r a m y h a n n a . c o m

V-ray for SketchUp

2013-05-20T16:30:00.000-05:00

Here is the latest batch of renderings we produced at Tiltpixel. These were all modelled in SketchUp and even rendered in SketchUp using Chaos Groups V-ray for SketchUp plugin.

We were quite impressed with how much we could push v-ray inside SU. It's the same old rendering engine, and even supports proxies and vray-RT. For objects that were high poly, we relied on using .vrmesh files. It has everything else in it from render elements and vray 2 sided material to distributed rendering.

Break up those panel reflections!

2013-04-16T23:20:00.002-05:00

I'm going to share a tip on shaking up those reflections among panels. I can't take credit for this, as the talented Pierre-Felix Breton brought this to my attention last year at AU. I don't know how often I've seen this but often.

If you want to create a distorted reflection, then you add a noise map to the bump parameter. Fine. Unfortunately in 3d this map gets distributed evenly among all objects, making all the panels act/look like one single panel.

In real life however each panel has its own unique distortions and should look like this (I exaggerated the effect for emphasis):

Achieving this effect in 3dsmax is really easier than you think. All you need to do is select a few panels, and apply a UVW Xform modifier to them. In this example I selected every other panel. The only parameter I adjusted was the "U Offset" amount (set to 1'). This simply shifts the previously applied maps to this modifers settings.

But there is a trick to getting this to work correctly with procedural maps such as the noise map. Be sure in your noise map you change the Coordinates Source from Object XYZ to Explicit Map Channel, otherwise the UVW Xform doesn't know what to do with those procedural maps.

Check out our latest renderings

2013-03-22T12:41:00.000-05:00

Wondering what's been going on at tiltpixel.com? Check out some of our latest renderings

Rename those Objects!

2013-03-21T10:07:00.000-05:00

Did you know there is a tool for renaming objects in your scene? I've been using 3dsMax for a while now, and I"m still finding small little amazing time saving tools in it so I thought I would share. This would have been useful years ago!

Often I have a scene that looks like this. Where in hindsight it would have been good to name the object initially, but now they are in my scene with a generic name.

Simply select all of those objects you want to rename together, and go to Tools->Rename Objects.

It will pull up this simple little tool. It gives you all the options to keep the existing name, give it a prefix name, add numbering or remove the first so many digits. Simply enter the parameters you want to get changed and click Rename. Instantly all of your selected objects will be renamed with your chosen parameters.

A great tool and a huge time saver!

I'm now at TILTPIXEL

2013-02-11T16:59:00.001-06:00

Ok, it's official. I now work for tiltpixel. It's a studio that provides high-end creative renderings, and I'm honored and privileged to be a part of it. What does this mean for me here? It means I'm doing more of the same thing I've been doing, only i can now focus even more time to rendering and process. So hopefully I will have more time to dedicate to my blog as well. So stay tuned, I'm not going anywhere and there are big things to come. You can also like us on our facebook page: http://www.facebook.com/Tiltpixel

Interview with Marcin Gruszczyk

2012-12-12T11:49:00.000-06:00

Typically my posts are my how to's and tips and tricks, but every now and then, I think it's important to get inspired and highlight some good talent. I don't do alot of interviews, but I have been a fan of Marcin's work for sometime and thought it important to put the spotlight on him as his work is very inspirational. Not only are his renderings beautiful, but there is alot of technical materials and lighting going on!

rh: Marcin, for a while now I’ve been a fan of your work, and as someone who is in the arch-viz rendering industry, your work is very fresh and new as your works subject matter, airplanes, are very different from typical architecture or automotive renderings. You’ve managed to take what could be very mundane shots of something in midair, and turn them into amazing compositions and works of art.

Marcin: I am very happy to be able to work together with many dedicated and talented people. In fact, all of the work for Lufthansa is a team effort and as a CG guy I have learned so much. Thankfully, our client appreciates our work - which is unfortunately not the case for so many CG artists out there. At times, I'm still able take a short break sometimes and work on something personal, either torturing a teapot with dynamic simulations or rendering a car.

rh: How did you get into doing renderings?

Marcin: I got in touch with computers and graphics in the C-64 era, did my first 3d renderings on Amiga later but lost interest in computers after that, mainly because of my other passion, the guitar. Years later, I started doing CD designs and flyers for my band on the PC and got hooked again; it was exciting to see how much 3d has developed in the meantime. At that time, I have been working in a print house doing pre-press work and layout and played around with 3d software at home, hoping to work in the 3d industry someday. I then got to work at a company which produced 3d real-time architectural presentations and later, in 2006, met the folks in Frankfurt and started working with them.

rh: What software’s do you use to create your renderings, and what is your workflow?

Marcin: Workflow will always differ depending on the job, but for most of the Lufthansa images our workflow has been as follows:

If NURBS data was supplied, I've used MoI3d to review, remodel or repair the data when needed. I then imported this as NURBS into 3ds max through the Npower importers and finally converted to mesh or poly objects. Getting NURBS data from manufacturers is a huge problem because of IP restrictions, so we ended up modeling a lot as NURBS in Moi3d and in 3ds max. I had an excellent library of reference images which allowed me to model with a high level of detail and was very useful as a guideline for the materials and textures. Brazil was used for rendering and shading. It is always a critical point for me so I usually spend a lot of time creating materials and textures, and doing many test renderings.

Once everything was in place we could start working on the images with Jens Polkowski from Lufthansa and photographer Jens Goerlich.

We did a lot of previz before the backplates were shot. The client's idea was to produce a consistent set of images that would be later used in a story about the new Boeing 747-8 in the Lufthansa magazine. We determined the best camera angles and lenses and basically defined and documented everything in 3ds max before the actual shooting. Jens had detailed information on the camera's levels, tilt angles and focal lengths and could shoot backplates that would later match our cameras in 3ds max. It turned out to work surprisingly well and helped a lot to maintain the consistency Lufthansa had in mind across very different shots.

After the shooting, I've been given bracketed shots of the environments, done with a Canon 5D and a fisheye lens. We've used PTGui to stitch these as equirectangular panoramas and Photomatix to generate the HDRs, and because they were shot shortly after the backplates were taken they had almost identical lighting conditions.

rh: Most of the backgrounds in your Lufthansa renderings are in fact photos. How does your photographer manage to get such beautiful photographs, and how do you get the lighting and reflections from the photos into your renderings?

Marcin: As an airline, Lufthansa needs to show both air-to-air and ground shots of their airplanes, showing them in different situations, lighting conditions and perspectives. Jens has been on several helicopter and airplane flights to capture beautiful and interesting skies and cloudscapes. For ground shots, we typically get a more defined request.

It's not easy for anyone to imagine a huge airplane somewhere on the ground when it's not actually there. We'll often sit together before a shooting in order to look for unseen perspectives or to just get a better picture of the airplane's shape. Out on location, he tries to capture an interesting picture of the area while keeping in mind the plane's position and dimensions. Jens is able to tell which focal length the camera I'm using in 3ds max after looking at the screen for a second so he really knows what he's doing out there.

As for the CG part, he usually shoots the backplates and the HDR environment in the same lighting conditions so they are easier to work with and mostly require only a basic color correction.

For lighting and reflection environments in air-to-air shots, I needed a different workflow due to the fact that one cannot shoot an environment HDR up in the air. I'm using Vue and Terragen to produce panoramic HDR environments, trying to mimic lighting and cloud formation from the real-world background plates.

Sometimes, I'll model some of the landscape's features in 3ds max and use camera mapping to map the background image onto it so it reflects in a more realistic way.

I tweak a lot until the CG objects look convincing. Sometimes it may be necessary to add fill lights or to change the sun's position in order to get a more interesting image or emphasize a certain area of the airplane. Additionally, if the backplate and HDR environment don't match, it may be necessary to apply curves or color corrections on the HDR images.

Luckily, there are plenty of free plugins and scripts to work around 3ds max' limitations. I have used Martin Breidt's ImagePlane and Overscan scripts, some of Neil Blevins' scripts, some of BlurBeta's maps and materials, Cuneyt Ozdas' ColorCorrect map, BerconMaps and many more.

rh: The details of the airplanes are exquisite, down to the screws and bolts. How much of the detail in the airplane is modeled geometry and how much of it do you rely on texture maps?

Marcin: Since airplanes are so huge they need a certain level of detail so they don't look like toy models when rendered. You really need to see screws, rivets and all kinds of small parts to create the illusion of a real-life object.

After working several years for Lufthansa, I know which parts of the airplane are the most interesting to them and try to keep that in mind when modeling or texturing. Something that's likely to be seen in a close-up will get a lot of attention. For example, we've produced a series of images with the airplane up in the air over beautiful landscapes, with the camera looking at the engines. They're all quite highres and require a very high level of detail to look convincing - so we started to model every single screw and rivet in our recent models. While it may sound like it's way too much work, it actually saves a lot of time that would otherwise have to be spent on painting and tweaking textures, which also needs considerable amounts of time when going for very high resolution output.

Also, an airplane is covered with markings almost everywhere. Most of those that are visible from a greater distance need to be painted and mapped. Some of them require different mapping projections and many of the main texture setups take up several map channels per object because of this.

When creating materials, I try to instance textures and maps as often as possible. It can get very tedious to replace a color map for example when it is used in many different materials. If you take care of those things at the beginning it'll eventually pay off.

rh: You have a very keen eye for composition in your renderings, do you pick the camera shots yourself, or is it a collaboration effort?

Marcin: This is always a collaboration effort. Often, Jens Polkowski from Lufthansa will approach us with an idea, either suggesting a certain composition or leaving it up to us to come up with something based on their marketing needs. We'll then work a bit on different camera angles and send them back to Lufthansa so they can give us feedback or decide which of them get produced.

I have learned a lot from both of them in this regard. Funny enough, sometimes they'll be able to imagine the rendered result much better than I am, and it still happens regularly!

rh: Although most of the backgrounds in your renderings are photos, you have also been successful in creating cg generated backgrounds. How do you get such realism in clouds, and mountains?

Marcin: For the image below of the Airbus A380 and Boeing 747-8, I have used Terragen and World Machine to produce the background. As a first step, I created an alpine landscape in Terragen and exported a part of this landscape as a heightfield map. In World Machine, a geological erosion effect was applied and exported back to Terragen in order to render the final image with clouds and proper lighting. I also rendered out a HDR environment and used this 3ds max to render the airplanes; the final composite was then done in Photoshop.

It's really fun to work with these programs and the possibilities are almost endless. It is something I want to use more in the future as you can get some beautiful results without having to rely on real-world photographs.

rh: How much of your renderings are touched in post, and what tools do you use to create the final composites?

Marcin: While I think Photoshop lacks some essential features (full HDR support or a non-destructive workflow for filters and masks are good examples), I still think it is the best tool for the kind of work I am doing. I try to avoid rendering in passes as Brazil doesn't support them in a convenient way. Instead, I try to get everything right in the rendering and only render out an additional shadow pass for better control over the shadows and masks for certain objects to avoid painting masks. After that, I'm doing some basic level, color and contrast corrections and leave the real work to Meike Wittenstein, a real Photoshop sorceress, who gives it a final touch.

Thank you, Ramy! Last but not least, I'd like to credit some others involved in this project as well, namely Fjodor Tscherkassow, an incredibly talented 3d modeler, and Martin Repplinger, who's been responsible for stitching the HDR environments.

rh: Marcin, thank you for all the rich information, and taking the time to answer these questions. You have definitely set the bar in many aspects. Your images are very inspirational, and fresh as your process is different than most out there!

You can see more of Marcin's work on cgsociety.org or his website mswee.net, and much like his renderings, Marcin went above and beyond, and has shared with us some of his insight to his process creating his awesome materials shared here:

Adding real-life look to materials

Since I have received a lot of positive feedback on the job, particularly on shading, I'd like to give a brief overview over how I added dirt to the materials. I am working with Brazil and because it's only rarely used out there, I will try to explain the idea behind the approach instead of showing screenshots of the material setups.

When working on the first model and doing the first test renderings I quickly realized that I'll need to introduce some kind of dirt on a material level to get a realistic look and a sense of scale. Painting maps for every single part of a huge model wouldn't have been feasible. Even if I would have taken the time to do so, it would have been a nightmare to manage the materials and would have taken a considerable amount of memory - only to add something for what should be solely a subtle effect.

The solution was to use a map that would paint a gradient on sharp edges of a mesh in a procedural way – I have used the f-edge plugin to do that and used that map as the blend mask in a Blend Material (see my notes at the end of the article on how to achieve similar effects with other plugins). This mask would then blend between a clean and shiny material (Material 1) and a more or less dirty and worn version of the clean material (Material 2). To take this a bit further and make it a bit more convincing, the f-edge map is distorted by another map, in this case by a Bercon Mapping map (there's also a free map called Warp Texture which can be used for the same effect). Using this in a ColorCorrect node will give you even more control.

Another technique to simulate dirt is to use an ambient occlusion map and use different procedural noise maps to make it look less regular. Again, if used as a blend map in a Blend Material to blend between two versions of a material, it can add realism to the shading without much effort.

Generally speaking, I often blend many procedural maps in a Composite Map and use them as diffuse or reflection textures. With huge models with thousands of objects and hundreds of materials, working with painted textures quickly becomes a pain - you always need to go through a chain of steps, unwrapping or taking care of the mapping projection, painting the textures, loading them in 3ds max and taking care of map channels etc. Procedural setups can help to quickly get a convincing level of realism in your materials. However, keep in mind that using procedurals will add to the rendertime and the more complicated your setups are, the slower they'll render.

Note:

Both techniques were already described by Neil Blevins in the past:

http://www.neilblevins.com/cg_education/vertex_map_wear/vertex_map_wear.htm http://www.neilblevins.com/cg_education/ambient_occlusion_rust/ambient_occlusion_rust.htm

For the worn edges technique, his approach is different as it is based on vertex colors. As a consequence, it is not completely procedural as it needs to be setup per object, it requires a rather high resolution of the mesh to look nice and it generates an additional mapping channel, thus requiring more memory than a completely procedural approach and will add to the file size.

For the ambient occlusion mask technique, I have used different maps to achieve a similar effect, but the general workflow is the same.

• f-edge is a commercial 3rd party map, available from http://www.ddag.org/products.html

• A similar effect can be achieved with the Gradient Edge map from Vladislav Gavrilov (http://www.vg2max.spb.ru/GradientEdge.htm). Vray users can use the VrayDirt map with the 'invert normal' option for a similar effect

• Keep in mind that most of the maps used here won't work with mental ray

Feel free to e-mail me if you have any questions regarding this workflow or for anything else – mg[at]mswee.net

-----------------------------------------------------------------------------------------

Copyrights for all images: Deutsche Lufthansa AG, Jens Goerlich, mo cgi

My 2 most commonly used maxscripts

2012-11-27T01:03:00.000-06:00

Over the years, there are 2 maxscripts that I use significantly more than any others, so I thought I would mention them and share them with you.

Arch&Design Utilities v2

This is a script that was originally written by Pierre Felix Breton and Zap Anderson back in 2006 I believe, and is super at changing some A&D material settings globally. I tweaked it and added the "Turn Back Face Culling OFF in all materials" button, hence the v2. The biggest benefit for using this script is that when you render a raw AO element, you can be certain that every object can have AO assigned to it, and you won't be getting surprised black objects in your pass (as long as you're only using A&D in your scene)....very powerful little script and you can download it here:

mrA&D_utilities2.ms

Unified Sampling

This is one I've mentioned before, but the current script by Artur Leao has many other mr features. This one is his older script, which is just unified sampling. It's so simple, it makes it worth using, and the benefits of using unified especially with moBlur are huge. What is nice about this script is you can save the settings in the file itself and it will render without a hitch on a farm or through backburner. You can download this file here:

unifiedSampling_mr3.9_2012.ms

What I particularly like about the .ms format for these scripts, is all you have to do to use them is click and drag them into your 3dsmax viewport. If you really want them into your UI though, simply open the scripts through a maxscript dialouge in 3dsmax, highlight all the text and click and drag it onto your main toolbar. From there you can change the icon of the button by right clicking on it. Simple as that!

AU2012

2012-11-06T08:50:00.002-06:00

Has it really been another year? Autodesk University is coming again, and I will be at AU2012 again this year giving a lab:

Rendering with mental ray®: Tips and Tricks

My lab will go through my cshs animation ; I'm taking one of the camera shots, and I will go through how to texture, light and render it to create this type of animation.

See you at AU!

Creating 3D logos from a jpg

2012-10-23T17:28:00.000-05:00

Lately I've been creating quite a few seals in my renderings, and thought I'd share a really fast way to bring 2D logos and seals into your 3d models. To do this well, you have to have illustrator, but there may be other vector programs out there that do this that I'm not aware of. Typically I look first for a vector version of a logo, usually I find them embedded in pdf files such as handbooks with a logo on the title page. However, if I can't find a vector image, then I can still work with a raster image.

So I start off with an image, typically found off the internet. I use Illustrator to convert a 2d image to vector, and it does the job fairly well. There are several things to look for when picking your image to get this to work well:
• Choose the highest resolution image you can find
• Images with strong contrasts in color work better

Simply open the image in Illustrator. With the image selected go to Object->Live Trace->Tracing Options. I always check "Preview" to see what my image will look like when it's converted. I also check "Ignore White" to remove white as an object and simply use the lines. The advantage to this is that you can get true holes in your logo, which is an effect impossible to create with opacity and bump alone. Text can be a bit tricky with the conversion and often doesn't look quite right, but you can adjust some of the Trace Settings to tweak them. Then hit Trace.

After it converts the image to vector, don't forget to hit the "Expand" button to bake the shapes in. Now I simply save the file as an .ai file. I use version 8, so Max will import it correctly, but newer versions of Max may handle newer .ai files.

.ai imported into Max

Why .ai files? Your initial thought may be to export from Illustrator as a .dwg. The problem with doing this is you get thousands of vertices in your spline, and it makes 3ds Max angry. Importing .ai files reduces the number of vertices and makes Max happier.

.dwg imported into Max

Now that I have a spline in 3dsmax, I simply add an Extrude modifier to the spline and watch the magic happen. You will notice that the result isn't perfect, but modeling something with this detail could simply take hours, and using this technique gets the job done in minutes!

WSOL renderings

2012-10-16T22:52:00.002-05:00

Here are the stills from the WSOL project. Even though the animation was done in v-ray, I chose to do the stills in mental ray. In the end, I preferred the lighting quality I was getting out of mental ray over v-ray. I even ran some iray tests on a Tesla C2075, and the mr results were very close to iray. Because I only have 1 dedicated GPU for rendering, I still prefer to use conventional CPU methods. For me it's still faster.

Modeling
This model became a little cumbersome near the end as the building had 3 floors and the exterior and interior renderings all came from the same model. It was primarily modeled in SketchUp. The foliage, people, and entourage are an assortment of max library files.

Lighting
Most lighting is done with the mr Daylight System, with addition of photometric lights and mr sky portals for the inside of the building. For indirect illumination I used the combination of photon mapping and final gather. For the indoor renderings the sun was set to around 1pm to get strong light into the windows. For the exterior renderings, I adjusted the sun to 7pm to give more dramatic shadows and warm tones.

Rendering
For sampling I used unified sampling to give me detail in far away objects, and faster renders with motion blur. I keyframed the movement of the people so I could get true 3D blur. These were all rendered at 4000px, with an 8 core Xeon with times varying from 1 hour - 4 hours.

WSOL

2012-10-04T15:30:00.000-05:00

Here is the latest and hot off the presses. This is an animation that I've been working on for the last month. I initially was going to animate it in mental ray, but was getting really long render times for frames. So I switched to vray, and it made the render process much easier. Here's a link of a mr / vr comparison of this project. There's not much special about the technicals. The post work was done in After Effects.

3D Camp | Houston

2012-08-29T22:02:00.000-05:00

If you're in the Houston area and like 3D mark your calendars September 29th at UofH for 3DCAMP Houston. This is a day-camp-mini-conference-un-conference-event. I will be on a panel with the talented Jorge Tiscareno and Scott DeWoody to do a session titled: Design Visualization: The Process of Visualization + Digital Design in Architecture. We will hash through various 3D processes that we use, what methods work, and what don't. I'm sure we'll get into discussing sketchup, revit, 3dsmax, mental ray, vray, photoshop...it should be good! did I mention it's $15?

see you there!

DMC vs. Unified

2012-08-21T22:08:00.000-05:00

Ok. So as we're continuing to learn together, Zap has pointed out to me, that with unified sampling, you can actually get away with using fewer individual samples and it seems to still maintain the integrity of the rendering. I managed to actually get really fast renders using unified with low glossy/DOF samples but they turned out quite grainy. Because mr and vr are not a 1:1 relationship...there are so many different things going on to really compare. I decided, rather than compare render speed/time, to make comparisons on the image alone. So my approach for this test was that I would shoot for the same render time for both engines. Taking this approach removes the argument that one is faster than the other. This way I can focus purely on the quality of the image.

v-ray (Adaptive DMC) 9min 30sec

mental ray (Unified): 9min 31sec

Based on similar times, you can see the disparity between the two. Trying to ignore the DOF and moBlur between the two and focusing on sampling, there is actually not that much difference. However I noticed that Vray seems or appears to have somewhat of a Gaussian blur look or filter to its sampling. While unified looks sharper, it also appears more noisy. I've noticed this is really evident in shadow samples from mrSkyPortals as an example. Perhaps there is a way to create a blurring of these samples at a 2D level without blurring the 3D image.

So there is no answer here, but more an open ended thought.

Until next time...happy rendering.

EDIT:

So as someone suggested, using Triangle or Mitchell filter really seems to fix the sampling. This example is using Triangle filter with the default 2,2. Oh, and by the way the render times were really good. The unified settings were min:1, max: 100, quality: 4.0, error threshold: 0.05. Also glossy, dof, and shadow samples were all set to 1. The results were actually quite good:

mental ray unifiled (Triangle filter): 2min 11sec

Unified Sampling

2012-08-16T01:45:00.001-05:00

Boy is it hard to keep up with our ever changing technology of 3D. Even though this is old news (2 years old...at least) and there are several posts about unified sampling simply by doing a Google search, I'll still bring it up. In my last post, I mentioned that I had used unified sampling for my MR renderings, and someone asked about it, so I figured I'd post my findings and impression of it.

It was incorporated with mr in 3ds max 2012. Although there is no user interface, there are maxscripts that invoke all of the unified sampling string commands thanks to folks like Artur Leao and Thorsten Hartmann. In fact the script I use is an older simplified maxscript from Artur, and you can get it here. It's a smart script that actually saves the data into the file so you can render through Backburner. It still doesn't work with distributed rendering though.

So here are my comparisons of unified sampling with standard sampling in mental ray. For all of the examples below, standard sampling was set to a min: 1 and max: 16 and a spatial contrast of 0.05 for RGB. However I did tweak the unified settings for some of the tests. These were also rendered on a 8 threaded i7 chip so nothing fancy here.

Motion Blur

For this test, I put some serious moBlur on these teapots to really push the AA engines. In general unified was much faster. However I did have to increase the min on unified to at least 4 samples to get comparable visuals. Even with increasing the min samples it was still faster than standard.

standard: 2min 50sec

unified: 1min 45sec

Depth of Field

For DOF I used the Bokeh camera shader with 16 samples and a bias of 10 for both tests. In this case for unified sampling, I used a min - max of 1 - 50, and an error threshold of 0.1 and it still turned out decent even though it has a little noise.

standard: 4min 42sec

unified: 3min 37sec

Glossy Reflections

For this test again I really wanted to push the samplers, so the glossy reflections on the teapots and floor both have 32 glossy samples. In this one I left the unified error threshold at 0.1 and you can see the noise on the edges of the teapots, however if I had decreased this my feeling is that the render would look even better without taking a serious time hit.

standard: 2min 27sec

unified: 1min 45sec

All Together Now

So for the last test, I turned everything on: moBlur, DOF, and glossy reflections. This is where unified really stands out over traditional sampling. All of these bells and whistles brought standard sampling to its knees and a screeching halt at almost 1.5 hrs of render time for a 600 px image...yikes! For this example I set unified to min: 2 and max: 50 with an error threshold of 0.1. Again even with the threshold that high it still turned out fairly decent, especially for the time.

standard: 1hr 25min 10sec

unified: 24min 9sec

Unified handles things like scattered samples quite well, especially in combination. Where I've found unified sampling especially helpful are for those pesky electric cables in my renderings that are like 100 feet away from the camera. Unified is really good at catching things like grass, fur, and real fine objects for every frame. This is really critical for animations where you may get flickering lights at a glancing angle.

I've heard some say unified sampling is mental ray's equivalent of v-ray's DMC sampling. Not sure about that though as I still find v-ray's sampling to be superior to anything out there currently. I guess now the real test is to try this same scene with v-ray and see what I get. Although it's not quite apples to apples...er teapots to teapots.

EDIT:

So I said that I would try this with v-ray. In the CPU battle between mr and vr, there is a reason that Chaos Group just celebrated a successful 10 years, and why everyone is using vray. Using Adaptive DMC at the default settings in vr 2.0, I threw all of the same bells and whistles at vray. Motion blur, DOF, glossy samples...and v-ray ate it for lunch! Not only did it render 2.5 times faster, but the quality of sampling just looks better. Judge for yourself, but my opinion, v-ray is still king!

v-ray (Adaptive DMC): 9min 54sec

BES renderings

2012-07-31T15:09:00.001-05:00

Here is a new batch of images that I've recently finished. This is another project that was done very quickly. My good friend, and even a better architect, Luis Ayala designed this building. He created the model in SketchUp. After receiving his model, I imported it into 3ds Max. It wasn't quite render ready, and there were a few things that I tweaked on the model. Much of the entourage were low poly, so I modified some of the geometry with editPoly / turboSmooths. The windows were thin planes without mullions, and I created splines from the faces, and rendered them as rectangles to give the window frames.

These renderings are somewhat of a departure from the hyper-real. Lately I've been tired of looking at renderings with too much chromatic aberration, DOF, vignetting, etc. So for this series there was very little post work done. In some ways it's almost more of a graphical series because of the colors than anything else.

A couple of technical specs:

This was rendered with mental ray. I used the mr daylight system. I created mr sky portals in most of the large windows, and created photometric rectangular lights in most of the recessed ceiling lights. For all of my lights I turn off "Automatically Calculate Energy and Photons"; essentially removing the lights from the GI calculation. I was able to get away with using the default settings for FG and GI. I also used the Bokeh shader for the Lens output to render DOF. For image sampling I used unified sampling (min:1, max: 100, quality: 1). It's much faster than the standard mr sampling, especially with DOF and glossy sampling.

flickr set of these images

VisDay Recap

2012-07-03T14:19:00.000-05:00

VisDay was a big success, and I was glad to be a part of the Inaugural event. I'm hoping we will be able to bring something like this back to the US. For me it was very reminiscent of VisMasters DMVC, and was a place where 3D Artists could get together and speak the same language.

Lon Grohs (Chaos Group) was the keynote speaker and did a great job summing up the evolution and history of arch-viz. Daniel Flood (FloodSlicer) amazed us with some wild 3D stereoscopy. Pat Carmichael (HKS) showed off some mind-blowing high poly scenes in realtime...off a laptop to say the least! Scott Ballis (Atomic 3D) shared some of the logistics of working in a full production Arch-Vis studio, and showed some of the latest and greatest coming out of Australia. Phil Read (M-SIX) unveiled some of the newest innovations in production workflows with VEO. I got to put my 2 cents in as well showing off some mental ray animations in from 3dsmax.

It was short and sweet, and reminded me of everything we love about 3D. My hat off to Dan Jürgens, Chris Needham, and Wesley Benn for putting on a great event. Here are some photos from VisDay:

Lock those cameras!

2012-06-27T09:22:00.000-05:00

Often as I'm working, and moving around in Max, I have found that somehow between switching views I've accidentally moved my camera, and can't get it back to where it was before. This is especially critical for cameras with animations on them. If they are moved, then the animation still play...they just get shifted to wherever the camera was accidentally moved to.

So once I've got my cameras set where I want them, I simply select all my cameras and camera targets and go over to the Hierarchy tab. If you click on Link Info, you will see a Locks rollout. Here you can lock the location of your cameras.

No more worrying about moving a camera. This is especially useful if you're using a 3D mouse. I've also found that if you have a 3D mouse on a computer net-rendering using server, the 3D mouse can still shift your camera...strange! Locking your cameras ensures your render will come out correctly!

Apologies for not posting as much as I would like to. I've been quite busy and I'm working on several 3D projects that I might be able to share in the future :)

RTC VisDay 2012

2012-05-19T23:31:00.000-05:00

Next week I'll be in Australia speaking at the inaugural VisDay prelude to RTC 2012. I will be presenting my process behind CSHS and how I used mental ray to do it. I will also be giving a lab during RTC, with a step-by-step process to produce similar results.

See you there!

Simple lighting that always works!

2012-05-08T14:55:00.000-05:00

I've had many people ask me how do I create lighting that works for every scene. If you're using indirect illumination or light bouncing, the answer is: there is no lighting that works for every scene. I will show you a technique that doesn't use indirect illumination, by creating a simple lighting setup that will always work, and renders really fast. So this setup is really a cheat. But it can really work to get a project out if your production time is crazy short.

Rendering with this technique

This is a trick that I learned from a friend several years ago, and I'm finally coming around to posting about it. This is also a trick that adds the Ambient/Reflective Occlusion map to your rendering (I also have alot of people ask me how to add AO to a render). Basically the idea is that you plug the AO map into an omni light, and the occlusion gets added to all of the geometry in the scene.

This trick is done using 1 omni light, and it is the only time that you should ever use an omni light if you're rendering with mental ray. If you forgot where it is, you'll have to switch from Photometric back to Standard.

The main catch to making this work is that the omni light must be placed at 0,0,0. If the light is anywhere else in the scene, the occlusion won't line up with the geometry in the scene, and you'll get really psychedelic results!

These are the settings for the omni light. Under Advanced Effects be sure to check Ambient Only. This will disable the shadow options. Also check Projector Map, and insert an Ambient/Reflective Occlusion map here (it's found under Maps->mental ray). To access its settings you can click and drag an instance of the map into a material editor slot.

I usually leave most of the settings at default. But I almost always increase the Max distance to something like 4' all the way to sometimes 10'.

To see the results of this in your renderings be sure to turn off any exposure control settings. You can also turn off any Indirect Illumination in the render settings.

That's it! What I really like about this setup, especially for complex animations, is that there is no light calculating process, and the renders from frame to frame are very stable and consistent. The soft shadow effect (ao) also adds virtually no extra render time to your finished rendering.

If you have a tight production turnaround, and are concerned about the lighting in a scene not being designed yet, this is a great solution.

Two more renderings with this technique:

jellyfish - iray style

2012-02-12T22:58:00.000-06:00

I may have not posted too much on iray for as much as I'm really liking it. So here is another iray post. Ok, for many applications, iray has become my second favorite rendering engine next to mental ray. For many reasons it can really be much better. If the scene fits into your GPU memory, then it is almost always faster than mental ray or vray. DOF is free, meaning it doesn't slow down the rendering eninge. There aren't any irradiance maps, photon maps, final gather maps, or light cache maps to calculate or read from. Which can take a lot of hassle out of the design -> production process.

iray | 1200px X 600px | 10 min.

So about a year ago my son and I were watching Finding Nemo, and during the jellyfish sequence, I thought, I would love to try to tackle those fish with my own rendering. I also wanted to get something similar using iray. So here is a little breakdown of my homage to Pixar.

So I started off with a single jellyfish itself. It's no more than a sphere for the head, and a series of planes and splines for the tentacles. The image on the left is the texture that I created for the head. I found a texture of a basketball for the bumps, then recolored it in photoshop to give it that purply-stripy look.

For the broad tentacles I wanted something that looked leafy, so I thought I could use a lettuce texture as reference. I looked for something that had a vein quality to the leaf.

So I took this leaf texture into Photoshop, copied it several times to create a tentacle arm, colored it using Hue/Saturation with the Colorize option checked until I found a pinkish-purple hue that matched the jellyfish head. I then played with the brightness/contrast to make the veins really pop.

So I could have spent alot of time modeling this tentacle, but instead I just projected this texture onto a plane. So I also created a cutout-alpha channel map to control the opacity of the plane.

So to really make the tentacles look more natural, I added a twist modifier to give it the appearance that it was thin and fluid. Then I added a shell modifier to give it some thickness. Then finally I added a bend modifier to give it the impression that they were being pulled by the head of the jellyfish.

I gave the jellyfish multiple tentacles, and added some strands simply by using some splines. Then to complete the jellyfish, I grouped it and added a bend modifier to the whole fish to really give it the impression that it's moving and swimming in water.

Here is a view from the camera showing what the rendering will look like. I intended to use DOF, so my field of view was very shallow with a focal length of maybe 80-100mm. I also added some particulate to simulate particles in the water. They were simply low poly spheres that were placed at varying distances to give the illusion of depth. They don't look like much in the viewport, but with DOF make a big difference in the overall environment.

For the caustics, I actually faked it with a neat trick. I created a plane that sat on top of the jellyfish. I applied an A&D material to the plane that had a caustics texture map to control the cutout opacity. I used a program called Caustics Generator, that creates fake caustic texture maps. It can also create image sequences of moving caustics, which is great for animations.

I then created two spotlights above the plane so that light would only be cast through the caustics openings simulating a caustics look. I also created a light from behind the camera to give it a photographers flash look. So it felt that the lighting was coming from the camera itself.

One, more thing. I did try this scene out with mental ray, and it was crazy slow compared to iray, especially with DOF turned on, running over an hour long. Even though I love mr, this was an example where iray was much better. Much like a mechanic has different tools to fix a car, so should 3D artists be familiar with all of their tools to get the job done correctly. Mental ray, Vray, Maxwell, and Iray are all great render options, but there is usually a best option depending on the situation, so don't ge too attached to a rendering engine...technology is changing too fast for that!

SpaceMouse Pro

2012-01-31T22:46:00.000-06:00

I have to admit that during the holiday break, I couldn't stay away from 3D software. I also found that as I would navigate through 3D space, my left hand would naturally reach for my Space Navigator that wasn't even there. That's when I realized that I really did use this thing. So now I'm convinced that the Space Navigator is a must have.

Well 3Dconnexion late last year released their newest product, the SpaceMouse Pro. It's a hybrid between the SpacePilot Pro and the SpaceExplorer. After giving it a test drive for several weeks, I have to admit it really is their best product yet. Not only is it sleek looking, but it has just the right buttons that need to be used and no more than that.

Probably the most distinguishing buttons that the SpaceMouse Pro has are the large buttons on top. I found these to be really useful for setting up tools that I use often in SketchUp & 3ds Max. The great thing about assigning your regular tools to these buttons, is once they're assigned, you can remove them from your menu bars and free up even more space in your viewport.

I'm using SketchUp as my example, but this applies to any software. In SU, I assigned 1,2,3,4 to my visual styles toolbar. These are styles that I switch between alot during modeling, and now that they're assigned to these buttons, I can simply remove the tools from my palette. What's even better is they're positioned right where I need them as I'm navigating so I don't have to take my hand off the navigator itself.

Another toolbar in SU that I constantly use are Views buttons to jump to my top, front, right, left, etc. Now instead of needing the toolbar on, or even keyboard shortcuts, this mouse has these buttons built right in. I guess the designers knew that 3D artists use these views alot for editing models and figured to add them in. Of course if you want to customize these buttons, the software with the driver allows you to change the buttons to any function you want.

Along with the views button is the addition of some other frequent buttons used to navigate in 3D, such as shift, ctrl, alt, and esc. It might be genius, because adding these buttons next to the navigator can almost mean never letting go of the navigator in SU when inference snapping a line to lock to an axis even during a pan!

Lastly, the SpaceMouse Pro has a Fit button, that zooms into objects visible in your scene. Which is very handy, and allows you to do this, again, without leaving the Navigator.

Really all of these buttons could be used on the keyboard as shortcuts, but the beauty about having them around the orbiter is that you don't have to lift your left hand as often to leave your navigation just to do simple functions such as shift and esc. I would recommend this tool for anyone who is a serious 3d artist.

A word of caution though...once you start using it, just be sure to take it everywhere with you or you'll find yourself reaching for something that's not there!

Merry Christmas!

2011-12-23T10:46:00.001-06:00

It's been a good 2011, and thanks to all the readers and comments posted here! Merry Christmas to all of you guys! Hoping to add more posts in the new year.

The above image was rendered using iray. The tree was downloaded from evermotion.com

Is That a Photograph? Architectural Photography for 3D Rendering

2011-11-29T09:16:00.001-06:00

My virtual course goes live today! Best of all it's free for you to see. You have to register for an AU account, but it's worth it, because there are many classes worth viewing. You can also download all the handouts for the courses as well.

The course description:

It is not enough to know how to create great 3D renderings. You have to make images that really sell, and to do that you need the knowledge of a photographer. This class covers worthy architectural photography principals and explains how to apply them to 3D renderings. You will learn how to translate good image composition, lighting, and staging to your renderings. This class will also discuss photographic phenomenon such as vignetting, barrel distortion, glare, and other issues and describe how to fine-tune these photographic flaws to your advantage to create a rendering that really sells.

Studio Lighting Setup

2011-11-17T12:04:00.001-06:00

Showcasing your next best product in a white studio lighting environment seems very simple. In actuality, doing this correctly or "physically accurate" is really more complex than you'd think.

If you tried to render the above image the way you thought you should do it, you would probably would have set your background environment to white, then place all of your objects on a white plane, then hit render. But this is what you get:

It's not exactly what you wanted. So to fix this, the answer lies in a production material called the Matte/Shadow/Reflection shader. This is a material that does some nice trickery and saves you alot of work to render your scene with a different "environment." This can be found under the mental ray materials in the Material Map Browser. Simply apply this material to your ground plane. Of course you'll have to change the "Camera Mapped Background" color to white if that's what you want.

The scene setup is very simple for this example. I have 2 photometric area lights, and a plane that my teapots are sitting ontop of. That's it. Simply apply the Matte/Shadow/Reflection material to your plane and you're done...you have now created a white studio light setup. You may also notice that the objects in the studio rendering, are lighter than they are in the regular rendering. This material renders your objects to bounce light as if the scene were white, so your objects are also reflecting the white scene, hence physically accurate lighting.

The MSR material can also be used to create mattes to render certain objects that need to reflect an environment, and can be composited later in post.

There is one thing to note: I've found this material to act really weird if Exposure is turned on. So before you start setting your lights, bear in mind you'll want to set up your scene without exposure.

This is what happens if you use the MSR material with mr Photographic Exposure Control...weird!

Maxwell Strikes Back

2011-11-11T16:39:00.001-06:00

So way back when, we all thought Maxwell was going to be the answer and the end all be all for renderings. The visuals were promising, but the workflow was painful, and always felt like it broke my production process. They had a plugin for SketchUp and for 3dsmax. Both versions required Maxwell's own 3rd party software rendering program, and so it wasn't really rendering within the software of choise. Vray for SU never really worked right either. Adjusting materials were cumbersome, but it just didn't like some large scenes built in SU. So I continued to stick with my tried and true workflow: model and texture in SU, import into 3dsmax, add lights, tweak materials and render.

Well perhaps Maxwell has learned from its past, because they just released a brand new plugin for SU. It's so streamlined into the software, infact there isn't even an install. Just a maxwell.rb file with a folder that gets placed in the plugin folders and that's it. If that wasn't good enough, they provide it for free, which allows you to render up to 800px. Now if you want to render larger images, the licensed version is nothing to kill your wallet at $95!

The icons are very simple aside from the buttons that give you scene stats, take you to Maxwell's website for documentation, and the help button, there are very few options. The Fire button simply opens the render frame buffer window up. When you hit the fire button, it bakes everything from your scene so that it can render. This sounds painful, but it really isn't. For large SU files the longest it took was a mind blowing 30 seconds. At that point the frame buffer window becomes a realtime render window much like VR-Realtime. The pencil icon in the middle opens your settings window. The eyedropper icon selects your materials. The disc looking icon and the f.d. ruler select your DOF, and make choosing it very easy.

The settings that really sold this to me were the materials. They are so simple, and they are almost a seamless 1:1 with what you have already created in SU. So if you tweak a material in SU it gets tweaked in Maxwells material browser. So Maxwell just has the same materials that you already created, with some extra settings. You can change the materials reflection type to tell it to render as metal, lacquer, velvet, light emitting, satin, etc. But what is really clever is that MW automatically creates a grey-scale bump map for your texture map....do I hear a "sweet". It gives you options to invert. What is even cooler is if you change the brightness or color on the map in the MW settings, it changes them in your SU material settings too. Of course if you really want to get fancy, you can unlink the MW texture from SU, but it's just a toggle and you can always turn it back on at any time!

The camera settings are what you would think they are. There are parameters to specifiy the render size. Then there are the camera settings themselves. Much like mental ray in 3dsmax, you can specify your fstop, shutter and ISO, or just your EV value. The cool thing here is that changing your fstop affects your amount of DOF just like a real camera. Also you can choose if you want the focus to work manually, or semi-automatic, or totally automatic, based on the scene objects and distances.

The next settings are the environment or lighting settings. The first option is to choose your type of skylight. You can use the default sky dome, the physical sky (which is linked to your sun angle from SU), or a simple hdr map of your choice. Then you have the option to turn the sun on or off along with other atmosphere settings like turbidity, ozone, etc.

Then, much like vray, you can toggle a ground plane on and off as well.

The last settings I'm going to cover are some custom settings. You can tell MW to use just front or both front and back faced materials. Also how it will respect hidden layers. There is also a material override option for simple testing. Then there is a section for color management. you can tell it to use sRGB, or the myriad of color options. The burn tells MW what to do with highlights, and then there is gamma.

The last section, simulens, I haven't played with yet, but have a hunch that it's really cool. I'm guessing devignetting is simply vignetting in your rendering. There are options to add aperture maps, which i'm guessing control the way DOF creates its bokeh effect. But these are all photo-phenomenon tools that will add realism to the rendering.

That's pretty much it. I was really amazed at how simple and robust the UI was. I'm not sure if its because my computer is any faster but I could guess that the rendering algorithm is faster at computing the data. It seemed that my renders were going quite quick, and gives iray a run for its money. Where some of my scenes wouldn't fit into memory for iray in 3dsmax, maxwell in SU was able to handle them fine. I have a feeling it has to do with SU's small memory footprint, then again, there may be some wizardry coming from the Spaniards!

20min - Still some speckling as I had a huge copper cylinder casting caustics, but not bad

UPDATE (2012-02-02): So just recently downloaded the latest Maxwell for SU, and they did make some revisions in their software. Unfortunately the software now requires an .exe install, and is a little larger, but hopefully that will only make it faster. Also it now requires Microsoft Silverlight, but that's not so bad.