footage from your fleet of vehicles that you might need to hold on for a certain duration

  • You’re an active streamer or YouTuber and you want to keep a copy of your videos in case something happens

  • You film coursework for online learning

  • You want to get into film-making and have a ton of multi-angle/B-roll footage

  • You like to vlog regularly

This is where we start to look at transcoding our content - to take that one-time (hopefully small!) image quality hit to substantially reduce the storage footprint.

[Back to Top]

Picking a codec (and encoder)

Every codec is a compromise of compression, quality, speed and ease of playback; the impact of these compromises will vary depending on your specific needs. To start,

  • How much compression do you need?

  • How long will you keep this video for?

  • How will the video be consumed now?

  • How critical is image quality?

  • What resources do you have available to encode?

The first step is to categorize whether or not GPU encoding is a good fit. There are some easy options:

  • Streamers: short of a dedicated recording machine, almost every other scenario will likely be best served by GPU encoding

  • Content creators: this will vary depending on how fast you have to get content delivered and whether or not you will ever go back to reference it later, perhaps as B-roll. In most case, at higher settings, GPU encoding will offer enough compression and performance to justify any hit to quality

  • Security or dash-cam footage: if the priority is getting the footage offsite, then GPU encoding is a no-brainer; the encoding time spend crushing down the file size will be more than paid back on savings uploading the file

CPU vs GPU

CPU encoding will categorically give you a better image quality and almost certainly give you a substantially more compressed file. You also have more codecs to choose from (that don’t need ‘special’ hardware) and as you upgrade your equipment you can keep your existing workflow - everything just runs faster.

GPU encoding though, is stupid-fast. While you won’t get the same amount of compression as CPU encoding, you can still get enough compression that it may not matter. While the image quality will never match CPU encoding, as you work with higher resolution videos, there’s more leeway (quality-wise) in what you can get away with. This ultimately comes back to just how silly-fast the encode process is.

Depending on your specific scenario, the best suggestion is to run some tests that represent your typical convent and your typical workflow to narrow down which codec and encoder makes the best sense.

This is something I’ll cover in much more depth in a future post.

[Back to Top]

 

Related

Now that we have a basic understanding of how video encoding works, let’s take a deep dive into the question we actually want to ask — what codec should I use?

Read More
 ]]>Guide: Video Encoding, Part 1 - The BasicsWeekend Getaway to San FranciscoAVDPWed, 03 Feb 2021 05:05:00 +0000https://www.minmaxgeek.com/blog/san-fran-20166020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:60373005eda7b812ca5871e6So I got engaged in August 2016 in San Francisco: I got my girlfriend’s boss to send her to a weekend training conference in California. For my cover story, I was already out of the country the week before (in South Korea no less) for work - so reason to suspect anything. In truth, I was hiding out at friend’s house for a few days before flying out the day before to make sure everything was in place. There were a few moments of panic during the final stretch but spoiler alert: everything worked out.

This trip was a bit of an excuse to rent and try and out some nicer camera gear than we had at the time. I spent most of this trip with the Canon 6D and 16-35mm f/2.8 L USM II.

Quick Links

Thursday

Twas the night before

I flew out the day before both to make sure everything was in place and ready but also to avoid the risk of both of us being at the airport at the same time. It’s a super weird (and cool!) to be in an airport that is shut down overnight.

View fullsize Spent the night at the airport
 Spent the night at the airport View fullsize Super eerie overnight
 Super eerie overnight View fullsize An increase in 'the mount of blue'
 An increase in 'the mount of blue'

On site!

That all went down at Hotel Nikko. There was a short list of hotels that hit my Goldilocks requirements:

  • It had to be a boutique-esque hotel to fit my cover story

  • A certain distance from the airport by taxi — she would be taking a cab from the airport and I needed to have a predictable time window of her arrival; it also needed to be close enough to a train-station stop as I would be taking the train

  • Central enough that we weren’t spending too much time getting to places

  • Needed to be a tower-style since I was planning on having the window open behind me and needed light

  • The onsite restaurant needed to be something I wanted to try

Cost was also a factor: while this certainly wasn’t inexpensive, the price-spread in a two block radius was ridiculous (there were some hotels going for triple the rate).

View fullsize Checking In
 Checking In View fullsize Scenic backdrop
 Scenic backdrop View fullsize She would be checking in the next day
 She would be checking in the next day View fullsize Even the bathroom
 Even the bathroom View fullsize Incredible view
 Incredible view

[Back to Top]

Friday

So after the charade ended, we went to see the Golden Gate Park. There are more ‘exhibits’ to visit here than we had time to really visit — we really only hit up the Japanese Tea Garden. The Japanese Tea Garden was super tranquil and very much a self-contained world inside the greater park.

Golden Gate Park

View fullsize Conservatory of Flowers
 Conservatory of Flowers View fullsize IMG_7199_1343x2000_Squooshed.jpg
 View fullsize Gorgeous fountain
 Gorgeous fountain View fullsize Shakespeare Garden
 Shakespeare Garden View fullsize Shakespeare Garden
 Shakespeare Garden View fullsize Japanese Teak Garden
 Japanese Teak Garden View fullsize IMG_7228_3334x2000_Squooshed.jpg
 View fullsize IMG_7229_3059x2000_Squooshed.jpg
 View fullsize IMG_7230_3030x2000_Squooshed.jpg
 View fullsize IMG_7231_3082x2000_Squooshed.jpg
 View fullsize IMG_7239_2023x2000_Squooshed.jpg
 View fullsize IMG_7240_1601x2000_Squooshed.jpg
 View fullsize IMG_7241_1333x2000_Squooshed.jpg
 View fullsize IMG_7250_1472x2000_Squooshed.jpg
 View fullsize IMG_7258_3000x2000_Squooshed.jpg
 View fullsize IMG_7261_1694x2000_Squooshed.jpg
 View fullsize IMG_7267_2514x2000_Squooshed.jpg
 View fullsize IMG_7282_1333x2000_Squooshed.jpg
 View fullsize IMG_7371_3245x2000_Squooshed.jpg
 View fullsize IMG_7309_2206x2000_Squooshed.jpg
 View fullsize IMG_7380_2825x2000_Squooshed.jpg
 View fullsize IMG_7391_3853x2000_Squooshed.jpg
 View fullsize IMG_7395_3607x2000_Squooshed.jpg
 View fullsize Dutch windmill
 Dutch windmill

[Back to Top]

Cupid's Span the Bay Bridge

Before dinner, we had a bit of time to visit Cupid’s Span and the Bay Bridge as luckily, it was only a few minutes’ walk from the restaurant. Dinner was at EPIC Steak and spoiler, the restaurant lived up to the name.

In a bit of a funny turn of events, I had made my reservation using OpenTable and requested a specific table and noted that this would be a post-engagement dinner. EPIC called me the day prior while I was waiting for an unrelated call from the hotel - I looked at the area code and just assumed it was the hotel and went to the front desk (and well… I don’t ever really take phone calls). Turns out EPIC was calling to double check that the engagement dinner was still a go - I’m guessing that they’ve had things go south before (ouch)!

View fullsize IMG_7410_3000x2000_Squooshed.jpg
 View fullsize IMG_7429_3286x2000_Squooshed.jpg
 View fullsize IMG_7424_2865x2000_Squooshed.jpg
 View fullsize IMG_7425_3155x2000_Squooshed.jpg
 View fullsize IMG_7422_3087x2000_Squooshed.jpg
 View fullsize IMG_7436_3412x2000_Squooshed.jpg
 View fullsize IMG_7437_3657x2000_Squooshed.jpg
 View fullsize IMG_7440_3387x2000_Squooshed.jpg
 View fullsize IMG_7445_2099x2000_Squooshed.jpg
 View fullsize IMG_7453_3745x2000_Squooshed.jpg
 View fullsize IMG_7458_3252x2000_Squooshed.jpg
 View fullsize IMG_7474_2635x2000_Squooshed.jpg
 View fullsize IMG_7482_3853x2000_Squooshed.jpg

[Back to Top]

Making our way back

At night, the Bay Bridge lights come on and it’s really very pretty (this is why I had the table-request for dinner). We spent a bit of time here struggling to take low-light photos before heading back to the hotel as the fog rolled in. The camera struggled with the low-light even though I think it performed admirably considering it’s age (and that I still jittery with excitement).

View fullsize Pier 14
 Pier 14 View fullsize The Bay Bridge Lights
 The Bay Bridge Lights View fullsize Fog
 Fog View fullsize IMG_7515_1798x2000_Squooshed.jpg
 View fullsize Back at the hotel
 Back at the hotel View fullsize Fog
 Fog

[Back to Top]

Saturday

Battery Spencer & the Golden Gate Bridge

I had a super ambitious plan for Saturday: to start from the far side of the bridge and work our way back across the bridge towards Fisherman’s Wharf. I had thought originally to only go as far as the near-side of the bridge and take pictures from there but to fly a bazillionty miles out to only go to the south side of the bridge felt silly. It was definitely worth it: the north side was definitely less crowded! It was very surreal to see the fog up close and have pockets of visibility come and go.

View fullsize Like something out of a movie
 View fullsize The (former) gun emplacements
 View fullsize IMG_7560_3410x2000_Squooshed.jpg
 View fullsize IMG_7613_2999x2000_Squooshed.jpg
 View fullsize IMG_7615_3400x2000_Squooshed.jpg
 View fullsize The North side visitor center
 View fullsize IMG_7636_3510x2000_Squooshed.jpg
 View fullsize IMG_7653_3919x2000_Squooshed.jpg
 View fullsize Crossing under the bridge
 View fullsize Coming out from the under the bridge
 View fullsize IMG_7691_1331x2000_Squooshed.jpg
 View fullsize Horseshoe Bay
 View fullsize Fog
 View fullsize IMG_7738_1701x2000_Squooshed.jpg
 View fullsize IMG_7740_2246x2000_Squooshed.jpg
 View fullsize Looking south, Fort Point
 View fullsize Looking north, Fort Point
 View fullsize IMG_7747_2703x2000_Squooshed.jpg
 View fullsize IMG_7751_3465x2000_Squooshed.jpg
 View fullsize Southern visitor center
 View fullsize An absolute unit of a 'cable'

[Back to Top]

The longest walk of our lives

While the actual distance may not have been that much (about 5km), we needed to be at the Wharf in time for the next surprise. We stopped at the Wave Organ for a bit of a rest before continuing on our way.

View fullsize Crissy Field Beach
 Crissy Field Beach View fullsize Random Lighthouse
 Random Lighthouse View fullsize IMG_7817_3051x2000_Squooshed.jpg
 View fullsize IMG_7821_2896x2000_Squooshed.jpg
 View fullsize Wave organ
 Wave organ View fullsize Aquatic Park Pier
 Aquatic Park Pier

[Back to Top]

Bay Cruise

The last surprise of the day, I booked a Bridge 2 Bridge Cruise - it’s a super cost effective way of getting some closer wide shots of the bridge. There are a couple of nearby museum warships as well but we didn’t get a chance to visit.

View fullsize Bridge 2 Bridge Cruise
 Bridge 2 Bridge Cruise View fullsize Ship Museums
 Ship Museums View fullsize IMG_7957_2398x2000_Squooshed.jpg
 View fullsize IMG_7963_2674x2000_Squooshed.jpg
 View fullsize Alcatraz
 Alcatraz View fullsize IMG_8061_3746x2000_Squooshed.jpg
 View fullsize Bay Bridge
 Bay Bridge View fullsize IMG_8084_2896x2000_Squooshed.jpg

[Back to Top]

Misc

The architecture of SF is very unique: there are ‘whimsical buildings’ scattered amongst all of the modern towers. There’s also a crazy variation in inclination!

View fullsize The Hilton
 The Hilton View fullsize That incline!
 That incline! View fullsize IMG_8123_1544x2000_Squooshed.jpg
 View fullsize IMG_7850_3596x2000_Squooshed.jpg
 View fullsize IMG_7847_1777x2000_Squooshed.jpg

[Back to Top]

How did it go down?

What was supposed to happen

  • I booked the room for three nights (since I was arriving a day early); as far as my girlfriend was concerned, this was a two night stay

  • Upon checking in, the desk agent was supposed to go get a manager immediately — it was literally the first line in the reservation notes. The manager would then finish the rest of the check-in and let her know that there was a hotel tour for her after she dropped off her bags (this is a somewhat common occurrence in the hotel industry)

  • I expected that there would be about a 5-10 minute delay as she dropped off her bags, freshened up and came downstairs for the tour

  • During the hotel tour, I would let myself back into the room

  • The hotel tour would end back at the room where I would be waiting

How it actually went down

  • Shift change at this hotel is around 0750 — I specifically wait until after shift change to come downstairs to meet up with the new shift leader and verify everything is good to go

  • My girlfriend arrives at around 0950. The desk agent for the shift was a newbie - complete with an ‘In Training’ badge. I’m guessing muscle memory took over and she let it slip that it was a three night stay before being quizzically correct by my girlfriend that it was only two nights

  • It was at this point she read the reservation notes before running off to go find a manager. She’s disappeared in the back office for a good five minutes.

  • Thankfully the manager smooths everything over however in the hubbub, nobody suggested that the tour was starting soon (remember how I mentally allotted 5-10 minutes?)

  • Over the course of the next 30 minutes, I’m sweating bullets wondering if my girlfriend had snuck out the stairs or something — and I’m being watched like a hawk by a security guard

  • I cave in and call one of the managers asking him to call up to the room to gently suggest ‘HURRY UP’

  • A few moments later, I see my girlfriend come out of the elevator and head off on the tour - we can finally get this show on the road

  • I get upstairs only to find out I’ve been locked out of the room. When making keys for a room, you can either Add key-cards to a room, or Replace them. During the excitement, newbie replaced my keys so. Another call to the manager and he meets me at the room, bringing along the same security guard who’s been eyeing me.

  • The security guard overrides the door and I’m in. At this point I don’t really an idea of how much time has passed and how far along the tour has progressed so I frantically setup a couple cameras

  • I get a few moments to panic quietly to myself before I hear someone opening the door…

So fun times, I captured my heart-rate during all of this…

FAQ

  1. Did she say yes? Yes

  2. Did you have a speech? Did you get down on one knee? Yes and yes

  3. Did you ask for permission? Yes

  4. Was she surprised? Yup

  5. Why did you propose then in August?

    • I wanted to line up my proposal with a meteor shower; on the off chance that we could catch it during the trip but also because it would give us something celebrate. August lined up perfectly for the Perseid’s Meteor Shower

    • Based on the historical weather trends for SF, the weather conditions were great

    • Our five year anniversary had passed in July previously and I had really played up the notion that I might propose then - but didn’t (in fact, I’d been teasing it for almost a year at this point). After the anniversary came and gone, she let her guard down

  6. Why San Francisco? Did you consider any other locations?

    • I wanted to pick a location that was so preposterous that she’d never suspect it

    • The shortlist was San Francisco, Santorini, and Paris. Of the three, San Francisco was the most cost effective to get to

    • I considered a few locations in Canada as well but it was (sadly), cheaper to go to San Francisco (!!)

  7. How far in advance did you plan this? Just over three years to put the plan together (with a few hiccups along the way)

  8. Why did you use South Korea as a cover story? It was in a sufficiently different time zone that she wouldn’t be thrown off if I didn’t reply and I had access to photos that I could share to suggest I was in Korea.

[Back to Top]

 

Related

If you want to see the story unfold from her perspective…

Read More
 

Permalink

]]>Weekend Getaway to San FranciscoGuide: Getting Better Video CallsAVTechDPMon, 01 Feb 2021 21:25:00 +0000https://www.minmaxgeek.com/blog/guide-better-video-calls6020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:60317e507806406cda6091e0Well 2020 introduced a lot of people to the concept of working remotely and with that, the growing pains of transitioning a workforce to collaborating remotely. Those of us that have done conference calls in the past know the pain this brings. After putting together Skype Station (read about it here), I wanted to learn how to take advantage of the fancy equipment we have and make sure that we weren’t contributing to the bad call-experience for others. Looking past some of the more amusing gaffs, every call seems to have one (or more!) of:"

  • Users not using headphones, earbuds, a headset or anything to prevent feedback

  • Sadly, many times, this is the user who doesn’t mute themselves when not speaking

  • Has the camera aimed up their nostrils or down at the top of their head

  • They have a bright window at their back which throws off the exposure entirely

  • Or has strong overhead lights on so their face is covered in shadow

  • They sound like they are in a glass coffin or are surrounded by pots and pans

Let me say I’m not an expert at any of this - but YouTube and the Internet at large has been a thing for some time now and it’s not hard to learn. What may be surprising is just how accessible it is to improve your video-call game.

Quick Links

Starting Point

We have to start somewhere so just to get started, let’s assume your current video-call setup comprises of:

  • A regular laptop with a built in webcam, microphone and speakers

  • A generic meeting room or kitchen

I will also assume that everyone has earbuds or headphones of some sort available to them. So far, this is a pretty accessible bar; when we talk about webcams, I’ll assume those setups start with a generic 1080P webcam that sits on top of the display.

For all of the audio samples on this page, I use:

  • Audacity: 44.1 Stereo, 32-bit float

  • Normalized to -6 dB

  • To post the audio samples here, I encoded them all to MP3 (-m j -V 2 -q 0 -lowpass 18.5 --vbr-new -b 32)

[Back to Top]

Audio is King

At first I found this to be a weird one to wrap my head around: that audio would be the most important component of a video-call. But it’s no joke: it doesn’t matter if you’re setting up for a company-wide conference call, streaming a game, a quick call with friends and family, a sales pitch to a potential customer or even a local recording locally - if you have bad audio, it’s all for naught. Just think of all the times you’ve clicked on a video clip which had terrible audio and you just noped right out of there.

Your guests/audience will forgive bad video - they won’t forgive bad audio.

What might be worse is when audio is just a little bit off; this happens often when users have ‘a good mic’ but failed to spend the time to set it up or use it correctly. Slightly-off audio is particularly bad on professional calls (think sales or support & training type calls) - subconsciously your audience is going to be trying to figure out why the call sounds off. Listening to this for any extended duration is not only exhausting but it means they aren’t focusing on your message.

Thankfully, it’s easy to improve your audio without spending too much time, effort or money.

What is good audio?

I’m not an audio engineer, nor am I a vocalist or someone who speaks for a living (thank god!); in short, I likely have a mediocre voice and ears. With that out of the way (and we have to start somewhere), let me propose the following as ‘good audio’.

How to check your audio?

As tempting as it might be, don’t rely on the built-in ‘test call’ functionality of conference apps. They are designed for quick ‘does it work’ testing; when you use them you are potentially introduce three variables:

  1. The test call audio is handled differently than normal calls — compare this Skype Test call sample which was recorded using the same equipment as above. You can hear the prompt-voice sounds great, but my sample recording sounds extremely muddy.

  2. Network load — if, at the exact moment of your test, your network has a slow-down? The app may overly compress your audio reducing the quality

  3. If you use multiple applications (Skype, Zoom, Teams, Meet etc.), they may each use a different compression scheme (or adapt to changing network conditions differently) - you might sound okay on Zoom at 1PM but you sound horrible on Teams at 2PM.

To test your audio, you’ll want to capture your audio as best as it sounds locally; this way you can compare apples to apples and when you’re done, you know that the audio you are sending through the call will be as good as possible before being compressed. I use a piece of software called Audacity: it’s free to use and works on almost any computer. There are tutorials for getting Audacity setup but for my quick testing I do five things:

  1. Select the input device (the microphone)

  2. Select stereo (shouldn’t matter but might as well)

  3. Select your output device: speakers or headphones etc. Since every pair of headphones/speakers sound different, remember to use the same device to play back!

  4. Record and (optional) Normalize

  5. See how it sounds! and (optional) Export it (so you can compare later)

Improving your audio for free

There are two things you can do to substantially improve your audio - for free!

  1. If you’re not wearing a headset, headphones or earbuds, stop being an acoustic pain in the ear for everyone and put them on. Wearing headphones keeps the audio from the call from feeding back into the microphone causing a feedback loop. It should go without saying that you should be muting yourself when not actively speaking as well.

  2. Every room will sound different. Simply changing the room you’re taking a call in can have a drastic effect on how you sound!

Spoiler: If your room has a lot of hard surfaces — bare walls, hardwood floors, hard ceiling, lots of glass windows and/or hard desks, then it’s a bad room! You can can do a quick ‘clap-test’ to check if a room is going to be terrible: clap sharply and if you can hear the clap echoing or reverberating then move on to the next room - that room is going to sound terrible.

I recorded some samples using the same built-in webcam mic in different rooms - some rooms are quite bad. All of these recordings roughly represent ideal conditions: I’ve controlled for the background room noise and I’m keeping my mouth about 6” from the microphone.

If all of the rooms you have available sound echoey, you’ll need to reduce the hard-surface footprint of the room. You can do this for free by temporarily bringing in things into the room that might soften up the surfaces: a rug, blankets, some canvas prints or even pillows and cushions. Scattered around, these softer materials will greatly help to reduce the amount of echo and bounce present in the room.

Cost effective ways to improve your audio

In simple terms, the most effective way to improve your audio is to get the microphone closer to your mouth. That’s it! We’re all familiar with the night and day difference this makes: just think of the difference between someone talking on the phone normally and then switching over to speakerphone! This is why microphones built into webcams, meeting room conference-call phones or multi-user puck microphones never sound good — they’re simply too far from your mouth. A difference of one foot can result in some noticeable changes (again, this is a ‘best case scenario’ recording)

With built-in/webcam mics, when you are close enough for decent audio, you’re usually too close for the webcam to focus or frame you properly! The most cost effective way to address this is to get a lav(alier) or lapel mic. A lav mic typically clips onto your collar or shirt and gets the microphone much closer to your mouth. This has three major benefits:

  1. Being closer to your mouth, the overwhelming majority of what the mic pics up will be you (less background noise, less echo)

  2. If you happen to look or turn away, your audio quality will remain the same

Even a stupid-cheap lav mic will give you a night and day improvement over the built in microphone.

Obviously the sky’s the limit if you want to throw money at it or if you have a specific needs, i.e.:

  • You want the mic to appear in the shot and it needs to fit an aesthetic

  • You need to capture audio without going through the fuss of wiring up each time

  • You specifically need to be wireless

  • You need to record multiple users

Remember: getting a more expensive or ‘better’ mic does not mean your audio will sound better. If you’re in an echoey room or you’re not speaking into the microphone (poor microphone technique), then all that fancier microphone is going to do is amplify the bad.

Improving audio in post

Our Skype Station is in a spare bedroom — hardwood floor, hard walls, hard closet, hard ceiling - an echoey mess. Getting a ‘fancier’ mic just means I more cleanly pick up the nasty sounding room. To get value out of that nicer mic, work needs to be done in post (between the microphone and the video call).

[I’m not even remotely an expert on audio] In OBS I have the following actions on my audio chain (breakdown here):

  1. Noise Gate

  2. Noise Suppression

  3. Compressor

  4. EQ

  5. Limiter

All of these filters are either included with OBS or are free; all it takes is a little bit of practice and experimentation (remember to use the same speaker/headphones when testing audio so you’re not introducing another variable!). With my specific room and audio filters, we can see a very drastic improvement.

[Back to Top]

Lighting makes or breaks video

Avoiding the two most common mistakes

Thankfully, making the best of your video is much less involved. For the basics, there are only two things to avoid.

  1. Don’t have a bright window at your back! Having a window at your back puts you at risk of having the background blown out and the subject - you! - being underexposed and dark. More advanced cameras will handle this better but ideally, you’ll want to avoid having a window at your back if possible.

  2. Don’t rely on ceiling/room lights. Of the two gotchas, this one is much more forgiving: when you have overhead lights, your eyes will be shrouded in shadow.

Here are some samples of both gotchas when running on a webcam, a smartphone and a full-sized camera.

Logitech C920 + Bright Background
Logitech C920 + Bright Background

Representative of a ‘good’ 1080P webcam, bright windows obliterate your subject. Built-in webcams will fare much worse and the few upmarket webcams may fare a bit better.

OnePlus 7T + Bright Background
OnePlus 7T + Bright Background

Smartphones have come a long way in image quality, low-light performance and dynamic range. For something that fits in your pocket, this is incredible

Sony A7Riv w/ 17-28 f/2.8 + Bright Background
Sony A7Riv w/ 17-28 f/2.8 + Bright Background

A placeholder for a higher end setup: a much bigger sensor and better glass means that this setup is much more forgiving. Other than cost, the tradeoff is that more work in post needs to be done.

Logitech C920 + Overhead Lights 1
Logitech C920 + Overhead Lights 1

When there is more ambient light to work with, the impact of the racoon shadows is less obvious

OnePlus 7T + Overhead Lights 1
OnePlus 7T + Overhead Lights 1

Again, this looks fantastic, especially considering it’s a phone.

Sony A7Riv + 17-28 f/2.8 + Overhead Lights
Sony A7Riv + 17-28 f/2.8 + Overhead Lights

I don’t think I’d notice the racoon shadowing if I wasn’t specifically looking for it. If your setup has a relatively modern camera, this will be so much more forgiving.

Logitech C920 + Overhead Lights 2
Logitech C920 + Overhead Lights 2

When the amount of available room light drops though, the webcam can’t keep the eyes acceptably exposed. Autofocus starts to suffer too (even though “there is lots of light”) and being a webcam, I can’t set the focus.

OnePlus 7T + Overhead Lights 2
OnePlus 7T + Overhead Lights 2

No surprise here, phones have come a long way, in their lower light performance. The racoon shadow is pronounced here but otherwise, this is workable.

Sony A7Riv + 17-28 f/2.8 + Overhead Lights 2
Sony A7Riv + 17-28 f/2.8 + Overhead Lights 2

This is a really good example to show that, even a fantastic setup will fail when given bad lighting. Everything here is sharp, there’s good dynamic range (all things considered) but there’s nothing we can do about the racoon stripe.

[Back to Top]

Soft vs Hard light

The hardness/softness of a light is determined by the type of shadows that are created. The three properties that will make a light softer are:

  1. Bigger lights are softer. A bigger light has more opportunity for light to diffuse, or scatter in different directions

  2. Being closer to the subject. (Factoid: even though the sun is super bright and super big, it’s also super far away, making it a harsh light)

  3. Being more diffused. Diffused light is light that is scattered evenly by the time it reaches the subject - while the sun in a clear sky is extremely harsh light, if it’s cloudy, by the time the light makes it through the clouds, it’s very diffused and extremely soft.

You can see here, the night and day difference between a hard light and a soft light.

Hard Light
 Hard Light Soft Light
 Soft Light

[Back to Top]

The five elements of a properly lit scene

To take lighting to the next level, we only need five things. Note that you don’t need all of these items nor do you need to buy “photographer” or “cinematic” grade equipment - a floor lamp with some high CRI bulbs which you can get at Walmart is a great place to start.

  1. A key light. A key light is just a fancy way of saying the primary light responsible for lighting up a subject. If you only have one light, this can be placed centered with the subject. If you have some kind of fill light, this can be off to the side.

  2. A fill light. A key light will make shadows on a person’s features (i.e., nose, chin) and a fill light will dampen those shadows down a touch. Note that a fill light doesn’t even have to be a real light — you can use a large piece of white craft board to bounce the key light onto the other side of the subject’s light. If you have a window that isn’t beaming light directly at a person’s face, the window could be a fill light. If you’re using an actual light, you can play around with the brightness of the key and fill light to change the mood.

  3. A hair (or shoulder) light. This is a subtle light that you place behind and off to the side of the subject and adds a tiny bit of light to their shoulder or hair - this little bit of edge lighting keeps the subject from blending into the background.

  4. A background light. This is another subtle light that you place behind you and gives the background a bit of light (or color). The purpose of this is to separate the subject from the background and give the entire shot some depth.

  5. A background. This doesn’t need to be something bought but could literally be a bookcase or shelf of doodads. Plain colored background are boring and sterile so this is an opportunity to spruce up the background a bit.

 

5 Elements of Lighting

  1. Key light. Here, the light is on the subject’s left — as evidenced by the shadow on the nose and cheek

  2. Fill light. We have a fill light on the subject’s right filling in what would otherwise, be a dark cheek

  3. Hair light. I set this to a yellow tone to make it stand out, this separates the user from the background.

  4. Background light. Another RGB bulb, just to give some color and provide a bit of interest to the background.

  5. Background. Here we have a backdrop, but some green paper and a virtual background would work as well, or pulling a bookshelf into the frame.

 

The above example shot was taken at night and with only the mentioned four lights.

[Back to Top]

Camera placement is important too!

In our ‘starting point’, I assumed that the common video-call platform was a laptop with a webcam built in or possibly, a standalone webcam clipped onto the monitor. No matter what your camera setup is, you’ll want to avoid having the camera too low (something common with built-in webcams) or too high (more common with standalone webcams clipped to monitors). Ideally, the camera would be at or slightly above eye level (looking down).

Camera too low!
 Camera too low! Camera too high!
 Camera too high!

[Back to Top]

What about a better camera?

Throughout this entire post, I’ve specifically glossed over the topic of what webcam to get. There are certainly milestones at which point it makes sense to make the leap to a better camera but even with a better camera, it’s all for nothing if the audio or the lighting is a disaster. There’s one more element that I’ve not addressed yet and it’s the platform you are making a call on.

Supported Resolutions, by Platform

It’s exceptionally frustrating that the information for Zoom and Teams/Skype (the two big players) is nearly impossible to find or is presented in such a silly manner, it might as well be obfuscated. Below is what I’ve been able to ascertain to be the maximal video quality. [Last updated 2020-02-21]

  • Google Meet: 720p / 720p (Send/Receive)

  • Cisco Webex: 720p / 720p (Send/Receive)

  • Zoom

    • 1:1 call, from what I can tell, this is 360p/360p video

    • Group Calls: 720p/720p (Send/Receive) or better, depending on what paid plan you have… but currently unavailable

    • Worth noting: screen sharing is supposedly uncapped (if you can pipe a virtual camera to the screen share)

  • Microsoft Teams

    • 1:1 calls, 1080p/1080p (Send/Receive)

    • Group Calls: 540p/540p (Send/Receive)

    • Worth noting: screen sharing is uncapped (if you can pipe a virtual camera to the screen share)

Camera milestones!

Assuming that lighting and audio are accounted for, I would say there are four tiers:

  1. A cheap ‘anything better than the built-in one’ webcam - an ultra cheap webcam can be a bit of a hit and miss game requiring a few test purchases to see if you got a model that works for you.

  2. A Logitech C920 or similar - if you don’t want to spend the time or money on trial-and-error with knockoff webcams, this will get you a good baseline experience

  3. Just using a smartphone - this can get you some spectacular image quality with a device you already have. If you’re directly connecting to a call with a phone, you’ll want to make sure you have power and some form of getting better audio in and out of the call. There are even options for iPhone and Android phones to be used as webcams on a computer.

  4. Varying degrees of using a dedicated camera connected either via USB or through a capture card. This is a super open ended bracket - anything from an older GoPro with HDMI output to a DSLR that can be used as a webcam via USB to a dedicated mirrorless camera connected to a USB capture device.

This topic tends to be nebulous and light on details just because there are so many potential options out there and what works for one person doesn’t work for another.

[Back to Top]

Bringing it all together!

At a very high level:

  1. Get the audio and video connected to the computer. This will vary based on the equipment:

    • Using a built-in webcam and mic requires no work

    • Using a USB webcam or microphone should be straightforward

    • If you’re using a capture device via HDMI, you’ll want to ensure you have a clean HDMI signal - steps to check this will vary based on your equipment

  2. Install some of streaming software such as OBS. You’ll also want to install a virtual camera plugin and a virtual audio cable. We’ll need these later.

    • Use the Video Capture Device to get OBS to read your video signal

    • If you are using a microphone routed through the camera, you may not need to do any additional work depending on your capture card. Use Audio Input Capture to get OBS to read the audio signal

  3. Do any post work on your audio and video sources

    • For audio, I talk about this earlier

    • For a video, in many cases a good starting point will to use a LUT.

  4. Prepare the outputs

    • For the audio

      • We want to mute or disable all the devices that we aren’t outputting

      • For the audio source(s) we do want to treat as our ‘microphone’, navigate to Advanced Audio Properties and set Audio Monitoring to one of the ON states

      • The last step is to go OBS settings —> Audio —> Advanced and set the Monitoring Device to the virtual cable.

    • For the video

      • Just check the menu Tools —> VirtualCam that the service is running (you can set to automatic if you prefer)

  5. Configure your conference application

    • In your conference app of choice, set the webcam to VirtualCam and set the Microphone to the Virtual Audio Cable. Set the output to whatever headphone output you have

In a future post, I’ll go through more thoroughly.

At a high level, what we ultimately want to do is “get audio and video into a computer, do some cleanup and use that as our audio and video for a web-call”.

[Back to Top]

 

Related

Every video-call (or streaming) setup is a bit different. I learned a lot about setting up the lights, audio and video equipment when building Skype Station. Click through to see what all went into Skype Station.

Build Log
 ]]>Build Log: Skype StationTechAVDIYDPMon, 01 Feb 2021 03:42:00 +0000https://www.minmaxgeek.com/blog/skype-station6020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:602f338b45e33468b0fa37f9My wife works remotely and spends quite a bit of time in online meetings. This was a perfect opportunity to tie in all the different hobbies: woodworking, photography and tech. In addition to using her as a guinea pig for testing out the audio & video equipment, once the station is complete, it can be used to do recordings.

Quick Links

Where we started...

For almost as long as I can remember, I’ve always had a webcam and a headset.

I bought my webcam back in 2014. Back then, this was the de-facto premium webcam of choice and sadly there just hasn’t been that much innovation in this space. The C920 has spawned a few minor variants this never really made substantial and meaningful improvements until perhaps the 4K Brio in 2017. The Brio was a upgrade in every meaningful way but also came with twice the price tag. While it was an upgrade, it wasn’t that big of an upgrade and frankly, I barely used my webcam as it were.

Fun note: with the COVID pandemic of 2020, I saw the pricing of the Brio go as high as $400 (about four times the price I paid for the C920)

My current headset of choice is the Arctis 7. For years, I used the Logitech G930 (plural!). For my needs, these headsets where great except for one crucial failure — the headband and cup mechanism was held in place by a piece of plastic. This piece of plastic was held captive by four screws. The way the mechanism worked, this piece of plastic was under constant strain and over time, would fracture; once broken, the ear-cups would dangle and be useless. I bought more of these headsets than I care to admit. The big selling point for the Arctis was that the design of the headband and pivot couldn’t fail in this fashion.

View fullsize Logitech C920
 Logitech C920 View fullsize Arctis 7 Wireless Headset
 Arctis 7 Wireless Headset

[Back to Top]

Goals and Constraints!

The overall objective of this project is to have a dedicated station ready to go at a moment’s notice for calls, streaming or recording. This is more so for my wife as my calls are generally voice-only and I’m quite well setup for that at my desk (Shure SM58 + Behringer X1204USB). The guiding theme for this project will be ‘how can we make this even better or more seamless?’

At a super high level, this project has three, somewhat nebulous objectives:

  1. A dedicated station

  2. Improved video

  3. Improved audio

One of the challenges in this project is keeping all of the equipment accessible and easy to use — we don’t have a camera operator, focus puller or audio engineer so everything just needs to work without too much extra fuss. From a technical perspective, the plan is to route as much as I can through the camera equipment and have a single capture stream into the computer - no syncing in post, no juggling different audio sources.

Once the basic construction is done (which should be the easy part), we have a lot of potential equipment we cans mix and match to try and find a good balance. This project will be a bit more open-ended and also be a learning experience for me as I explore ‘how can we make this better’. I’m super grateful that this project synergizes so well with a bunch of preexisting hobbies. Thankfully, we have a spare bedroom we can partially convert into a meeting space; for the desk, it will be super compact and super simple — this is a virtual meeting space, not a sprawling full-time computer desk. For the time being, the desk just needs to hold a simple computer/monitor and have space for a keyboard and some notes. Once the other equipment matures, I’ll revisit making enhancements to the desk.

[Back to Top]

Step 1: Starting with a desk

We don’t have a lot of room depth to work with so the desk needs to be super compact. Using up leftovers from my other projects, this build only took an afternoon to finish.

View fullsize Building up the table top
 Building up the table top View fullsize Padding out the Interior
 Padding out the Interior View fullsize Build Inspector
 Build Inspector View fullsize Laminating the top
 Laminating the top View fullsize Tacking with Pins
 Tacking with Pins View fullsize Trimming the edges even
 Trimming the edges even View fullsize Easing over the edges
 Easing over the edges View fullsize Insetting the legs
 Insetting the legs View fullsize Back brace / foot rest
 Back brace / foot rest View fullsize Construction complete
 Construction complete View fullsize Test Fitting
 Test Fitting

[Back to Top]

Step 2: Equipment Configuration

I wanted to plan for a few different configurations:

  1. A single user on a simple call

  2. One (or more) user(s) on a call, potentially with the need to record

  3. A single broadcast, but using two camera angles

View fullsize Primary (single user) configuration
 Primary (single user) configuration View fullsize Multi-person/record configuration
 Multi-person/record configuration View fullsize Multi-cam configuration
 Multi-cam configuration

[Back to Top]

Step 3: The PC

The base requirements for Skype Station are pretty relaxed with the only real load being simple video transcoding and piping that stream to a virtual webcam so that applications like Skype or Zoom can use as they normally would. The old laptop we’re currently using has Intel Quick Sync support which gives it the ability to punch well above its weight for basic x264 transcodes. Beyond this, Skype Station needs to run office and collaboration software — nothing intense.

The only thing of note is that Skype Station is two floors away so WiFi performance could conceivably be a problem. We ran wifi fine for quite some time and then one day we ran into nonstop connectivity issues - I bought a powerline networking kit (since we rented at the time) to try and short-circuit the problem (I also ran a length of ethernet cable up the stairs in a ‘standby’ location for troubleshooting or when true wired performance was required).

Q4 2020 $0 Computer MSI GT70 Laptop $0 Intel Core i7 3610QM, 4c/8t16GB, 500GB SSD17" 1080P, GTX670M 3GB $0 Keyboard Logitech G510 $0 SKU: 920-00253018 Physical Macro Keys, 3-profilesIntegrated Audio, Display Screen $0 Mouse Microsoft Arc Touch $0 1000 dpiNovel folding mechanismSuper long battery life $0 Network NETGEAR PLP2000 Powerline $0 Powerline AV2Dual gigabit portsPass through power outlet $0 Primary Capture Device Elgato Cam Link 4K $0 Input: 4K30, 1080p60Output: 1080p60Entirely USB powered $0 Secondary Capture Device Y&H Video Capture $0 Input: 1080p60Output: 1080p60Entirely USB powered

[Back to Top]

Step 4: Audio and Video Equipment

There is a bunch more equipment churn with Skype Station both as I figure out what equipment that we do have works best that we can spare (since I don’t want to be swapping gear in and out if it’s needed elsewhere) and as I learn more about this space.

Video Equipment $0 Camera Sony RX10 $0 1" sensor, ISO125-25600Set to 1080P60FW50 dummy battery $0 Lens 24-200 f/2.8 (built in) $0 Built in electronic ND filter62mm filter thread $0 Field Monitor Feelworld FW279 $0 7" 2200 nit, 8-bit IPS4K input, 1200p output Lighting Equipment $0 Key Light NEEWER P280-2.4G $0 140 bi-color LED (3200-5600K), 95+ CRIBattery or DC power, wireless remote $0 Fill Light NEEWER P280-2.4G $0 140 bi-color LED (3200-5600K), 95+ CRIBattery or DC power, wireless remote $0 Hair Light IKEA Lersta Lamp $0 52" height, 6" lamp shadeT TECKIN RGB WiFi Smart Bulb $0 Background Light CANVAS Table Lamp $0 6" diffuse shadeT TECKIN RGB WiFi Smart Bulb Audio Equipment $0 Microphone Adapter Sony XLR-K3M $0 Allows for 3 audio inputsDirectly connected via the Multi Interface Shoe $0 Microphone Sony ECM XM1 $0 Unidirectional condenser microphone21mm diameter, 50Hz-20kHz $0 In-ear Monitors KZ ZS10 Pro IEM $0 4x balanced armature, 1x dynamic driver7-40kHz, 3.5mm TRS $0 Wireless Audio Avantree Pre-Paired Bluetooth TX/RX $0 Can be charged while in useConverts any headphones into Bluetooth Other Equipment $0 Tripod Vanguard Alta Pro 2 263AV (with PH-114V head) $0 15 lbs max, arca swiss video headApprox. 54" without extending the center column $0 Teleprompter Glide Gear TMP100 $0 Supports tablets up to 10x9Easy to fold-down & setup $0 HDMI Duplicator Techole HDMI Splitter 1-in, 2-out $0 4K30, 1080p60, 10.2Gbs bandwidthIncludes USB power input for longer runs $0 Light Stand (x2) Pipishell 14" Webcam stand $0 Built in ¼-20 thread, 32.7" max reachMax 1.5 kg, 2" clamp capacity $0 Microphone Stand Ohuhu Microphone Stand $0 31-52" height30" reach $0 Backdrop AIIKES 5x7 Vinyl Backdrop $0 Antique brick patternHung using clips & ½" EMT on Command Hooks $0 Footrest Cozy TT Heated Footrest $0 12" x 18" x 5"Only uses 90w $0 Chair IKEA Markus $0 Repurposed from the basement officeHighback, tilt-lock Video Equipment $0 Camera Sony RX10 $0 1" sensor, ISO125-25600Set to 1080P60FW50 dummy battery $0 Lens 24-200 f/2.8 (built in) $0 Built in electronic ND filter62mm filter thread Lighting Equipment $0 Key Light NEEWER P280-2.4G $0 140 bi-color LED (3200-5600K), 95+ CRIBattery or DC power, wireless remote $0 Fill Light NEEWER P280-2.4G $0 140 bi-color LED (3200-5600K), 95+ CRIBattery or DC power, wireless remote $0 Hair Light IKEA Lersta Lamp $0 52" height, 6" lamp shadeT TECKIN RGB WiFi Smart Bulb $0 Background Light CANVAS Table Lamp $0 6" diffuse shadeT TECKIN RGB WiFi Smart Bulb Audio Equipment $0 Microphone Sony ECM-W1M $0 Super compact, wireless lav mic (AAA)Bi-directional audio passthrough $0 Headphones Cheap Earbuds $0 Super cheapTRS, for connecting to the laptop directly) $0 Wireless Audio Avantree Pre-Paired Bluetooth TX/RX $0 Can be charged while in useConverts any headphones into Bluetooth Other Equipment $0 Tripod Vanguard Alta Pro 2 263AV (with PH-114V head) $0 15 lbs max, arca swiss video headApprox. 54" without extending the center column $0 Light Stand (x2) NEEWER Basic light stand $0 3-6.5 ftExtremely lightweight $0 Backdrop AIIKES 5x7 Vinyl Backdrop $0 Antique brick patternHung using clips & ½" EMT on Command Hooks $0 Footrest Cozy TT Heated Footrest $0 12" x 18" x 5"Only uses 90w $0 Chair IKEA Markus $0 Repurposed from the basement officeHighback, tilt-lock Video Equipment $0 Camera Sony A7 III $0 Set to 4K 24, 100MbitFZ100 dummy battery $0 Lens Tamron 28-75 f/2.8 $0 SKU: A03667mm filter thread Lighting Equipment $0 Key Light IKEA Lersta Lamp $0 52" height, 6" lamp shadeCree LED 2700K, 80CRI Audio Equipment $0 Microphone Sony ECM-W1M $0 Super compact, wireless lav mic (AAA)Bi-directional audio passthrough $0 Headphones Cheap Earbuds $0 Super cheapTRS, for connecting to the laptop directly) Other Equipment $0 Tripod AmazonBasics 52" Carbon Fiber Tripod $0 8 lbs max, arca swiss ball headApprox. 43" without extending the center column $0 Footrest Cozy TT Heated Footrest $0 12" x 18" x 5"Only uses 90w $0 Chair IKEA Markus $0 Repurposed from the basement officeHighback, tilt-lock

[Back to Top]

Milestone(s)

The original goal of this project was to give my wife a dedicated space to have her calls and this project epically sailed right past that goal. This is very much an exploration of ‘how can we make this a better’. Since starting this project, I’ve learned a ton on lighting, video and most importantly, audio (although I’d still consider myself quite the beginner). This project is an exercise in controlled feature creep (if there can be such a thing).

With this project, I set out to create a flexible but compact mini-studio. Skype Station is going to be a bit of an ever-evolving station so this will be more of snapshots over time.

View fullsize Just getting it into the room
 Just getting it into the room View fullsize Switched camera &amp; added lights
 Switched camera & added lights View fullsize Using webcam stands
 Using webcam stands View fullsize Adding a backdrop
 Adding a backdrop View fullsize Hanging the backdrop
 Hanging the backdrop View fullsize XLR Audio
 XLR Audio View fullsize Mic placement
 Mic placement View fullsize No more tissue papers
 No more tissue papers View fullsize Teleprompter
 Teleprompter View fullsize Close-up of the flip-view
 Close-up of the flip-view View fullsize Milestone 2
 Milestone 2

[Back to Top]

 

Related

Now that Skype Station is in place, I need to do some trial and error (and learning) to get the most out of the setup. Read more to see the common mistakes that might be holding back your setup.

Read More
 ]]>Build Log: Skype Station