[argyllcms] Re: help with camera profile
- From: Iliah Borg <ib@xxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Sun, 29 Jul 2012 18:57:40 -0400
On Jul 29, 2012, at 5:45 PM, Ben Goren wrote:
> On 2012-07-29, at 12:55 PM, Iliah Borg wrote:
>
>>> If I try to do the same with the 7D profile, the error gets higher (27
>>> as peak) and even after removing 3 patches it stays to 17. The "bad"
>>> patches are always orange/yellow.
>>>
>> it might depend on the output colour space and how they get their. I mean,
>> have you checked for clipping?
>
> For example, sRGB is only big enough to encode 96% of the ColorChecker gamut,
> and only 83% of the Digital ColorChecker gamut.
There is another flip to it. Some sensors and some profiles exceed Lab, and
deltaE in Lab may be not very informative. Even mapping to Lab needs very close
attention. We do not have perceptual here, so we may want to aim for
preservation "memory colours", sacrificing colour in shadows; or we may use
some luminosity mapping to avoid negative numbers - but in that case we are in
a non-linear space; we may try to skip Lab altogether; or we just cut the
dynamic range in conversion. An old way of making RGB-RGB transforms is also
sometimes useful.
--
Iliah Borg
ib@xxxxxxxxxxx
Other related posts:
