RE: Sanity check, ITU and IEC definitions of sRG are identical, right?

Note that the ITU document is a Report (informative) and not a Recommendation (which would  be a normative specification).

Craig Todd
Dolby Fellow
Dolby Laboratories
1275 Market St.
San Francisco, CA  94103
T +1 415-558-0221  M +1 415 672-0221
ct@dolby.com<mailto:ct@dolby.com> | http://www.dolby.com


From: Chris Lilley <chris@w3.org>
Sent: Wednesday, May 1, 2019 8:18 AM
To: public-colorweb@w3.org
Subject: Sanity check, ITU and IEC definitions of sRG are identical, right?


Learned people,

Currently, all W3C specifications (bar one, which is being fixed [1]) refer to the IEC definition of sRGB as authoritative. [2]

Multimedia systems and equipment - Colour measurement and management - Part 2-1: Colour management - Default RGB colour space - sRGB<https://urldefense.proofpoint.com/v2/url?u=https-3A__webstore.iec.ch_publication_6169&d=DwMDaQ&c=lI8Zb6TzM3d1tX4iEu7bpg&r=_F90w7eDD4MBjtHx1p0KJg&m=86nhgkWcBMyAZtdf1CJ-FZR_ODNPVLcBs7NNCK2zIV8&s=IYyuGYlT2vemcxSBHQWFTfhyHDZes9JGpETCZEBJ4sc&e=>.

However, the link goes to the IEC web store[3], where one can pay real money for a copy. In consequence, many people don't have a copy (including myself, hence this mail).

I recently learned [4] (and sorry, I didn't know this before) that sRGB is also defined by the ITU, on p.14 of:

Television colorimetry elements. Report  ITU-R  BT.2380-2 (10/2018)
https://www.itu.int/dms_pub/itu-r/opb/rep/R-REP-BT.2380-2-2018-PDF-E.pdf<https://urldefense.proofpoint.com/v2/url?u=https-3A__www.itu.int_dms-5Fpub_itu-2Dr_opb_rep_R-2DREP-2DBT.2380-2D2-2D2018-2DPDF-2DE.pdf&d=DwMDaQ&c=lI8Zb6TzM3d1tX4iEu7bpg&r=_F90w7eDD4MBjtHx1p0KJg&m=86nhgkWcBMyAZtdf1CJ-FZR_ODNPVLcBs7NNCK2zIV8&s=Zxm1z8sjWiUl4oGKRc3fp_jLMVvPhs5DL-XT8Y2ctIM&e=>

and this specification has the merit of being freely available to download.

I see that the ITU report does reference IEC 61966-2-1 so I presume that the chromaticity coordinates, EOTF and OETF are copied correctly from there.

I notice that the EOTF breakpoint is specified to more significant figures, 0.40449936 rather than 0.04045 which I have seen elsewhere.

I would have greater peace of mind, before submitting a pull request on the spec database to point to the freely available specification, if someone who has read both specifications could confirm for me that the definitions are identical?

[1] Web Content Accessibility Guidelines refers to an obsolete 4th working draft of the IEC specification, which still has the erroneous rounded-off coefficient that led to a discontinuity in the transfer function.
Although the impact on 8bit per component systems is invisible, I now have agreement to fix that
https://github.com/w3c/wcag/issues/360#issuecomment-453741912<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_w3c_wcag_issues_360-23issuecomment-2D453741912&d=DwMDaQ&c=lI8Zb6TzM3d1tX4iEu7bpg&r=_F90w7eDD4MBjtHx1p0KJg&m=86nhgkWcBMyAZtdf1CJ-FZR_ODNPVLcBs7NNCK2zIV8&s=0PPYZJiOTCA6FD5IEiZiOANkjJ4AbHbms3XXNv-wdA4&e=> and following
[2] https://www.specref.org/?q=srgb<https://urldefense.proofpoint.com/v2/url?u=https-3A__www.specref.org_-3Fq-3Dsrgb&d=DwMDaQ&c=lI8Zb6TzM3d1tX4iEu7bpg&r=_F90w7eDD4MBjtHx1p0KJg&m=86nhgkWcBMyAZtdf1CJ-FZR_ODNPVLcBs7NNCK2zIV8&s=blQS3hXydiwA2j9rzuJF4CKBHTAafd9RDIzg-2PhHq4&e=>
[3]  https://webstore.iec.ch/publication/6169<https://urldefense.proofpoint.com/v2/url?u=https-3A__webstore.iec.ch_publication_6169&d=DwMDaQ&c=lI8Zb6TzM3d1tX4iEu7bpg&r=_F90w7eDD4MBjtHx1p0KJg&m=86nhgkWcBMyAZtdf1CJ-FZR_ODNPVLcBs7NNCK2zIV8&s=IYyuGYlT2vemcxSBHQWFTfhyHDZes9JGpETCZEBJ4sc&e=>
[4] https://github.com/w3c/wcag/issues/360#issuecomment-487476773<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_w3c_wcag_issues_360-23issuecomment-2D487476773&d=DwMDaQ&c=lI8Zb6TzM3d1tX4iEu7bpg&r=_F90w7eDD4MBjtHx1p0KJg&m=86nhgkWcBMyAZtdf1CJ-FZR_ODNPVLcBs7NNCK2zIV8&s=k3PIiA0oDHLQLNvFImi8mxZ2rNokitQf97DxGIQcgz4&e=>

--

Chris Lilley

@svgeesus

Technical Director @ W3C

W3C Strategy Team, Core Web Design

W3C Architecture & Technology Team, Core Web & Media