Do you think it might be useful to get in contact with the ISO committee? For all practical purposes, the standard is "wrong" and should be updated.
I think this is the committee in question: http://www.iso.org/iso/home/standards_development/list_of_iso_technical_committees/iso_technical_committee.htm?commid=45020 cheers, -Mark On Wed, Jun 1, 2016 at 7:02 PM, Larry Gritz <[email protected]> wrote: > My guess is Photoshop was the first tool to consider this metadata and got > it wrong, and everyone else is copying Photoshop. > > > Yeah, that seems very likely to me. > > And here we are, about to take the only identified instance of software > that got it right, and break it in order to be compatible with all the ones > who got it wrong. :-( (I see no practical alternative to this plan.) > > > On Jun 1, 2016, at 1:34 PM, Jonathan Gibbs <[email protected]> wrote: > > This is just a totally fascinating study of modern software. It's > fascinating that you have a spec as widely adopted as JFIF/JPEG, and > clearly the original engineers did a good job specifying some important > meta-data. I'm sure they had all sorts of use cases where this was > important. > > And yet (a) very little software handles it at all, and (b) those who do > get it wrong. > > When does Software Archeologist become a well-paid job? That would be an > interesting thing to be called out of retirement to do some day. :) > > My guess is Photoshop was the first tool to consider this metadata and got > it wrong, and everyone else is copying Photoshop. > > FWIW, Houdini's "mplay" image/flipbook viewer seems to ignore it too. > > --jono > > > On Fri, May 20, 2016 at 1:36 AM Kevin Wheatley <[email protected]> > wrote: > >> now I'm at work the best documentation I have is from ITU T.871 which >> is ISO/IEC 10918-5: >> >> Units for the H (horizontal) and V (vertical) densities: >> = X'00': units unspecified; H and V densities, expressed in dots per >> arbitrary unit, specify only the pixel aspect ratio (width:height >> pixel aspect ratio = Vdensity:Hdensity). >> >> which is more explicit than the JFIF 1.02 de facto standard, which was >> what I read last night. JFIF 1.02 I thought could be read in an >> ambiguous manner. >> >> This is much better, but reinforces something I mentioned to some >> people yesterday; specifications should come with some form of >> examples/reference implementation, especially when tying together >> multiple standards or when there are variations supported due to >> combining multiple proposals. >> >> Kevin >> _______________________________________________ >> Oiio-dev mailing list >> [email protected] >> http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org >> > _______________________________________________ > Oiio-dev mailing list > [email protected] > http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org > > > -- > Larry Gritz > [email protected] > > > > _______________________________________________ > Oiio-dev mailing list > [email protected] > http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org > >
_______________________________________________ Oiio-dev mailing list [email protected] http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org
