00:50 <yuni> lina: I'm still upset about the golem thing, but Phoronix gave you credit for your acheivement :)

00:50 <alyssa> https://www.phoronix.com/scan.php?page=news_item&px=Asahi-Linux-First-Triangle

00:51 <yuni> alyssa: Wow, swooping in to paste a link instead of letting people do one Google search ...

00:51 <alyssa> blurgh

00:54 <Tramtrist> Just saw that about 10 minutes ago. Congrats lina 😻

01:08 <alyssa> blu

03:48 <rkjnsn> Random thought: I think I recall seeing something a couple of days ago about a blob that turned out to be YUV <-> RGB conversion matrices, but there was still part that was unidentified. Since the Macs support Dolby Vision playback, could the last part be for converting from the ICtCp or IPTPQc2 color space?

08:16 <lina> yuni: I saw that!! ^^

08:17 <lina> rkjnsn: Do you have the coefficients for that? Check #6 here: https://gist.github.com/asahilina/2a80fe5e4476de56412508ec78245b9b

08:22 <lina> It seems to use limited range luma but with a different offset, since you get negative RGB at Y=16/256

08:25 <chadmed> if its ICtCp then the valid inputs are [0,1] for I and [-0.5,0.5] for Ct and Cp

08:25 <lina> Doesn't sound like that then...

08:25 <chadmed> yeah

08:25 <rkjnsn> I'm afraid I don't know too much about it. I just remembered reading that Macs support Dolby Vision and (separately) that Profile 5 Dolby Vision uses IPTPQc2. The Wikipedia page (https://en.wikipedia.org/wiki/ICtCp) has some matrices.

08:26 <chadmed> rkjnsn keep in mind that those matrices are transformed from the LMS colour space, not RGB

08:26 <chadmed> so the coefficients would be wrong until you apply the HLG transfer function and whatever special sauce dolby use

08:27 <chadmed> but ICtCp fits BT2020 into -1 to 1 for luma and -0.5 to 0.5 for chroma information so it cant be those anyway

10:38 <jannau> is there value in trying agx_1tri.py on t6000? has anyone already tried the mesa driver on macOS t6000?

11:12 <lina> I think someone did and it worked with some issues, so some numbers probably need tweaking...

11:12 <lina> 1tri won't work since it hardcodes t8103-specific data in lots of places right now. Lots of power management / hardware related constants change for each SoC.