Todays minor update introduces a new export format option for Pro users, important bug fixes and a lot of internal work that paves the road for the next major version update!
A previous update introduced tiled rendering to support very high resolution exports that do not overwhelm the WebGL rendering core. There was a rounding error in the tile calculation that partitions the image rectangle into the sub-grid for rendering that caused a half pixel shift for exported images, resulting in a slightly blurred image. This has now been fixed and full resolution exports are pixel-perfectly-crisp now.
Previously, LUTs exported for an ACES working space could contain NaN (Not a Number) errors in the lattice, resulting in "holes" in the LUT file. This was due to the inability to index out-of-range values which can occur in ACES. This has been fixed by extending the index range for the internal output transformers and NaN errors should be a thing of the past.
This update introduces Tiff as a new export format available for Pro Users. There's not much to say about this other than it's a standard tiff implementation that supports an alpha channel. There's generally no quality gain over lossless PNG but TIFF might come in handy for some workflows so choose it when you need it.
Color.io allows you to design looks that can be integrated into third party ACES color managed workflows. I added documentation on how to export LUTs for an ACES working space in Davinci Resolve with some pointers on how to setup your node structure and ACES transforms: https://www.color.io/user-guide/using-color-io-with-davinci-resolve
A lot of things have changed under the hood as I'm preparing the first major update since Color.io soft-launched one month ago. Thanks to the amazing feedback I got from users via email, this feedback board and via lift-gamma-gain.com, a clear vision of where the app should be headed next, has been established:
Right now, all adjustments are scene specific, meaning that the same color grade is applied to every image in a scene. While this is very useful, it would be even better if some adjustments could be made on a per-image basis. This requires some architectural changes of the processing core but it's the right direction to take. Things like grain and halation as well as basic exposure and saturation corrections make more sense to be applied as a per-image pre-processing step before color grading. This will also open the door to masking and additional image correction tools for photographers.
The next major update will introduce advanced film acutance and grain modeling emulation that I'm super excited about. What makes this new model incredibly powerful is that it breaks the image apart and re-builds it, pixel-by-pixel, from physically accurate grain patterns with varying density. This approach completely changes the pixel structure of digital photos, introducing a subtle but powerful meta layer that further blurs the lines between digital and film. This module will replace the current grain implemention and will be part of the image specific adjustments. I'm currently working on optimizing the real time rendering of the acutance distribution.
Some of you probably remember the subtractive CMY channel mixer I had in an early alpha release of the app. I sacrificed it for adding the grain and halation modules into the main interface for the public release but since grain and halation will be moved into the image editing tool section of the next version, you can expect a return of the cmy channel mixer into the luminance panel. It is a great tool to "finish" a look. I'm thinking about splitting it up into shadows and highlights, similar to the Refraction emulator. Let me know your ideas!
As always, thank you so much for using the app and for providing me with the feedback and bug reports that make all of this possible. I launched Color.io with the sincere desire to improve and extend features based on user feedback to create the most useful tool for my own work and others. I could't be more grateful for the response the app has received so far and looking forward to where we're going next!
Jonathan | Color.io
