Hello again,
I'm answering my own question again, after discovering that it is the "-s"
flag compression factor that is causing the discontinuities. Quite strange
since I've never found this to be a problem with inkjet RGB printer
profiles for matte media, which have relatively small gamuts too.
With "-s 1", I see that the discontinuities and bumps in the gradient of
dark colours, especially more saturated dark colours, are very significant.
Going to 20 (20%) reduces the discontinuities by quite a bit. Taking it up
to 100 reduces it the most, although it's still present. I tried converting
several of my pictures with important shadow detail using "-s 20" and "-s
30" variants, and I could still see a gross tone reversal in the dark
colours. Only "-s 100" approaches an acceptable smoothness, but then a lot
of the in-gamut colours lose way too much saturation, obviously since this
setting controls the gamut compression.
Here are the profiles, to save time re-computing them. They do take a long
time to build.
https://www.dropbox.com/sh/euert20x9f0oo8y/AABIWSwOj85bFguL4Wo5DwBka?dl=0
The file names detail the commands that were used to build them, for
example:
KHLIndigo10000-Nevia-limit400-CalGlossyDefaults-31Aug18-SpectralM0-averaged-kh-s1-ca25-cg15
(the third hyphen from the end indicates the "-s" flag setting which was
used. There are four profiles in the folder,for "-s 1", "-s 20", "-s 30"
and "-s 100".
Graeme, hope you could look into this. I wonder if it has got to do with
the fact that this is a CMYK printer profile, rather than an RGB one.
Thanks!
Regards,
Samuel
On Wed, 5 Sep 2018 at 07:43, Samuel Chia <inksandpaper@xxxxxxxxx> wrote:
Hi everyone,
Would appreciate any help in trying to figure out why the Argyll-built
profile is showing discontinuities in the shadows! Thank you.
Regards,
Samuel
On Mon, 3 Sep 2018 at 09:57, Samuel Chia <inksandpaper@xxxxxxxxx> wrote:
Hi,
I have a separate issue so I thought it might be less confusing to put it
out in a separate thread.
Because I'm using an iSis and having trouble using ArgyllCMS to generate
patch sets for CMYK profiling (an HP Indigo press) for an iSis since it's
not supported, I tried a 3000 patch set which was created in i1Profiler,
measured the patches, and asked ArgyllCMS to build a profile after
converting the spectral measurements text file to ti3.
This is the argument string I used for colprof:
colprof -v -qh -l400 -tlp -s 20 -cmt -ca:27 -cg:7 -dpp
KHLIndigo10000-Nevia-limit400-CalGlossyDefaults-31Aug18-SpectralM0-averaged-kh-s20-ca27-cg4
Here is the ti3 file which was used:
https://www.dropbox.com/s/txarthlw4ksvfgp/KHLIndigo10000-Nevia-limit400-CalGlossyDefaults-31Aug18-SpectralM0-averaged-kh-s20-ca27-cg7.ti3?dl=0
Here is the profile that was created:
https://www.dropbox.com/s/pyipen1bnud5y6w/KHLIndigo10000-Nevia-limit400-CalGlossyDefaults-31Aug18-SpectralM0-averaged-kh-s20-ca27-cg7.icm?dl=0
As one can see, both the Relative Colorimetric and Perceptual tables show
discontinuities in the gradients. For Perceptual, which is more important
to me, the multiple tone reversals are less though still present in the
dark colours. The neutrals are somehow unaffected.
When I generated an ICC profile in i1Profiler from the same measurement
data, this is the profile I get, which is smooth with no discontinuities.
https://www.dropbox.com/s/1iorgeyaite41k2/KHLIndigo10000-Nevia-Calglossy-18-08-31-M0averaged-TIC400-bs30mb100bcmedplusbw50-c0s0ng0-sm0.icm?dl=0
Perhaps I have entered something wrong into the argument string which is
causing Argyll to be unable to create a smooth profile? Thank you for your
time looking into this.
Regards,
Samuel
