Re: Unwanted restriction for JPEG images that have colour profiles.

Hi Chris,

Well spotted! I'm happy with that wording.

Kind Regards,
Anthony

Chris Lilley wrote:
> Hello ,
> 
> I just noticed a problem in Appendix H.
> http://dev.w3.org/SVG/profiles/1.2T/publish/jpeg.html
> 
> 
>   SVG Tiny 1.2 UA's should convert Y,Cb,Cr values compressed in the
>   JPEG image to RGB as defined in the JFIF specification [JFIF] and
>   may assume that the RGB values are sRGB.
> 
> While that usefully covers what colour space is meant by RGB, in the
> absence of other information (since the JFIF specification is silent
> on the matter) it should obviously not take precedence when
> information *is* supplied (ie an ICC colour profile in the JPEG
> image). The current wording seems to forbid this, which is not wanted.
> 
> For Tiny, colour management is optional but profiles that build on
> Tiny may require it. I suggest therefore this modified wording; it
> adds no new conformance criteria but makes the intention clearer:
> 
>   SVG Tiny 1.2 UA's should convert Y,Cb,Cr values compressed in the
>   JPEG image to RGB as defined in the JFIF specification [JFIF] and
>   may assume, in the absence of a colour profile, that the RGB values
>   are sRGB.
> 

Received on Friday, 18 July 2008 04:02:12 UTC