Wayland (the protocol and architecture) is still lacking exact consideration for coloration management. Wayland also lacks improve for excessive dynamic vary (HDR) imagery which has been spherical in movie and broadcasting industry for a while now (e.g. Netflix HDR UI).
While there are successfully established instruments and workflows for the skill to whole coloration management on X11, even X11 has no longer gained improve for HDR. There were plans for it (Alex Goins, DeepColor Visuals), nonetheless so some distance as I do know nothing really materialized from them. True now, the suited skill to survey HDR narrate on a HDR monitor in Linux is to utilize the DRM KMS API straight away, in totally different words, no longer utilize any window machine, that means no longer using any desktop atmosphere. Kodi is one in every of the one just a few applications that may per chance well perhaps stop this in any appreciate.
That is a yarn about starting the efforts to fix the direct of affairs on Wayland.
Coloration management for Wayland has been talked about on and off for a massive selection of years by dozens of people. To me it changed into glaring from the launch that coloration management architecture on Wayland have to be mainly totally different from X11. I believed the level to server have to be part of the coloration management stack as an replace of an untrusted, unknown entity that have to be bypassed and overridden by applications that fight each totally different for who will get to configure the level to. This notion changed into wildly controversial and it took a lengthy time to rep my level all the blueprint by, nonetheless over the years some coloration management consultants began to expose to new suggestions and totally different other folks joined in the notion as successfully. Whether these new suggestions are really better than the techniques of old remains to be viewed, though. I feel the promise of getting everything and more to work better is some distance too immense to no longer are attempting it out.
The discussions started numerous times over the years, nonetheless they consistently dried out mostly without any tangible progress. Coloration management is a wide, deep and complex subject, and the desired talents, info, pastime, and available time did no longer attain collectively unless rather no longer too lengthy ago. Individuals did write draft protocol extensions, nonetheless I’d claim that it changed into no longer really unless Sebastian Wick started constructing on top of them that things started intriguing ahead. However one particular person can’t push such a wide effort on my own even for the easy motive that there have to be at least one reviewer sooner than the relaxation will even be merged upstream. I changed into very lucky that since summer 2020 I were ready to work on Wayland coloration management and HDR for bettering ChromeOS, letting me improve Sebastian’s efforts on a on a regular foundation foundation. Vitaly Prosyak joined the state this year as successfully, researching the blueprint to combine the two apparently totally different worlds of ICC and HDR, and how tone-mapping may per chance be applied.
I have to also stamp the previous efforts of Harish Krupo, who submitted a predominant Weston merge count on, nonetheless unfortunately on the time reviewers in Weston upstream were no longer out of the ordinary available. Even sooner than that, there were experiments by Ville Syrjälä. All these are really mostly old-usual by the on-going work.
Presently the full of life other folks all the blueprint by the subject are me (Collabora), Vitaly Prosyak (AMD), and Naveen Kumar (Intel). Sebastian Wick (unaffilated) is still spherical as successfully. None of us is a coloration management or HDR educated by trade, so we are all studying things as we poke.
The inspiration for the coloration management protocol are ICC profile files for describing each output and narrate coloration spaces. The aim is for ICCv4, also allowing ICCv2, as these are identified and supported successfully mainly. Adding iccMAX improve or the relaxation else may per chance be seemingly any time in some unspecified time in the future.
As coloration management is all about coloration spaces and gamuts, and excessive dynamic vary (HDR) is also very out of the ordinary about coloration spaces and gamuts plus prolonged luminance vary, Sebastian and I made a decision that Wayland coloration management extension have to cater for every from the starting. Combining former coloration management and HDR is a rather new thing so some distance as I do know, and I’m no longer determined we possess out of the ordinary prior art to tainted upon, so that is an enticing research lunge as successfully. There is rather just a few prior art on HDR and coloration management individually, nonetheless they possess an inclination to possess traditional variations that makes the mixture no longer glaring.
To abet us keep focused and level to to the neighborhood about what we really intend with Wayland coloration management and HDR improve, I wrote the fraction « Wayland Coloration Management and HDR Kind Needs » in coloration.rst (draft). I very out of the ordinary counsel you to read it so that you rep a image what we (or I, at least) want to aim for.
Elle Stone explains in their article how coloration management have to work on X11. As I predominant to steer certain of repeating the wide electronic mail threads that were had on the wayland-devel mailing list, I wrote the fraction « Coloration Pipeline Overview » in coloration.rst (draft) more or much less as a response to her article, making an strive to level to in what techniques Wayland may per chance be totally different from X11. I feel that knowing that fraction is paramount sooner than any individual makes any statement on our efforts with the Wayland protocol extension.
HDR brings out of the ordinary more causes to assign apart coloration home conversions in the level to server than perfect the thought that every applications have to be coloration managed if no longer explicitly then implicitly. Most of the desktop applications (successfully, literally all perfect now) are using Similar old Dynamic Vary (SDR). SDR is a fuzzy thought referring to all former, non-HDR image narrate. Attributable to this fact, your desktop is mostly 100% SDR. You lag your love new HDR monitor in SDR mode, that means it appears to be like to be like perfect love any old monitor with nothing love. What whilst it is advisable to survey a HDR video? The monitor gained’t level to HDR in SDR mode. In the occasion you merely swap the monitor to HDR mode, you are going to be blinded by the total over-intellectual SDR applications. Switching monitor modes may per chance well perhaps furthermore reason flicker and take away a runt of time. That may per chance well perhaps be a rather harmful user experience, perfect?
A answer is to lag your monitor in HDR mode the total time, and possess the window machine compositor convert all SDR utility windows properly to the HDR luminance, so that they notice well-liked in spite of the HDR mode. There will consistently be applications that will never improve HDR in any appreciate, so the compositor doing the conversion is form of the suited skill.
For the protocol, we are presently exploring the utilize of relative luminance. The motive is that folk notice at shows in wildly varying viewing environments, below well-liked assign apart of job lights for instance. The atmosphere and non-public preferences impact what monitor brightness you desire. Additionally shows themselves will even be wildly totally different in their capabilities. Most prior art on HDR makes utilize of absolute luminance, nonetheless absolute luminance has the state that it assumes a particular viewing atmosphere, in most cases a unfortunate room, such as a movie theatre. If a level to server would convey a movie with absolutely the luminance it changed into mastered for, in most circumstances it would be some distance too dusky to survey. Whether using relative luminance on the protocol stage appears to be like to be an very perfect recommendation or no longer, we are able to analyze.
The Wayland coloration management and HDR protocol extension proposal is identified as wayland/wayland-protocols!14 (MR14). Since it’s some distance a extremely lengthy operating merge count on (the bar for touchdown a new protocol into wayland-protocols is excessive) and there are numerous other folks working on it, we started using sub-merge-requests to regulate the proposal. You may per chance well perhaps procure the sub-MRs in Sebastian’s fork. In the occasion you possess a trade to point out, that is the blueprint to whole it.
Obviously using sub-MRs also splits the review discussions into numerous locations, nonetheless on this case I feel it’s some distance an very perfect thing, since the discussion threads in Gitlab are already massive.
There are a whole lot of spacious and diminutive start questions we haven’t had the time to sort out yet even among the many full of life neighborhood; questions that I feel we must possess some tentative answers sooner than soliciting for wider neighborhood feedback. There is also no area time desk, so don’t exhaust your breath. This work is at possibility of do away with months still sooner than there may per chance be a full tentative protocol, and per chance years unless these aspects will seemingly be found on your well-liked Wayland desktop environments.
In the occasion that you can perhaps very successfully be an educated on the topics of coloration management or HDR shows and narrate, or a developer drawn to contributing to the undertaking, that you can perhaps very successfully be warmly welcome to affix the advance.