Talk:Color appearance model

IPT non-constant lines of hue

The IPT section mentions "the issue of non-constant lines of hue" but no where up the page are lines of hue explained, or even mentioned. Pnorman (talk) 01:25, 22 March 2018 (UTC)

What I think the editor meant is that some of the models have the property that changing the physical brightness of a stimulus might result in a hue change, rather than just a lightness change.

The developers of IPT argue that theoretically, according to them anyway, that shouldn't happen. In the IPT model, if you multiply the XYZ vector by a constant K, this first commutes with the HPE matrix, then can be pulled out of the power function and finally commutes with the IPT matrix, resulting in a final constant scaling factor of K·⁴³ in the IPT vector. So within their theoretical framework it preserves hue.

The thing is, there are good reasons to think the physical brightness does affect perceived hue. So I doubt the validity of the raison d'être of IPT and it contributes nothing else to the field. In particular, it doesn't even attempt to model human colour vision. I'd argue that even the most recent models are only a first step, but still it's something and that cannot be said for IPT. — Preceding unsigned comment added by 92.67.227.181 (talk) 16:19, 2 July 2022 (UTC)

CAM20u

Any info on CAM20u? — Preceding unsigned comment added by 77.61.180.106 (talk) 23:32, 7 March 2021 (UTC)

CIECAM16

So the news kind of passed me by, but you guys apparently didn't notice it either: CAM16 is an official CIE standard now: CIE 248:2022. — Preceding unsigned comment added by 92.67.227.181 (talk) 00:57, 4 June 2022 (UTC)

Also, ‘various fixes and improvements’ is code for ‘I haven't actually read it’. And I know you haven't read it, because if you had this would have jumped out at you like a flashing neon sign: it uses one colour mapping matrix rather than two, eliminating not just an enormous numerical problem with the previous models, but also the most important outstanding point of psychophysical implausibility of the same. — Preceding unsigned comment added by 92.67.227.181 (talk) 11:06, 6 June 2022 (UTC)

Color appearance

"In 1931, using psychophysical measurements, the International Commission on Illumination (CIE) created the XYZ color space, which successfully models human color vision on this basic sensory level." This is very misleading: The LMS color space attempts to model the human sensory system. The 1931 XYZ color space models how humans perceive colors, while the cone functions were still unknown. 109.199.166.7 (talk) 13:17, 5 September 2022 (UTC)

Requested move 10 April 2023

Color appearance model → Color appearance – "Models" is just a smaller section of the article. fgnievinski (talk) 15:11, 10 April 2023 (UTC)