[spaces/rec2020] Use gamma 2.40 for display-referred rec2020 #669

[svgeesus]

rename previous implementation as scene-referred --rec2020-oetf so people can still use the older implementation if they want

See

• Define rec2020 color space to use 2.4 gamma w3c/csswg-drafts#12574 (comment)

[netlify]

✅ Deploy Preview for colorjs ready!

Name	Link
🔨 Latest commit	7254de4
🔍 Latest deploy log	https://app.netlify.com/projects/colorjs/deploys/68b86b8951eb2500085a4170
😎 Deploy Preview	https://deploy-preview-672--colorjs.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify project configuration.

[svgeesus]

Build gives a bunch of type linting errors completely unrelated to the files I changed.
Adding fixes to those should not be done in this PR, and should be done by those who wrote them and understand the types involved.

[facelessuser]

As far as implementation, everything looks fine. But related Rec. 2020 tests would need to be updated as some are failing:

• Rec. 2020 <-> Linear Rec. 2020
• APCA contrast, sRGB gamut in other color spaces

[facelessuser]

As a side note, I don't know when it happened, but it seems gamut mapping tests are not working: https://colorjs.io/test/

I'm thinking there may be some infinite recursion or something as it seems we blow the call stack:

Uncaught RangeError: Maximum call stack size exceeded
    at custom (util.js:75:11)
    at callee (objects.js:225:15)
    at Module.stringify (objects.js:212:9)
    at stringify (util.js:151:17)
    at output (util.js:74:9)
    at Array.map (<anonymous>)
    at custom (util.js:77:16)
    at callee (objects.js:225:15)
    at Module.stringify (objects.js:212:9)
    at stringify (util.js:151:17)

[facelessuser]

I don't see anything obviously wrong with gamut mapping or stringifying the output, so I suspect gamut mapping tests (and a few others) are broken due to calling some special htest check functions. This isn't an area I'm familiar with, but I guess we should open a separate issue to figure out what's going wrong there.

[svgeesus]

But related Rec. 2020 tests would need to be updated as some are failing:

Agreed, I will fix those

[svgeesus]

One reason so few rec2020 tests were failing is that an astonishing number use colors with either 0 or 1 coordinates which are insensitive to the actual transfer function.

[facelessuser]

LGTM