Fujifilm has recently announced its first ever medium format fixed lens camera, the GFX100RF. (For the record, RF = “Rangefinder style + Fixed lens”.) Priced at $4899 USD, it is also one of the least expensive (relatively speaking) medium format digital cameras on the market. The press release issued by Fuji contains links to all of the technical details and marketing fluff surrounding this interesting, but very niche camera.

Niche?

Fujifilm has made some very bold decisions regarding the design of this camera. Following the mold of the X100 Series, the GFX100RF is a rangefinder inspired fixed lens camera sporting the sensor and processing power of the GFX100SII. The lens is 35mm f/4 (equivalent to 28mm f/3.1 in 35mm full frame) and due to its wide field of view, IBIS was left out of the camera, unlike the X100VI which has IBIS.

One has to realize that medium format is not tolerant of one-hand camera operation regardless of how much IBIS it has. Medium format’s mantra is “details, details, details” and to capture this requires a rock-steady hand or (more often) a tripod.

This excludes most of the masses of photographers who shoot off the cuff. The more bourgeoisie will likely get one because it is very expensive and has lofty air about it and then return it because it actually demands photographic skills to create amazing images. These cameras are likely to be returned and I will be picking up one of those second-hand.

“Anyone can cook, but not everyone is a chef.” – Chef Auguste Gusteau, Ratatouille

I am in no way impinging on the design and engineering skills from the House of Fuji. Used properly this camera makes stunning images. All of the rules for sharp image capture apply here. The camera is ideal for landscape and travel photography, greatly reducing the amount of kit you have to drag around with you. For handheld work, you must be prepared to crank up the ISO to have sharp images. ISO noise is extremely manageable (I have tested this myself with sample RAW files graciously provided by DPReview) and with it turned up, faster shutter speeds can be used to offset hand movement. Due to copyright, I cannot show the images used in the DPReview article, however I can show a before and after crop showing the results of noise reduction on a high-ISO image.

Amazingly, Adobe Lightroom Classic can read the GFX100RF RAF files without issues. The original RAW file is on the left and the noise reduced, color profiled (I prefer ETERNA Cinema for night work) and perspective corrected DNG file on the right.

Traveling Lighter

My travel photo kit is the Fuji X100VI + Fuji Wide/Tele Adapters, EF-X20 Flash, spare batteries, charger and SD cards. It all fits in a small Tenba shoulder bag. The GFX100RF can make for an even smaller kit despite being slightly larger than my X100VI. Fuji has added a digital zoom feature to take advantage of the plethora of pixels provided by the GFX sensor. A rocker switch just below the front dial under the shutter release allows to switch from 35mm -> 45mm -> 63mm -> 80mm by cropping the image. You drop from 102 to 20 megapixels at the extremes, but 20 megapixels is fine for many use cases these days and those are a whole pile of lens adapters you don’t have to buy.

Another interesting feature is the addition of an aspect ratio dial. Following logically from the crop zoom approach, the aspect ratio dial allows you to switch between the aspects various classic Fuji cameras. JPEGs and HEIF files are hard cropped to the selected aspect ratio, but RAW files simply have the crop layered on and it can be modified in post.

Wrap up (for now):

The GFX100RF is quite a statement by Fuji, one of the few camera companies that do things for the love of photography and not just the bottom line. Fuji is not totally impractical, either. Component reuse and design decisions have lowered the barrier to entry in the medium format space. I hope to pick one of these up eventually and provide a more in depth report.

UPDATE

My local camera dealer (AVC Photo in Miami, FL) had a Fuji Takeover Day on April 4, 2025 and I attended. The GFX100RF was present in all of its glory and I got to play with one for a little while.

Fuji has done their homework with this camera. The lack of IBIS and mediocre lens speed (f4) is nicely offset by exposure profiles that favor keeping the minimum shutter speed in the automatic modes to 1/125 of a second. This, coupled with the excellent leaf shutter from the X100 series, does help mitigate hand movement.

Another fun feature I tried was the digital zoom and the aspect ratio dial. The image above was shot at 65:24 aspect ratio (the same ratio as the Fuji TX-1/Hasselblad Xpan) and makes for a quick one-shot panorama without the hassle of the barrel distortion created by panorama modes that involve sweeping the camera. Also, if you record in raw, the crops are an overlay and can be altered or removed in post-production. Please note that raw files are full sized (200 MB) regardless of crop, aspect ratio, or film simulation used. JPEGs and HEIF files are permanently cropped or filtered.

That’s it for this update. I am seriously contemplating this camera now.

Here is a quick round up of things I find interesting:

1. Darktable 4.8 has been released. Head over to https://darktable.org for more details.
   
2. I have traded in my Fuji X-H2 for a Fuji X100VI. I even traveled to Peru with it. Great travel camera in spite of what the naysayers are spewing.
   
3. Apple has introduced the M4 chip in an iPad of all things.I was hoping for the Mac Studio to get some M4 love, but that may not happen until next year which will see the release of the M5. This means that they will either: release it with an M4 family chip which will be immediately superseded by the M5 OR delay it again to put an M5 in it making us wait another year and further alienate the Mac desktop users.
   
4. I have managed to 3D print a medium format film camera. Many thanks to Dora Goodman and the team at Dora Goodman Cameras. More on this in another post.

The following is a summary of the main features added to darktable 4.6. Please see the user manual for more details of the individual changes (where available).

• Editing history is now periodically auto-saved (every 10 seconds) while editing images in the darkroom view. The auto-save interval can be changed (via a preference) or auto-save can be disabled entirely (by setting the interval to 0).
• A new processing module rgb primaries has been added. This module can be used for delicate color corrections as well as creative color grading. It allows the red, green and blue primary colors to be moved around using “hue” and “purity” controls. The underlying pixel operation is essentially the same as channel mixing.
• In addition, the sigmoid module now includes a new primaries section, which can be used to gracefully handle difficult lighting situations (e.g. LEDs) and tune the overall look of the image. Modifying these parameters can provide pleasing sunsets, improved skin tones etc. This feature can only be used with sigmoid’s per-channel mode and is loosely based on ideas from Troy Sobotka’s AgX and related work in the Blender community. The included “smooth” preset should provide a good starting point for further corrections using this feature.
• When working with the liquify and retouch modules, the full uncropped image is now always shown, with any crop indicated by displaying an overlaid rectangle. This allows for cropped-out parts of the image to be used without having to first disable the crop module and re-enable it when finished.
• When panning or zooming in the darkroom view, a low resolution placeholder used to be shown until the image was fully recalculated for the newly-visible region. Now, if any part of the previous view is still visible, that part will be immediately shown in high quality, with only the remainder of the image being temporarily shown in low quality until the pipe has finished. This has been achieved as part of a complete reworking of the darkroom image display, which has also removed some annoying jumps when replacing the low resolution placeholder with recalculated image data, or when switching between full and cropped view (when the crop, retouch or liquify modules are (un)focused).

Performance Improvements

• OpenCL is now initialized in the background immediately after launching the darktable UI, with progress indicated by a series of toast messages. Previously, this was done before the main UI was launched, which could lead to significant delays (sometimes more than a minute) between the user opening darktable and anything appearing on-screen, making it appear as though darktable had failed to start. Note that processing will be slower (since darktable will only use the CPU) until OpenCL is fully initialized.
• An OpenCL code path has been implemented for lens correction using embedded correction metadata.
• Image display speed in the map view has been increased by 25%.
• Export speed has been improved for JPEG 2000 and B&W TIFF images.
• The chromatic aberrations module is now approximately 10% faster when run on the CPU.

Other Changes

• The hot pixels module now supports monochrome images.
• The internal compositional guide widget has been removed from the retouch module as it is not required (you can still display the guide globally).
• A long-running import session can now be canceled by clicking on the cross in the progress bar on the bottom left of the screen.
• Auto orientation is now supported when importing AVIF/HEIF images (requires at least libavif 0.9.2 and/or libheif 1.16.0).
• It is now possible to visualize raster masks in the same way as other mask types by clicking the mask visualization icon.
• The dual demosaic mask visualization and performance have been improved.
• It is now possible to manually define additional vignette correction via the lens correction module.
• The linear ProPhoto RGB color space can now be selected in the LUT 3D module.
• The following changes have been made to the OpenCL implementation:
  • Benchmarking code has been removed as it no longer produces valid results on today’s computers.
  • The “pinned memory transfer” option has now been removed from the preferences dialog, though it can still be set on a per-device basis.
  • A new option “use all device memory” has been added to the preferences dialog. Select this option to use all GPU memory, with the exception of a 600MB “headroom” which is left free for use by other applications. A similar option was previously available in earlier versions of darktable, however the headroom is now increased to 600MB to account for greater use of GPU acceleration on modern systems.
  • A new set of options have been added to the preferences dialog, to allow selection of desired OpenCL drivers.
• A built-in “Display P3” color profile has been added for use in input, output, display, working and soft-proofing profiles.
• The highlight reconstruction module can now be used for all non-raw files.
• The scroll speed in the filmstrip can now be increased by holding Shift while scrolling (filmstrip will move by half of the visible images at a time). Holding Ctrl while scrolling changes the number of images shown (zooms the filmstrip) and thereby also changes the speed achieved when holding Shift.
• Exported PNG files now embed a CICP (Coding-Independent Code Points) encoded color profile in addition to ICC where possible.
• Exported AVIF files now no longer embed a superfluous ICC profile if the color profile can be encoded as CICP.
• Improved scaling and placement of images in the culling view now make better use of available screen space.
• When hovering over the sample patches in the global color picker module those areas are displayed on the central image and on the histogram (if the corresponding option is selected). It is no longer necessary to have the color-picker activated when doing this. This enhanced behavior is useful, for example, when color grading.
• It is now possible to display a tooltip for a slider or drop-down and a separate tooltip for the button to its right (commonly a color picker button). This was not possible in previous versions of darktable, and was an issue where the button was not directly linked to the slider (for example, a button for mask display).
• Holding the Ctrl key while double clicking a slider or drop-down in a processing module now restores any auto-applied preset.
• Where multiple drawn shapes are grouped (e.g. in a module’s mask) the mask manager module now displays a tick mark in front of the active combination mode (in the popup menu) when right-clicking one of the constituent shapes. This is usually much easier to read than the “Venn diagram” icons in the main module UI.
• Mnemonics have been added to dialog boxes and default actions have been assigned, so that pressing Enter will trigger the correct action and close the dialog.
• More OpenEXR 3.2.0 attributes can now be extracted and displayed in the image information module, where present.
• Lens and cameras filters have been added to the collection filters module.
• AVIF exports have been changed so that there is now no conversion to YUV for lossless. The quantizer selection logic has been updated and “lossy” is now the default.
• The content of the main preferences dialog is now (almost) entirely taken from a single definition file (darktableconfig.xml.in) making its layout much easier to modify.
• The drop-downs in the preferences dialog now use the same (bauhaus) widgets as the drop-downs in the rest of the application, offering a more consistent behavior.
• The configuration system now has improved number precision.
• The collections module has been reworked for more consistent sorting. All date/time sort orders can now be reversed (placing either the oldest or newest first). The film roll order can also be reversed when ordered by id (which is equivalent to the film roll’s first import date). The “sort film rolls by” preference has been renamed from “id” to “import time” to make this clearer. The collection filtering based on rating now uses proper text (like “rejected” instead of -1) and the numbers are replaced by stars. The collection module now displays color labels in the same order as in the rest of the UI.
• Adding a tag to an image is now considered to be an “edit” and causes XMP sidecars to be created/updated (unless XMP creation is disabled).
• New functionality has been added to automatically generate thumbnails in the background while the user is inactive in the lighttable view. Currently this functionality is only available by running darktable-generate-cache from the command-line. The new functionality is controlled via a preference setting that defines which thumbnail sizes to generate in the background (default “never”).
• Two new variables have been introduced to allow the camera crop factor (“EXIF.CROP_FACTOR”) and 35mm-equivalent focal length (“EXIF.FOCAL.LENGTH.EQUIV”) to be displayed.
• Color harmony guides in the scopes module are now saved and restored for every image. It is therefore no longer necessary to reset the guides when going back to a previous edit. This information is also stored in the XMP file so can be viewed when sharing edits or re-importing a collection.
• It is now possible to control chroma subsampling in JPEG exports. This allows the user to reduce the color resolution, often resulting in much smaller files that are virtually indistinguishable from images with more color information. On the other hand, certain images will look better when the chroma resolution is maximized. This includes images with small colored details surrounded by a solid background (such as screenshots with colored text).
• The option to ignore JPEG files when importing is now outdated, given that many cameras can also output HEIF images. This option has therefore now been changed to ignore all non-raw files.
• The lens correction module now supports embedded metadata from Olympus .ORF files for correction of distortion and chromatic aberration. This correction is equivalent to that applied by the camera body to in-camera JPEGs. For older bodies that do not apply chromatic aberration correction to the JPEGs, the required information is also excluded from the embedded metadata so darktable will be unable to apply corrections. Vignetting correction based on embedded metadata is not supported. However, if the camera’s “Shading Compensation” option is enabled, the vignetting correction will already have been applied to the data in the raw file.
• The shortcuts system has received several refinements:
  • Deleting or overwriting a default shortcut now moves it to the “disabled defaults” category from where it can restored by pressing Delete. It is no longer necessary to uncheck “load default shortcuts at startup” (in prefs/misc/interface) to keep it disabled.
  • Visual mapping mode now has improved mouse cursors to indicate whether the widget under the cursor can have a shortcut assigned or be added to (or removed from) the quick access panel.
  • The shortcuts tab in the preferences dialog now explains that it may be more convenient to use visual mapping mode.
  • When combining a shortcut with a mouse move (for example b+scroll), separate actions can be triggered by up and down moves. For example, b+scroll-up could cycle through the top panel options and b+scroll-down through the bottom panel combinations.
  • A problem with drop-down and slider popups opened via a shortcut, whereby they would immediately fill with the shortcut key character, was resolved. Those popups are now integrated into the shortcut system, so most non-alphanumeric shortcuts will continue to work. If a dropdown value is changed while the popup is open (for example via an automatic calculation in color calibration, a Lua script or a MIDI shortcut), the popup is correctly updated/repositioned. After a popup is closed it can quickly be reopened to enter another value by pressing Enter (as long as the corresponding widget still has focus).
  • On macOS the system shortcuts e.g. for copy and paste will now work according to the macOS keyboard defaults, meaning using the command key instead of the original Linux style control key.
• The ISO 12464 color assessment mode for the secondary preview window is now activated independently from the main window (and saved between sessions) with a toggle in the right-click popover of the “display second window” button (or using the default alt+b shortcut).

Bug Fixes

• Fixed issues with the OpenCL platform checks that were causing freezes.
• Fixed the size calculation of resizable widgets based on line size of their contents.
• Fixed a bug in the collection filters module where the conjunction of multiple filters was not handled properly.
• Fixed focus distance detection for Nikon Z mount lens corrections.
• Fixed incorrect cropping of sensor data for sRaw dng files.
• Apply the Lightroom color matrix only when importing a genuine Lightroom XMP sidecar file.
• Fixed a crash when increasing the number of recent collections.
• Fixed a crash when clicking+dragging the feather line on the path mask.
• Fixed a crash when applying the CMYK soft-proof ICC profile.
• The white borders for ISO 12464 color assessment (toggled with Ctrl+b) are now correctly sized and placed at all zoom levels and no longer flash when switching between low and high quality preview, both in the center view and secondary preview window.
• Fixed an issue downloading to Piwigo when the “on conflict” option was set to “don’t check”.
• Fixed several mouse scroll-wheel issues on macOS when used in combination with the Shift modifier key (color harmonies width, module height, geotagging date/time, …).
• When loading an XMP sidecar file from disk, resolved an issue where the assigned “duplicate number” was already present in the database. This has been resolved by instead recalculating the “duplicate number” at import time for all duplicates.
• For large collections on the lighttable and when using small thumbnails (more than 15 per line), using the scrollbar or scrolling the mouse to move up/down could be slow. The responsiveness of the lighttable in this scenario has now been drastically improved and is fast and responsive even with ~50k images and 20 thumbnails per line.
• Fixed pixel errors in RAW Chromatic Aberration and LMMSE demosaic resulting in decreased noise.
• Fixed the ‘avoid colorshift’ mode in the RAW Chromatic Aberrations module.
• Fixed the snapshot module so that snapshots for removed images are no longer accessible – this was previously causing crashes when attempting to display a snapshot of a removed image.
• Fixed a potential crash when using an unsupported ICC profile for soft-proofing.
• Fixed multiple issues when computing borders in the framing module. For a 0% border on the bottom or right-hand side a single pixel border was sometimes shown, depending on the export size/zoom level. Note that a 0% border doesn’t necessarily mean no border at all if the chosen aspect ratio doesn’t correspond to the aspect ratio of the image.
• The internal luminance mask in the tone equalizer has been made more resilient to distortion changes made with modules like lens correction, crop, etc. After activating a crop, the tone equalizer cursor will now show the proper mask luminance values when hovering over the image in the darkroom.
• Fixed calculation of required graphics memory for modules performing blending operations, possibly avoiding crashes related to OpenCL.
• Added the ability to calculate the crop factor for those cameras that do not include this information in their Exif metadata.
• Fixed various bugs related to feathering masks.
• Fixed a DPI scaling issue on macOS causing some GUI elements being rendered too small.

Lua

API Version

• API version is now 9.2.0

Bug Fixes

• Fixed scripts_installer to handle user names with spaces on Windows.

Add action support for Lua

Other Lua changes

• Allow access to image change_timestamp.

Notes

• When exporting to AVIF, EXR, JPEG XL, or XCF, selecting specific metadata (e.g. geo-tag or creator) is not currently possible. For AVIF, EXR, JPEG XL, and XCF formats, darktable will not include any metadata fields unless the user selects all of the checkboxes in the export module’s preference options.
• In order to support the correct display of numbers in darktable, the minimum supported Gtk version has had to be increased to 3.24.15. For people who need to build darktable with an older version, this can be achieved by removing line 241 of the darktable.css file on your system. See https://github.com/darktable-org/darktable/issues/13166.
• Starting with release 4.4 a new support policy regarding macOS versions has been put in place – darktable releases will now only support those macOS versions that are also supported by Apple. Release 4.6 therefore drops support for macOS versions older than 12.5.

Changed Dependencies

Mandatory

• ???

Optional

• Bump libavif to 0.9.2

RawSpeed changes

Camera support, compared to 4.4

Base Support

• Canon EOS Kiss F
• Canon EOS Kiss X50
• Canon EOS Kiss X90
• Canon IXY 220F (dng, chdk)
• Canon PowerShot SX220 HS (chdk)
• Fujifilm FinePix SL1000
• Fujifilm X-S20 (compressed)
• ImBack ImB35mm (chdk)
• Leica M10-D (dng)
• Leica M10-P (dng)
• Leica M10-R (dng)
• Leica M11-P (dng)
• Leica Q2 Monochrom (dng)
• Leica SL2 (dng)
• Leica SL2-S (dng)
• Nikon Coolpix A1000 (12bit-uncompressed)
• Nikon Z f (14bit-compressed)
• OM System TG-7
• Panasonic DC-TZ200D (3:2)
• Panasonic DC-TZ202D (3:2)
• Panasonic DC-TZ220 (3:2)
• Panasonic DC-TZ220D (3:2)
• Panasonic DC-ZS200D (3:2)
• Panasonic DC-ZS220 (3:2)
• Panasonic DC-ZS220D (3:2)
• Panasonic DMC-G10 (3:2, 16:9, 1:1)
• Panasonic DMC-GM1S (4:3, 3:2, 16:9, 1:1)
• Pentax K-3 Mark III Monochrome
• Pentax KF
• Samsung EK-GN120
• Samsung G920F (dng)
• Samsung G935F (dng)
• Samsung GX10 (dng)
• Samsung GX20 (dng)
• Sigma fp (dng)
• Sigma fp L (dng)
• Sigma sd Quattro (dng)
• Sigma sd Quattro H (dng)
• Sinar eVolution 75
• Sony ILCE-6700
• Sony ILCE-7CM2
• Sony ILCE-7CR
• Sony ILME-FX3
• Sony ZV-E1

White Balance Presets

• Leica M (Typ 240)
• Olympus E-P7
• Olympus TG-6
• Sony ILCE-7CM2

Noise Profiles

• Canon IXY 220F
• Nikon Z f
• Olympus E-P7
• Pentax *ist DS
• Sony ILCE-6700

Missing Compression Mode Support

• Apple ProRAW DNGs
• CinemaDNG lossless (Blackmagic, DJI, etc.)
• Fujifilm lossy RAFs
• Nikon high efficiency NEFs
• Samsung Expert RAW DNGs
• Sony downsized lossless ARWs (“M” for full-frame, “S” for full-frame & APS-C)

Suspended Support

Support for the following cameras is suspended because no samples are available on raw.pixls.us:

• Creo/Leaf Aptus 22(LF3779)/Hasselblad H1
• Fujifilm FinePix S9600fd
• Fujifilm IS-1
• GoPro FUSION
• Kodak EasyShare Z980
• Leaf Aptus-II 5(LI300059)/Mamiya 645 AFD
• Leaf Credo 60
• Leaf Credo 80
• Minolta DiMAGE 5
• Olympus SP320
• Panasonic DMC-FX150
• Pentax Q10
• Phase One IQ250
• Sinar Hy6/ Sinarback eXact
• ST Micro STV680

Translations

• Albanian
• Brazilian Portuguese
• Chinese (Simplified)
• Chinese (Traditional)
• Czech
• Dutch
• European Spanish
• Finnish
• French
• German
• Hungarian
• Italian
• Japanese
• Polish
• Russian
• Slovenian
• Ukrainian

Since the dawn of the graphical user interface, computer users were always on a quest to show better images on their screens. The problem was image file size. Back in the early 90’s, you were quite the techno-hipster if you had a 28kbps dial up connection and had connections with the US Defense Department if you had something faster. This is an IV drip compared to the multi-gigabit data torrents available to many consumers these days. Since getting a fatter connection was financially out of the question for most, the IEEE formed the Joint Photographic Engineering Group to figure out how to make high-color, high-resolution images much smaller than their TIFF counterparts. Working like badgers, the group published their specification and the JPEG file format was born. However, the spec is not a panacea and while it is ubiquitous, it has several well known shortcomings.

Fast forward to the digital camera revolution.

Early cameras adopted the JPEG file format exclusively and that became a barrier to entry for the professional photographer who, entrenched in the world of analog film, did not want to give up the flexibility having a film negative gave them. Camera makers responded with digital SLR cameras that were able to save in other file formats including TIFF and the native RAW file the camera’s sensor would produce. This latter development (no pun intended), with the addition of software applications to non-destructively edit the raw images and camera bodies that would allow pros to use their existing stable of lenses, is what opened the floodgates and started the practically wholesale conversion of the professional photo industry over to digital photos. In 2015 the “Moving Pictures Experts Group” (aka MPEG) published the finalized specs for the “High Efficiency Image File” format.

Fast forward to the smartphone revolution.

Apple ushered in this particular revolution and many, many competitors followed suit. All of these devices contained cameras of ever increasing quality with the ability to not only take photos but videos as well. However, these devices were hamstrung at first by a lack of storage (solid state drives lagged way behind their “spinning rust” counterparts) and someone had to science their way out of it. By about the time of the iPhone 7, Apple adpopted the High Efficiency Image File, or HEIF format for short, and inserted it into the newly released iPhone (there is an advantage to owning the ball, the bat, the mitts, bases, baseball stadium, the concession stands and the surrounding parking lot) and updated iOS 11 to seamlessly process them. The file was more compressed than an equivalent JPEG, supported 10-bit color long before anyone else, and lossless to boot. The only shortcoming was that it was not well supported outside of Apple’s walled garden.

Fast forward to present day.

Despite its flaws, pretty much all digital cameras to this day retain the ability to save images in JPEG. It’s still that ubiquitous. However, some camera makers (*cough* Fujifilm *cough*) have made the HEIF file format available on their cameras. Just a quick setting change and your JPEG files are now being saved as HEIF files. Adobe Photoshop, Adobe Lightroom, and the open-source darktable all support the HEIF file format with more and more being added every day. This is a good thing.

Why not use RAW, smart guy?

RAW files are amazing things in the hands of a professional who has a complete understanding of the strengths and weaknesses of the format. Exposure errors, white balance gaffs, and host of other things that can go wrong (short of leaving the lens cap on) can be addressed in post-production. But it also has the potential of becoming a crutch. “Fix it in post” is not a mantra I endorse. It promotes sloppy photography and leads to an endless parade of mediocre images. “Get in right in the camera.” is what I practice and it is what I teach my students. Let’s face it, I prefer to be out capturing images instead of being trapped in front of my computer post-processing images. Ansel Adams loved working in the darkroom. I am not Ansel Adams. Another shortcoming of shooting in RAW is that all of the fabulous film simulations provided in-camera by Fuji are aspirational only as the RAW file does not record that bit of post-processing in itself. I have shot RAW+JPEG for many years and I could never nail the look exactly.

Shortly after I picked up my X-T5 for walkabout camera duties, I discovered the option to record in HEIF instead of JPEG. Thus began my journey into the world of film simulations that frustrated me so often in the past. HEIF’s lossless 10-bit recording with baked in film simulations is an excellent “RAW-lite substitute”, providing editable source files that can be easily exported to JPEG for general consumption. The space savings are significant vs RAW (12.7Mb vs 47.9Mb in most cases – a 75% space reduction) also making it faster to load into memory for editing.

The left image above was shot in RAF and converted to JPG (with default settings) using darktable. The image on the right was saved in HEIF format with the Provia film simulation and exported to JPG using darktable. Both images were taken simultaneously with my X-H2 + TTArtisan 27mm f/2.8 lens writing RAW to the CFExpress card and HEIF to the SD card.

Since both images are shot with the Standard (Provia) profile, they are the closest in appearance. It’s when you start using other film simulations that things get tricky.

Here we have another pair of images this time with the Velvia simulation activated. Again RAF on the left and HEIF on the right and the same rules as before. The RAF discarded the simulation information and has the Provia profile applied. The HEIF file has the Velvia “baked in” and is preserved even with the JPG export. A lot less work and a lossless format to boot.

Hardware Support

According to Wikipedia, the following hardware supports HEIF:

• The Canon EOS-1D X Mark III, Canon EOS R5, Canon EOS R6 and Canon EOS R8 cameras use the HEIF format to capture images in an HDR display format that use the PQ tone curve, BT.2100 color primaries and 10-bit.^[55][56] “We’ve moved on to HEIF files,” Canon said in 2019.^[57]
• The Sony α1^[58] and Sony α7 IV^[59] offer capturing images in 10-bit HEIF format with an HDR format that uses HLG.
• The Fujifilm X-H2S^[60], Fujifilm X-H2^[61], and Fujifilm X-T5^[62] offers a choice of JPEG or 10-bit HEIF file formats.
• The Nikon Z9 and Nikon Z8 offer 10-bit HEIF file formats.
• Multiple Qualcomm Snapdragon SoCs support capturing images in HEIC format (e.g. Snapdragon 888,^[63] Snapdragon 662^[64]). Some of their latest SoCs also support capturing in HEIC with HDR (e.g. Snapdragon 8 Gen 1,^[65] Snapdragon 780^[66]).
• The iPhone 7 and later devices from Apple can capture media in HEIF or HEVC format.^[67]
• Android smartphones like Xiaomi 12, OPPO Reno 7 5G, Samsung Galaxy S21 5G can capture images in HEIF format.

Conclusion

While JPEG reigns supreme for cross-platform compatibility, and RAW for post-processing flexibility, HEIF weighs in with features the modern photographer wants: lossless images that are vastly smaller than RAW, 10-bit color, and in the case of Fuji cameras, benefit from advanced film simulations baked into the images. For paid photo shoots, I will shoot RAW + HEIF, but for my personal work HEIF meets all of my needs.

Introduction:

Darktable, the powerful open-source photography workflow software, has recently unveiled its latest major release, darktable 4.4.0. Packed with exciting features and improvements, this update aims to provide photographers with even greater control over their image editing and management. In this blog post, we’ll explore the key highlights of darktable 4.4.0 and how they enhance the user experience.

Enhanced User Interface:

Darktable 4.4.0 introduces a revamped user interface (UI) that improves the overall workflow and usability. The new interface design offers a cleaner and more intuitive experience, making it easier for both new and experienced users to navigate and access the various editing tools.

Advanced Editing Tools:

This release brings several new editing tools and improvements to existing ones, allowing photographers to achieve greater precision and creativity in their post-processing. Some notable additions include:

1. Advanced Tone Equalizer: The Tone Equalizer module has been significantly enhanced, enabling photographers to make fine adjustments to brightness, contrast, and color in specific tonal ranges. This tool provides greater control over the image’s overall tonality and can enhance details and textures.
2. Compressed RAW Support: Darktable 4.4.0 now supports compressed RAW formats, enabling users to work with more camera models and increasing compatibility across different devices.
3. Improved Denoise and Demosaic Algorithms: The denoise and demosaic modules have been refined, resulting in better noise reduction and demosaicing capabilities. Photographers can now achieve cleaner and sharper images, particularly when dealing with high ISO or low-light conditions.
4. Advanced Color Calibration: Darktable now includes enhanced color calibration options, allowing users to achieve more accurate and consistent color representation across various devices and displays.

Workflow Improvements:

Darktable 4.4.0 introduces several workflow enhancements aimed at streamlining the image editing process. Some notable improvements include:

1. Timeline View: A new timeline view has been added, enabling photographers to have a visual overview of their editing history. This feature makes it easier to navigate through different editing steps and revert to previous versions if needed.
2. Customizable Shortcuts: Users can now customize keyboard shortcuts to suit their preferences and workflow, further increasing efficiency and productivity.
3. Geotagging Enhancements: Geotagging functionality has been improved, allowing for easier and more accurate location tagging of images.

Performance and Stability:

In addition to new features, darktable 4.4.0 brings numerous performance improvements and bug fixes, resulting in a more stable and reliable software experience. These enhancements ensure smoother operation and optimized performance, even when working with large image libraries.

Conclusion:

Darktable 4.4.0 represents a significant milestone in the evolution of this popular open-source photography workflow software. With its enhanced editing tools, streamlined workflow, and improved user interface, darktable empowers photographers to unleash their creative potential and achieve stunning results. Whether you are a professional photographer or an enthusiast, this latest release is definitely worth exploring. Download darktable 4.4.0 and elevate your photography post-processing to new heights.

My erstwhile photography blog is back after a forced hiatus. Now that I have retired from my day job, I expect to keep this more up to date. Please visit often and read whatever blatherings I choose to post here.