Hi all,
you might check a method we have develop termed „discernible tonal steps“:
https://www.fogra.org/en/fogra-research/wc-digital-printing/digital-printing-current-projects/imagequality-35003/research-topics/tonal-steps/evaluation-the-number-of-tonal-steps-28cmyk29.html
<https://www.fogra.org/en/fogra-research/wc-digital-printing/digital-printing-current-projects/imagequality-35003/research-topics/tonal-steps/evaluation-the-number-of-tonal-steps-28cmyk29.html>
regards
Andy
Wir stellen den Hintergrund in den Vordergrund.
We put the background in the foreground.
—————————————————————————
Dr.-Ing. Andreas Kraushaar
Abt. Vorstufentechnik
Fogra
Forschungsinstitut für Medientechnologien e.V.
Einsteinring 1a
85609 Aschheim b. München
Telefon: +49 89. 431 82 - 335
Telefax: +49 89. 431 82 - 100
E-mail: kraushaar@xxxxxxxxx
Internet: www.fogra.org
-----------------------------------------------------------------
Geschäftsführer: Dr. Eduard Neufeld | Sitz des Vereins: Aschheim b. München |
Vereinsregister 4909
Am 13.10.2018 um 12:19 schrieb Paul Langmead <paul@xxxxxxxxxxxxxxxx>:
Hi Yves,
First of all, 16 bit doesn't expand a colour space, it simply adds more
colours in between, which will reduce banding if a gradient between two very
similar colours is present in your image. If print is your goal, then all of
this is rather moot, as even the best paper and ink combinations have small
gamuts compared to most capture devices (the gamut of Fujichrome Provia 100F
and the Hasselblad X1 scanner is astaunding). Furthermore, displays are still
only 10 bit, and, as of writing all modern operating systems offer standard
support only for 8-bit video output (MS is dabbling with 10-bit, but it's
certainly not mainstream at this stage - this will change of course).
As I understand it, the benefit to sending 16 bit data to the printer is that
you allow the printer software to convert the data to 8-bit before reaching
the print head rather than Photoshop - the advantage being that the printer
manufacturer probably better understands how to optimise for it's print head
than Adobe. This of course will change (and may have already with newer
equipment), but the weak link is still the paper/ink, which is of course why
we are all here. This is unlikely to improve more than incrementally in the
foreseeable future.
I guess what I'm trying to say is that being particular about your output
profiles is probably more important than worrying about bit depth at the
print head, but I'm hoping Greame will chime in if I'm wrong.
Cheers,
Paul Langmead
__________________________________________________________
Atelier Paul Langmead - Photography and Fine Art Printing
Wilhelminenhofstraße 68a, Berlin 12459
paullangmead.com <http://paullangmead.com/> / +49 (0)17677251462 /
paul@xxxxxxxxxxxxxxxx <mailto:paul@xxxxxxxxxxxxxxxx>
On Sat, 13 Oct 2018 at 11:00, Yves Gauvreau <gauvreau-yves@xxxxxxxxxxx
<mailto:gauvreau-yves@xxxxxxxxxxx>> wrote:
Graeme, Ben and Paul,
I have 2 objectives about this high bit depth printing stuff the first one
is that I have images of nature that present a very wide range of colors
often a more or less important part of them outside sRGB space, I always
shot raw in the hope to capture all of them, I use mostly use Lightroom with
its huge colorspace ProPhoto RGB as workspace again such that I don't clip
anything at birth as someone put it. I've setup Photoshop to use the same
colorspace as LR. I think if I convert voluntary or not such images to 8 bit
depth image I will lose quite a lot of data that may very well produce
prints of less than optimal quality. From the little I know, papers, ink and
printer combo are now capable of producing colors that cover a larger color
space then sRGB often filling a large part of Adobe RGB and even sometimes
they present colors protruding outside of AdobeRGB, so I want to exploit
this as much as possible. As for high bit depth, my thinking is that with
these bigger volume of colors, the distance between to colors should
increase and thus artefact like banding should be worst then with a smaller
distance between colors. Now this should practically be way to minute to
detect even considering noise and other factors if any at higher bit depth,
thus my interest.
A long time ago I came across a guy named Norman Koren at the time he
produced an image to help us adjust our monitor visually to have a specific
gamma, either 1.8 for a Mac and 2.2 for the window OS. I suspect this was
done using advanced knowledge about frequencies and a pretty good
understanding of visual perception.
The reason I'm writing this in the middle of the night for me, is that I
wonder if a similar approach using knowledge about frequencies and visual
perception would make it possible to kind of amplify the signal well above
the noise level to make it easily discernible that this print can only have
been made using a 16 bit printing process or not.
Adobe as a color print utility that I see often mention so that no color
management is use in the printing process for the creation of profiles,
maybe there is a possibility there as well. I've written to Adobe about this
high bit depth printing thing but I didn't mention this utility. They about
convinced me that none of their product such as LR and PS change the mode to
8 bit per channel during the printing process at any time on their side.
Anyone knows a physicist?
Thanks for all your effort,
Yves
-----Original Message-----
From: argyllcms-bounce@xxxxxxxxxxxxx <mailto:argyllcms-bounce@xxxxxxxxxxxxx>
[mailto:argyllcms-bounce@xxxxxxxxxxxxx ;
<mailto:argyllcms-bounce@xxxxxxxxxxxxx>]
On Behalf Of Graeme Gill
Sent: Friday, October 12, 2018 7:29 PM
To: argyllcms@xxxxxxxxxxxxx <mailto:argyllcms@xxxxxxxxxxxxx>
Subject: [argyllcms] Re: testing
Yves Gauvreau wrote:
Anyone as an idea if it is possible to test, check or verify if a printwas done at a high
bit depth of 16 bit per channel?
Hi,
it's going to be tough to do with a printer. It's hard to be sure
the depth is > 8 bits
with a display, and a display is much less noisy than the printing process.
Printing has a lot of noise, which covers up quantization. Primarily it is
screening
that adds noise, but paper itself usually has texture, which adds to the
randomness.
If you have a printer interface to a system that makes the 8 bits reasonably
visually uniform, then my guess is that you aren't going to be able to tell
the difference between 8 bit and something more.
If on the other hand you are dealing with a raw printing system, then some
parts of the response are likely to be very non-linear (due to dot gain),
and the difference between 8 and 16 bit control will be quite noticeable.
So test patches that span an area of high printer response (if it has
any), may be visible or measurable.
Cheers,
Graeme Gill.
