Rendering Intents

[Stalis Man]

Hi,



I have made four device links using the collink commands
such as:



collink -v -qh -i s -g srgb.icc cmy190214_SP.icm
sRGBcmy190214_s.icm for a saturated intent.



I am profiling a printer with just C, M and Y hence I
made the cmy190214_SP.icm profile without any K at all with:



colprof -v -qh -kz -l300 -iD50 -nP -nS -S srgb.icc
cmy090114.



I printed of the standard colour checker card I was
surprised that the perceptual intent looked rather ‘smokey’ compared to the
Saturated intent whilst the Relative looked too brown compared to the Absolute.



As a result folks seem to prefer the saturated intent whilst
I had expected the Perceptual to be the one they would want. Their choice is
down to the general darkening of the perceptual.



Is the perceptual being smokey/ darkened to be expected or
have I dropped an obvious clanger.



I was using X-Rite’s i1publisher before but couldn’t handle
CMY profiles too well and so far the results with Argyll are better, I just
want to make them even better.



Best Regards



Stalis

[Graeme Gill]

Stalis Man wrote:

Hi,

collink -v -qh -i s -g srgb.icc cmy190214_SP.icm sRGBcmy190214_s.icm
for a saturated intent.
I am profiling a printer with just C, M and Y hence I
colprof -v -qh -kz -l300 -iD50 -nP -nS -S srgb.icc cmy090114.

So the printer is defined to be CMY space ? If so, the -kz doesn't
apply - there is no K ink.

Or do you mean that you have profiled it as a CMYK printer ?
If so, -kz is not guaranteed to give you zero K ink, it just gives
you the minimum possible that achieves the target colors.

I printed of the standard colour checker card I was
surprised that the perceptual intent looked rather ‘smokey’ compared to the
Saturated intent whilst the Relative looked too brown compared to the Absolute.

It's not possible to comment without knowing the source of the image file,
the workflow used to print it, or be able to see the result.

As a result folks seem to prefer the saturated intent whilst
I had expected the Perceptual to be the one they would want. Their choice is
down to the general darkening of the perceptual.

Naturally people prefer "pop" if they see it in isolation and want impact.
"pop" is not accuracy though.

Is the perceptual being smokey/ darkened to be expected or
have I dropped an obvious clanger.

Check what you mean by CMY. To profile in CMY you need to create a CMY
chart and print with just CMY.

Graeme Gill.

[Stalis Man]

Date: Fri, 4 Apr 2014 16:28:15 +1100
Subject: [argyllcms] Re: Rendering Intents
Hi,

collink -v -qh -i s -g srgb.icc cmy190214_SP.icm sRGBcmy190214_s.icm
for a saturated intent.
I am profiling a printer with just C, M and Y hence I
colprof -v -qh -kz -l300 -iD50 -nP -nS -S srgb.icc cmy090114.

So the printer is defined to be CMY space ? If so, the -kz doesn't
apply - there is no K ink.
Or do you mean that you have profiled it as a CMYK printer ?
If so, -kz is not guaranteed to give you zero K ink, it just gives
you the minimum possible that achieves the target colors.

I printed of the standard colour checker card I was
surprised that the perceptual intent looked rather ‘smokey’ compared to the
Saturated intent whilst the Relative looked too brown compared to the Absolute.

It's not possible to comment without knowing the source of the image file,
the workflow used to print it, or be able to see the result.

As a result folks seem to prefer the saturated intent whilst
I had expected the Perceptual to be the one they would want. Their choice is
down to the general darkening of the perceptual.

Naturally people prefer "pop" if they see it in isolation and want impact.
"pop" is not accuracy though.

Is the perceptual being smokey/ darkened to be expected or
have I dropped an obvious clanger.

Check what you mean by CMY. To profile in CMY you need to create a CMY
chart and print with just CMY.
Graeme Gill.
Thanks for your response Graeme. I'll clarify. I am profiling a CMY printer and I did create and measure a CMY chart. I used the -kz because I wasn't sure how things worked internally, I'll retry without it. It's CMY because I am using dyefilm which doesn't support K. Regarding the 'smokiness' it is a very subjective term to use. What I was trying to explain was that having created a device link with a perceptual intent and another with an absolute intent that the Yellows looked quite different although within gamut. Absolute was fine whilst Perceptual looked as if I were looking through a dirty window. Maybe I am expecting too much? I understand that perceptual is supposed to be more pleasing to photographers but this from a large gamut to a smaller gamut is decidedly not. I'll look at the D65 / D50 missmatch. Best Regards Stali

[Krzysztof Tomczyk]

IMHO you have a RGB printer. I experienced the same symptoms in the past.
Best regards
Krzysztof

Post by Stalis Man

Date: Fri, 4 Apr 2014 16:28:15 +1100
Subject: [argyllcms] Re: Rendering Intents
Hi,

collink -v -qh -i s -g srgb.icc cmy190214_SP.icm

sRGBcmy190214_s.icm

for a saturated intent.
I am profiling a printer with just C, M and Y hence I
colprof -v -qh -kz -l300 -iD50 -nP -nS -S srgb.icc cmy090114.

So the printer is defined to be CMY space ? If so, the -kz doesn't
apply - there is no K ink.
Or do you mean that you have profiled it as a CMYK printer ?
If so, -kz is not guaranteed to give you zero K ink, it just gives
you the minimum possible that achieves the target colors.

I printed of the standard colour checker card I was
surprised that the perceptual intent looked rather ‘smokey’

compared to the

Saturated intent whilst the Relative looked too brown compared to

the Absolute.

It's not possible to comment without knowing the source of the image

file,

the workflow used to print it, or be able to see the result.

As a result folks seem to prefer the saturated intent whilst
I had expected the Perceptual to be the one they would want. Their

choice is

down to the general darkening of the perceptual.

Naturally people prefer "pop" if they see it in isolation and want

impact.

"pop" is not accuracy though.

Is the perceptual being smokey/ darkened to be expected or
have I dropped an obvious clanger.

Check what you mean by CMY. To profile in CMY you need to create a

CMY

chart and print with just CMY.
Graeme Gill.
Thanks for your response Graeme. I'll clarify. I am profiling a

CMY printer and I did create and measure a CMY chart. I used the -kz
because I wasn't sure how things worked internally, I'll retry without
it. It's CMY because I am using dyefilm which doesn't support K.
Regarding the 'smokiness' it is a very subjective term to use. What I
was trying to explain was that having created a device link with a
perceptual intent and another with an absolute intent that the Yellows
looked quite different although within gamut. Absolute was fine whilst
Perceptual looked as if I were looking through a dirty window. Maybe I
am expecting too much? I understand that perceptual is supposed to be
more pleasing to photographers but this from a large gamut to a smaller
gamut is decidedly not. I'll look at the D65 / D50 missmatch. Best
Regards Stali

[Stalis Man]

Hi Kryzysztof,

I am writing a little tool to take CMY encoded Tiff files and convert them to a file that I can send direct to the printer via USB bypassing its usual driver.

So I have no RGB in my work flow.

Best Regards

Stali



Subject: [argyllcms] Re: Rendering Intents
From: ***@poczta.onet.pl
Date: Fri, 4 Apr 2014 13:11:31 +0200
To: ***@freelists.org

IMHO you have a RGB printer. I experienced the same symptoms in the past.

Best regards

Krzysztof

Date: Fri, 4 Apr 2014 16:28:15 +1100
Subject: [argyllcms] Re: Rendering Intents
Hi,

collink -v -qh -i s -g srgb.icc cmy190214_SP.icm sRGBcmy190214_s.icm
for a saturated intent.
I am profiling a printer with just C, M and Y hence I
colprof -v -qh -kz -l300 -iD50 -nP -nS -S srgb.icc cmy090114.

So the printer is defined to be CMY space ? If so, the -kz doesn't
apply - there is no K ink.
Or do you mean that you have profiled it as a CMYK printer ?
If so, -kz is not guaranteed to give you zero K ink, it just gives
you the minimum possible that achieves the ta

rget
colors.

I printed of the standard colour checker card I was
surprised that the perceptual intent looked rather 'smokey' compared to the
Saturated intent whilst the Relative looked too brown compared to the Absolute.

It's not possible to comment without knowing the source of the image file,
the workflow used to print it, or be able to see the result.

As a result folks seem to prefer the saturated intent whilst
I had expected the Perceptual to be the one they would want. Their choice is
down to the general darkening of the perceptual.

Naturally people prefer "pop" if they see it in isolation and want impact.
"pop" is not accuracy though.

Is the perceptual being smokey/ darkened to be expected or
have I dropped an obvious clanger.

Check what

you
mean by CMY. To profile in CMY you need to create a CMY

chart and print with just CMY.
Graeme Gill.
Thanks for your response Graeme. I'll clarify. I am profiling a CMY printer and I did create and measure a CMY chart. I used the -kz because I wasn't sure how things worked internally, I'll retry without it. It's CMY because I am using dyefilm which doesn't support K. Regarding the 'smokiness' it is a very subjective term to use. What I was trying to explain was that having created a device link with a perceptual intent and another with an absolute intent that the Yellows looked quite different although within gamut. Absolute was fine whilst Perceptual looked as if I were looking through a dirty window. Maybe I am expecting too much? I understand that perceptual is supposed to be more pleasing to photographers but this f

rom a
large gamut to a smaller gamut is decidedly not. I'll look at the D65 / D50 missmatch. Best Regards Stali



--

Wys³ane za pomoc± K-9 Mail.

[Krzysztof Tomczyk]

Hi
So this is not the case which I experienced, but what I think is worth to consider is that some printers receive RGB only and internally do cmy(k) separations. I hope this is not yours case.
Best regards

Post by Stalis Man
Hi Kryzysztof,
I am writing a little tool to take CMY encoded Tiff files and convert
them to a file that I can send direct to the printer via USB bypassing
its usual driver.
So I have no RGB in my work flow.
Best Regards
Stali
Subject: [argyllcms] Re: Rendering Intents
Date: Fri, 4 Apr 2014 13:11:31 +0200
IMHO you have a RGB printer. I experienced the same symptoms in the past.
Best regards
Krzysztof

Date: Fri, 4 Apr 2014 16:28:15 +1100
Subject: [argyllcms] Re: Rendering Intents
Hi,

collink -v -qh -i s -g srgb.icc cmy190214_SP.icm

sRGBcmy190214_s.icm

for a saturated intent.
I am profiling a printer with just C, M and Y hence I
colprof -v -qh -kz -l300 -iD50 -nP -nS -S srgb.icc cmy090114.

So the printer is defined to be CMY space ? If so, the -kz doesn't
apply - there is no K ink.
Or do you mean that you have profiled it as a CMYK printer ?
If so, -kz is not guaranteed to give you zero K ink, it just gives
you the minimum possible that achieves the ta

rget
colors.

I printed of the standard colour checker card I was
surprised that the perceptual intent looked rather 'smokey'

compared to the

Saturated intent whilst the Relative looked too brown compared to

the Absolute.

It's not possible to comment without knowing the source of the image

file,

the workflow used to print it, or be able to see the result.

As a result folks seem to prefer the saturated intent whilst
I had expected the Perceptual to be the one they would want. Their

choice is

down to the general darkening of the perceptual.

Naturally people prefer "pop" if they see it in isolation and want

impact.

"pop" is not accuracy though.

Is the perceptual being smokey/ darkened to be expected or
have I dropped an obvious clanger.

Check what

you
mean by CMY. To profile in CMY you need to create a CMY

chart and print with just CMY.
Graeme Gill.
Thanks for your response Graeme. I'll clarify. I am profiling a

CMY printer and I did create and measure a CMY chart. I used the -kz
because I wasn't sure how things worked internally, I'll retry without
it. It's CMY because I am using dyefilm which doesn't support K.
Regarding the 'smokiness' it is a very subjective term to use. What I
was trying to explain was that having created a device link with a
perceptual intent and another with an absolute intent that the Yellows
looked quite different although within gamut. Absolute was fine whilst
Perceptual looked as if I were looking through a dirty window. Maybe I
am expecting too much? I understand that perceptual is supposed to be
more pleasing to photographers but this f
rom a
large gamut to a smaller gamut is decidedly not. I'll look at the D65
/ D50 missmatch. Best Regards Stali
--
Wysłane za pomocą K-9 Mail.

[Stalis Man]

Hi Krzysztof,

I am indeed doing cmy separations internally. Maybe madness but that's just how it is.

Best Regards

Stali

Subject: [argyllcms] Re: Rendering Intents
From: k2mil-***@public.gmane.org
Date: Mon, 7 Apr 2014 15:50:08 +0200
To: argyllcms-***@public.gmane.org

Hi

So this is not the case which I experienced, but what I think is worth to consider is that some printers receive RGB only and internally do cmy(k) separations. I hope this is not yours case.

Best regards

On 7 kwietnia 2014 08:44:36 CEST, Stalis Man <stalis_man-***@public.gmane.org> wrote:



Hi Kryzysztof,

I am writing a little tool to take CMY encoded Tiff files and convert them to a file that I can send direct to the printer via USB bypassing its usual driver.

So I have no RGB in my work flow.

Best Regards

Stali



Subject: [argyllcms] Re: Rendering Intents
From: k2mil-***@public.gmane.org
Date: Fri, 4 Apr 2014 13:11:31 +0200
To: argyllcms-***@public.gmane.org

IMHO you have a RGB printer. I experienced the same symptoms in the past.

Best regards

Krzysztof

Date: Fri, 4 Apr 2014 16:28:15 +1100
Subject: [argyllcms] Re: Rendering Intents
Hi,

collink -v -qh -i s -g srgb.icc cmy190214_SP.icm sRGBcmy190214_s.icm
for a saturated intent.
I am profiling a printer with just C, M and Y hence I
colprof -v -qh -kz -l300 -iD50 -nP -nS -S srgb.icc cmy090114.

So the printer is defined to be CMY space ? If so, the -kz doesn't
apply - there is no K ink.
Or do you mean that you have profiled it as a CMYK printer ?
If so, -kz is not guaranteed to give you zero K ink, it just gives
you the minimum possible that achieves the ta

rget
colors.

I printed of the standard colour checker card I was
surprised that the perceptual intent looked rather 'smokey' compared to the
Saturated intent whilst the Relative looked too brown compared to the Absolute.

It's not possible to comment without knowing the source of the image file,
the workflow used to print it, or be able to see the result.

As a result folks seem to prefer the saturated intent whilst
I had expected the Perceptual to be the one they would want. Their choice is
down to the general darkening of the perceptual.

Naturally people prefer "pop" if they see it in isolation and want impact.
"pop" is not accuracy though.

Is the perceptual being smokey/ darkened to be expected or
have I dropped an obvious clanger.

Check what

you
mean by CMY. To profile in CMY you need to create a CMY

chart and print with just CMY.
Graeme Gill.
Thanks for your response Graeme. I'll clarify. I am profiling a CMY printer and I did create and measure a CMY chart. I used the -kz because I wasn't sure how things worked internally, I'll retry without it. It's CMY because I am using dyefilm which doesn't support K. Regarding the 'smokiness' it is a very subjective term to use. What I was trying to explain was that having created a device link with a perceptual intent and another with an absolute intent that the Yellows looked quite different although within gamut. Absolute was fine whilst Perceptual looked as if I were looking through a dirty window. Maybe I am expecting too much? I understand that perceptual is supposed to be more pleasing to photographers but this f

rom a
large gamut to a smaller gamut is decidedly not. I'll look at the D65 / D50 missmatch. Best Regards Stali

[Graeme Gill]

What I was trying to explain was that having created a device link with a perceptual
intent and another with an absolute intent that the Yellows looked quite different
although within gamut. Absolute was fine whilst Perceptual looked as if I were
looking through a dirty window. Maybe I am expecting too much? I understand that
perceptual is supposed to be more pleasing to photographers but this from a large
gamut to a smaller gamut is decidedly not.

Hi,
it's to be expected that Absolute and Perceptual will look quite
different. Whether there is anything wrong with your profile is not possible
to say without the .ti3 file, the colprof parameters and the resulting profile.

Perceptual attempts to preserve the nuances of the reproduction, as well
as producing a visually pleasing result. If you are transforming to a very
small gamut, the compression needed so as to not loose the source gamut may not
have the same impact as a mapping that clips shadows or saturated colors.

Graeme Gill.

[Stalis Man]

Yes I understand all that Graeme, my concern was that 'perceptual' seems to be trumpeted as the intent most likely to 'succeed' and hence I wonder if I have done something wrong.

This morning I generated a profile with D65 to match that of sRGB's D65 and then generated some device links.

Perceptual and Saturated device links were marginally better whilst relative and absolute were decidedly worse. Relative for example gave a 'browner' black than absolute. (Absolute did however seem to have the cyan hue removed).

On a range of images and the basis of votes from a small number of friends the Saturated intent seems a clear winner.

It's just now what I expected before doing the work.

Thanks for all your help

Stali

Date: Mon, 7 Apr 2014 18:32:27 +1000
Subject: [argyllcms] Re: Rendering Intents

What I was trying to explain was that having created a device link with a perceptual
intent and another with an absolute intent that the Yellows looked quite different
although within gamut. Absolute was fine whilst Perceptual looked as if I were
looking through a dirty window. Maybe I am expecting too much? I understand that
perceptual is supposed to be more pleasing to photographers but this from a large
gamut to a smaller gamut is decidedly not.

Hi,
it's to be expected that Absolute and Perceptual will look quite
different. Whether there is anything wrong with your profile is not possible
to say without the .ti3 file, the colprof parameters and the resulting profile.
Perceptual attempts to preserve the nuances of the reproduction, as well
as producing a visually pleasing result. If you are transforming to a very
small gamut, the compression needed so as to not loose the source gamut may not
have the same impact as a mapping that clips shadows or saturated colors.
Graeme Gill.

[Ben Goren]

Post by Stalis Man
Yes I understand all that Graeme, my concern was that 'perceptual' seems to be trumpeted as the intent most likely to 'succeed' and hence I wonder if I have done something wrong.

It depends on what the goal is, as well as other constraints such as device gamut.

If you're doing fine art reproduction and the printer has a smaller gamut than the original, perceptual is very likely your best bet. It'll look a bit washed out in a side-by-side comparison, but that should be the only significant complaint.

If you're printing photographic originals and the photographer is a fan of "pop," especially HDR and similar aesthetics, saturation may well win the day.

And, for certain types of prepress work, only absolute is even theoretically acceptable.

Right tool for the right job....

b&

[Graeme Gill]

Post by Stalis Man
Yes I understand all that Graeme, my concern was that 'perceptual' seems to be
trumpeted as the intent most likely to 'succeed' and hence I wonder if I have done
something wrong.

Right, but we haven't established whether things are working as intended, or
whether something is awry.

Graeme Gill.