Incorrect TTML Conversions

[simontWork]

The linked TTML conversions were designed for subtitles - i.e. to maintain enough separation between a small number of colours to be visibly different.

The equations there are not colour accurate.

[palemieux]

Can you suggest an alternatives?

[simontWork]

Would prefer to put a code snippet in that implements the ITU conversions. It simplifies as the OOTF adjustment just happens to be unity at 80 cd/m2.

[simontWork]

Further on the TTML conversion, the important design limitation was to be able to undertake the transform in the graphics plane of a television. I don't think either of the TTML conversions are colour accurate.

[simontWork]

Could we use the following from a previous commit

Conversion from sRGB to HLG

Process:

1. Linearize using the sRGB EOTF
2. Convert from srgb color space to rec2100-hlg color space
3. Scale pixel values - See Note 1
4. Apply HLG Inverse EOTF to convert to HLG from Pseudo-Display Light with Lw = 302 cd/m2 - See Note 1
   • apply HLG Inverse OOTF
   • apply HLG OETF - See Note 2

Note 1: As rec2100-hlg is a relative format, the brightness of the virtual monitor used for mathematical transforms can be chosen to be any level. In this transform it is chosen to be 302 cd/m2 so that the brightness of diffuse white matches the diffuse white of sRGB (80 cd/m2). Using the extended range gamma formula in footnote 2 of BT.2100, this also sets the HLG Inverse OOTF to be unity. The value 0.265 is calculated by taking the inverse OETF of 0.75, the rec2100-hlg diffuse white level.

Note 2: See section 5.3 in ITU-R BT.2408-4 relating to negative transfer functions in format conversions.

function convertSRGBtoREC2100HLG(r, g, b) {
  const systemGamma = 1.0;
  const linearLightScaler = 0.26496256042100724;

  const r1 = srgb_eotf(r);
  const g1 = srgb_eotf(g);
  const b1 = srgb_eotf(b);

  const [r2, g2, b2] = matrixXYZtoBT2020(matrixSRGBtoXYZ(r1,g1,b1));

  const r3 = linearLightScaler * r2;
  const g3 = linearLightScaler * g2;
  const b3 = linearLightScaler * b2;

  const [r4, g4, b4] = hlg_inverse_ootf(r3, g3, b3, systemGamma);  // Can be removed, included for completeness
  const [r5, g5, b5] = hlg_oetf(r4, g4, b4);

  return [r5, g5, b5]
}

[palemieux]

@simontWork

See #107. Step (4) did not have matching lines in the pseudocode and it is not clear the relevance of Lw.