[argyllcms] Re: Poor Man's Colorimeter
- From: Zi Wang <ziwang84@xxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Sat, 13 Jun 2015 00:19:03 +0800
Hi,
Thank you guys for the answer!
On Jun 12, 2015, at 12:49 AM, Richard Kirk <richard@xxxxxxxxxxxxxxxx> wrote:
I have done a lot of this sort of thing. It can work if you restrict yourself
to one illuminant and one medium. So, if you can produce one set of
decent-sized patches which you can measure with the spectrometer, then you
can get a set of scanner RGB and corresponding spectrometer XYZ which you can
then use to interpolate. However, the interpolation will only convert a
particular set of CMY inks with a particular halftoning on a particular paper
to the XYZ for the same medium measured under the particular illuminant (or
any illuminant if you measure your illuminant, and it is broad-band enough to
correct).
On Jun 12, 2015, at 7:33 AM, Graeme Gill <graeme@xxxxxxxxxxxxx> wrote:
That's not going to work very well, because it is composed of different
colorants with different spectral characteristics. If the scanner was
colorimetric,
it wouldn't matter, but things like scanners and cameras aren't close to
human vision in terms of their spectral sensitivities.
First of all just to make sure I actually understand what you two are saying.
Let’s say if my samples are CMY and all of them are produced with exactly the
same batch of inks that I can get hold of. Then cooking up a good set of
patches from those inks will actually give good result. What I’m trying to
figure out is whether this works because CMY are direct complements to RGB or
because CMY are three primitives along with the additive nature of scanners and
cameras things will somehow survive the linear interpolation?
You also specifically mentioned CMY instead of CMYK, so is black going to cause
problem? I don’t know much about inks but adding even a tiny bit of carbon
black to some bright organic pigments really distorts the spectral graph. Is
this the same kind of business?
I get the impression that you are saying because a scanner never actually
measure the spectral reflectance curve therefore there’s not much we can do.
I’m wondering whether it’s too difficult or theoretically impossible to guess
the reflectance curve from sRGB data? Or at least if reflectance curve is
definitely not possible can we do any better than linear interpolation? I’m
thinking if we can work out the transparency of each of the RGB filters on the
CCD and the sensitivity of the CCD itself then we probably could guess the (or
in fact some) reflectance curve(s) by numerical method?
On Jun 12, 2015, at 7:33 AM, Graeme Gill <graeme@xxxxxxxxxxxxx> wrote:
That's a difficult path. If your samples are the product of a 3 color process
(i.e.
an CMY printer or a photographic process), then it may be possible to get
passable
results, but accuracy will get poorer, the more colorants that are involved
(i.e.
CMYK, CMYK + others).
The situation is I can live with one illuminant as long as it’s not too exotic,
say I will be happy with either D50, D65 or C. But restricting myself to one
set of CMY inks isn’t really possible (paper shouldn’t be a problem because I’m
actually dealing with opaque pigments). Realistically, even with the Wolf Faust
target the result is not completely useless. They generally turn out to be too
green or sometimes too blue as well. With educated guesses on how much to
compensate for the a* and b* value I still manage to hammer my way through the
whole thing. So if this is a really difficult path then I might just try to get
the best out of scanner calibration then leave the rest of the work to manual
labour. In this case what can I do to improve accuracy? In particular I’m
assuming a naive point of view that adding more patches to the target won’t
hurt as long as they are accurate, is this true or not?
Rough rule of thumb for an RGB or CMY device is that you need about 1000
sample
points for a good profile, and more if you want a high quality profile. This
is not easy to achieve for input devices.
1000 or even more sample points is not a problem for me as long as it’s worth
the effort. In fact it’s much better to know that I need a huge number of
sample points so I don’t spend time fiddling with an IT8 target that’s never
going to work. Does it make sense to create lots of patches with say 10 or 20
different pigments or something like that?
Cheers,
Zi
Other related posts:
