Pages: 1 2 [3] 4 5 6   Go Down

Author Topic: New color chart for high quality profiling in captures  (Read 7867 times)

hurodal

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 54
    • My website
Re: New color chart for high quality profiling in captures
« Reply #40 on: July 11, 2019, 10:40:11 am »

If you're referring to Kevin Raber, he hasn't been connected to this site for some time.

Oh, really?
Ok, I see now it's Josh. Is he michael's brother?

Regards,
Logged
Hugo Rodriguez

Coloratti member | PhaseOne certified professional | BenQ Ambassador

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #41 on: July 11, 2019, 10:51:32 am »

Ok, I think you’re going too far. I replied and gave arguments, screenshots and answers to every single critic that you told me; all I see from you is a person that simply jumps to another new small detail to criticize when I reply to the last one. And in this your last post, you're laughing about the project I've been working slowly for 4 years.
So the before's examples are using what target and built by you as custom camera profiles with what software? Did I miss that answer? If so, I apologize. Please clarify or point me to that text.
Quote
Let me tell you something, Andrew: when I came here to share this my project, I expected to have a nice discussion with many experts here.
That might have been a mistake as you've seen.
Quote
A discussion on where some would like my chart and some don’t.
I don't have enough data to tell you if I like or dislike it. Just like you don't have enough data to tell us how many pro photographers ALL use C1.
Quote
I will totally respect their opinion, of course, but as well as I would never throw shit over other person’s project, I expect them to not do that to me.

Shit? was thrown at the before and after examples as interpreted. Again, is the before image on your site a custom profile and if so, using what current target?
Quote
There always should be a respect for others work, because we can all discuss without going low.
Respect is earned.
Quote
So now you’re telling me that the everything else in the image looks 'WORSE'. And you laugh
I believe I made clear what I subjectively saw of that one image and what looked 'better' and all the rest that looked worse. The paper trail of my comments remain here to see.
Please show me the paper trail of yours, that perhaps I missed, that describes how your "before" examples were built using a CUSTOM camera profile as a comparison to the custom profile examples made with your target.
Quote
You might not seen the answer and images of the CC24 I've already posted, on where it can be seen that the profiled capture is almost spot on.
OT
Quote
But after you said that, now I can see that you ignore how phosphor colors behave, which in the end tells me that you might know Ps very well, but regarding camera profiling you're lacking quite some knowledge.
OT, I've ONLY commented so far on two attributes: A comparison that thus far appears to be unfair. A single image that I believe looks worse with the application of your profile. It's subjective of course.
Quote
Also, if you ever come to a Spanish forum on where I use to be, I will first welcome you (because that’s gentleman do when a new one comes and introduces himself) and then will start to discuss.
Then ask questions who's answers you wish to hear on a Spanish photo forum; I have no desire and will not be posting on any such site.
Quote
But I won’t try to find out any mistake or misspelling in your book just to hit hard with my finger on it.

I haven’t and don't believe anyone else is taking you to task about spelling are they? IF so, point that post out to me. Or stop playing the victim.
Quote
I don't know if you know Chris Murphy or if he is also here, but you can ask him how did I treated him when I managed to bring him to a big intl. photo show in Barcelona a few years ago. We both performed a conference about color management, together.
I know Chris quite well and for decades, but your comments are again off topic.
Quote
Would you like if I read your book and post here every single and small mistake I find out? And, trust me, I will find them. Even that it’s not in my mother language.
I haven’t and don't believe anyone else is taking you to task about spelling are they? IF so, point that post out to me. Or stop playing the victim.
Quote
Now, that said, let me answer the main question
:
Yes please: HOW were the 'before" images on your site produced with respect to the target used and the software used to build them? OR are they canned C1 camera profiles?
Quote
Now, let me tell you that I wasted enough time replying to this; I won’t be replying unless you keep a respectful position.
And I've wasted enough time too. But then, who came here asking for feedback? Wasn't me.
Quote
PS: Regarding your last post, If you ever asked me to add another examples to make the comparison more fair/complete/real/whatever, I’d kindly find some time to add them. Now you achieved the opposite from me.
I'm asking you to explain how ALL the 'before' images were produced with respect to a target and generation of the camera profile. That's all I'm asking. IF as I believe but may be wrong, the 'Before' images were all canned C1 profiles, your examples ARE bogus. And I explained HOW you could fix that issue and show us JUST what your target provides using an apples to apples comparison. CAN or DID you do this? Yes or no?
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #42 on: July 11, 2019, 10:52:39 am »

Oh, really?
Yes, really.

Quote
Ok, I see now it's Josh. Is he michael's brother?
No.  ???
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Alexey.Danilchenko

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 257
    • Spectron
Re: New color chart for high quality profiling in captures
« Reply #43 on: July 11, 2019, 11:31:04 am »

Shooting targets are not really high quality profiling. And it is not easy either. Too much investment into proper lighting setup, flat fielding, fighting with flare/glare from highly reflective patches and so on. The ultimate solution if you want quality in your profiles is to measure spectral sensitivity of the sensor, spectral transmission of the lenses and filters (so the complete spectral profile for camera+lens+filter could be obtained), measure the lighting spectra and build profile targeting specific set of colour samples you need for high quality reproduction from that data. That approach will allow any samples to be used and resulting profile optimised for those sample colours.

« Last Edit: July 11, 2019, 12:32:16 pm by Alexey.Danilchenko »
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #44 on: July 11, 2019, 11:32:20 am »

Shooting targets are not really high quality profiling. And it is not easy either. Too much investment into proper lighting setup, flat fielding, fighting with flare/glare from highly reflective patches and so on. The ultimate solution if you want quality in your profiles is to measure spectral sensitivity of the sensor, spectral transmission of the lenses and filters (so the complete spectral profile for camera+lens+filter could be obtained), measure the lighting spectra and build profile targeting specific set of colour samples you need for hi quality reproduction from that data. That approach will allow any samples to be used and resulting profile optimised for those sample colours.
+1!
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #45 on: July 11, 2019, 11:45:30 am »

This needs to be explained a bit more too:
Quote
It should be noted that in the developing there is no adjustment of any kind, except the WB and the color profile. No styles, no color editing, no layers, no masks or anything similar.
All images on the site with all Before shots? Because, with a 5x5 sampler in Photoshop of the before, I'm hard pressed to see where WB was set:

Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

hurodal

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 54
    • My website
Re: New color chart for high quality profiling in captures
« Reply #46 on: July 11, 2019, 12:49:42 pm »

Photographing a profiled monitor's image of a chart to make a profile works quite well, and isn't that difficult. But only if the monitor's white point is D50 and you want to use the profile to tag pictures the camera takes of the monitor. Otherwise it's useless.

A monitor has 3 primaries that are quite separated in wavelength. Especially wide gamut monitors. You will get large metameric error when applying these to real life camera images or even images of printed material. CYM prints at least spectrally gradually overlap and blend into each other.

Photographing an image in a monitor is easy, but capturing a very high quality capture from a monitor, with absolutely no hue shifting from corner to corner (even with IPS), 100% free of moire and 100% perfect focus and fine detail is very hard to achieve.
Logged
Hugo Rodriguez

Coloratti member | PhaseOne certified professional | BenQ Ambassador

Doug Peterson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4210
    • http://www.doug-peterson.com
Re: New color chart for high quality profiling in captures
« Reply #47 on: July 11, 2019, 01:43:37 pm »

It's very easy Andrew: it's the generic embedded profile into C1 (or ACR in that case). It's the profile that C1 automatically uses when you open your raw files. Canned profile? I've never heard that, don't know what you mean.
How do Phase one's engineers built that profile is something I don't know, but what do I know is that they aim to 'pleasant' colors, not faithful.

What you say is correct for Capture One Pro - they aim for pleasing color

Capture One CH on the other hand has profiles meticulously created and test for absolute accuracy.

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: New color chart for high quality profiling in captures
« Reply #48 on: July 11, 2019, 01:51:58 pm »

Photographing an image in a monitor is easy, but capturing a very high quality capture from a monitor, with absolutely no hue shifting from corner to corner (even with IPS), 100% free of moire and 100% perfect focus and fine detail is very hard to achieve.
Moire's easily removed by very slight defocussing. It isn't even necessary to remove it all since you are adding all the pixels within a subset of each patch. You have to work pretty hard before moire, even if visible, alters the average patch color more than a fraction of a dE.

You are correct that hue changes are an issues when off 90 degrees. There are also L* differences across the face of a monitor and they can be pretty significant. In fact they can vary even more than hue over just a fewn inches. Even with high end monitors. But that is overcome effectively by using a large, randomized, patch set and is helped significantly by the relatively (compared to printers) independent response monitors tend to have. I did some work on this as a possible approach to profiling monitors (not cameras) . It's quite easy and produces excellent profiles except that it has one, major, problem. Camera response is not intrinsically colorimetric so you also have to have a spectroradiometer to measure the primaries and set the white point. And if you have that you can just use it to create monitor profiles. So it was just an interesting experiment.

But all that aside, even with a "perfect" monitor one can't produce good camera profiles this way unless one wants to take pictures of images from the same monitor. Might be something an academic might do but there is no reason anyone else would that I can imagine.

Also, your printer charts should do quite well for repro work where your images are printed materials. At least those printed with CYMK processes.  I've used a similar approach here to profile my scanner. I wound up using around 2k of randomized patches. And the profiles worked quite well across my three printers. Better than a CC, IT8, or CCSG scanner profile.   But the bigger problem with scanners is something called large area crosstalk that is intrinsic to most all reflection scanner designs. It can be a much larger problem than metameric error and, like metameric error, can't be solved by any profiling technique. Luckily, it doesn't affect transmission scans nor is it a significant issue for cameras.

Careful not to conflate accuracy with what looks good. Accuracy is meaningful only for scene referred images where one can actually used dE stats. And generally, colorimetrically accurate prints are only useful for reproduction work where the goal is to make a copy indistinguishable from the original when viewed side by side. The same process will produce rather unappealing prints in almost all other uses.

As another aside, when I see PR stuff like claims of "Absolute Accuracy" from a camera profile I chalk the rest of the claims to marketing blather. Camera CFA responses are not linearly congruent to L/I or even very close.
Logged

hurodal

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 54
    • My website
Re: New color chart for high quality profiling in captures
« Reply #49 on: July 11, 2019, 01:54:19 pm »

This needs to be explained a bit more too:All images on the site with all Before shots? Because, with a 5x5 sampler in Photoshop of the before, I'm hard pressed to see where WB was set:

The problem with those who think that know everything is that is hard to make them understand they don't.
Do you know what color was the background or what lighting was used, sir?
That's the real question.

PS: I don't know why I'm still replying you.
Logged
Hugo Rodriguez

Coloratti member | PhaseOne certified professional | BenQ Ambassador

hurodal

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 54
    • My website
Re: New color chart for high quality profiling in captures
« Reply #50 on: July 11, 2019, 01:57:46 pm »

What you say is correct for Capture One Pro - they aim for pleasing color

Capture One CH on the other hand has profiles meticulously created and test for absolute accuracy.

I didn't know that CH profiles were different from the other versions. Does it come with these profiles for every single camera?
Logged
Hugo Rodriguez

Coloratti member | PhaseOne certified professional | BenQ Ambassador

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #51 on: July 11, 2019, 02:02:38 pm »

What you say is correct for Capture One Pro - they aim for pleasing color

Either or, the examples shown on the site ARE bogus, as the author of the target is comparing a canned camera profile to a custom profile.
AGAIN, a fair test would be to have both profiles be identical in creation EXCEPT for the targets being different. Then and only then would we see what that target brings to the party.
The examples (before) also are not fair to PhaseOne either, but that's kind of besides the point.
Quote
Capture One CH on the other hand has profiles meticulously created and test for absolute accuracy.
Then, if so, if the profiles are meticulously created for absolute accuracy (a dE report would be lovely), the new target isn't going to bring anything to the party.  ;D
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #52 on: July 11, 2019, 02:05:49 pm »

PS: I don't know why I'm still replying you.
"It is better to say nothing and be thought a fool than to open your mouth and remove all doubt".- Abraham Lincoln  ::)
Quote
I didn't know that CH profiles were different from the other versions.
Even more reason to dismiss the presentation.....
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Jack Hogan

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 798
    • Hikes -more than strolls- with my dog
Re: New color chart for high quality profiling in captures
« Reply #53 on: July 11, 2019, 03:34:49 pm »

Then, if so, if the profiles are meticulously created for absolute accuracy (a dE report would be lovely), the new target isn't going to bring anything to the party.  ;D

Well. it would be interesting to test whether C1 CH's 'absolute accuracy' is indeed so by capturing Hugo's target with it :)
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #54 on: July 11, 2019, 03:37:38 pm »

Well. it would be interesting to test whether C1 DH's 'absolute accuracy' is indeed so by capturing Hugo's target with it :)
Brilliant idea.

And if the claim is true, then Hugo should update all the "before" images on his blog with that rendering using this canned profile. Plus, I still think (beating a dead horse), create a CUSTOM profile using whatever target, other than his own, to compare with a profile made with his target.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

hurodal

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 54
    • My website
Re: New color chart for high quality profiling in captures
« Reply #55 on: July 11, 2019, 04:54:01 pm »

Shooting targets are not really high quality profiling. And it is not easy either. Too much investment into proper lighting setup, flat fielding, fighting with flare/glare from highly reflective patches and so on. The ultimate solution if you want quality in your profiles is to measure spectral sensitivity of the sensor, spectral transmission of the lenses and filters (so the complete spectral profile for camera+lens+filter could be obtained), measure the lighting spectra and build profile targeting specific set of colour samples you need for high quality reproduction from that data. That approach will allow any samples to be used and resulting profile optimised for those sample colours.

Well, I think a pro photographer and/or a color management technician can do some things to achieve a quite fine reproduction without all that. Just with a good camera, good lighting and a good chart and software.
How many photographers or color mngmt consultants do have such equipment?
I don't want to seem ironic, but if we follow your recommendations, what would be the next step? Call the NASA?
Logged
Hugo Rodriguez

Coloratti member | PhaseOne certified professional | BenQ Ambassador

hurodal

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 54
    • My website
Re: New color chart for high quality profiling in captures
« Reply #56 on: July 11, 2019, 05:11:00 pm »

Moire's easily removed by very slight defocussing. It isn't even necessary to remove it all since you are adding all the pixels within a subset of each patch. You have to work pretty hard before moire, even if visible, alters the average patch color more than a fraction of a dE.

I don't agree with that. Yes, with slight defocusing you avoid high resolution moire, but unless you're capturing the monitor at a very low res (i.e: just filling a small rectangle in the frame), you'll have other moires at different resolutions. Even focusing way behind or ahead the monitor, and even with a 16MP camera you have some sort of interference.
Maybe moire is not the right word, but you won't get a 100% clean image like if it was a print.


You are correct that hue changes are an issues when off 90 degrees. There are also L* differences across the face of a monitor and they can be pretty significant. In fact they can vary even more than hue over just a fewn inches.

OLED monitor can be a solution with that. I've calibrated several SONY A250 for some big cinema studios but couldn't try this experiment.

Even with high end monitors. But that is overcome effectively by using a large, randomized, patch set and is helped significantly by the relatively (compared to printers) independent response monitors tend to have. I did some work on this as a possible approach to profiling monitors (not cameras) . It's quite easy and produces excellent profiles except that it has one, major, problem. Camera response is not intrinsically colorimetric so you also have to have a spectroradiometer to measure the primaries and set the white point. And if you have that you can just use it to create monitor profiles. So it was just an interesting experiment.

Very interesting. That sounds similar to the characterizaction of the embedded EIZO colorimeter, but handmade.

But all that aside, even with a "perfect" monitor one can't produce good camera profiles this way unless one wants to take pictures of images from the same monitor. Might be something an academic might do but there is no reason anyone else would that I can imagine.

Also, your printer charts should do quite well for repro work where your images are printed materials. At least those printed with CYMK processes.  I've used a similar approach here to profile my scanner. I wound up using around 2k of randomized patches. And the profiles worked quite well across my three printers. Better than a CC, IT8, or CCSG scanner profile.
That statement makes a lot of sense.

But the bigger problem with scanners is something called large area crosstalk that is intrinsic to most all reflection scanner designs.
I've never heard about that. Is that related to film, flatbed, virtual drum? Can you elaborate, please?

It can be a much larger problem than metameric error and, like metameric error, can't be solved by any profiling technique. Luckily, it doesn't affect transmission scans nor is it a significant issue for cameras.

Careful not to conflate accuracy with what looks good. Accuracy is meaningful only for scene referred images where one can actually used dE stats. And generally, colorimetrically accurate prints are only useful for reproduction work where the goal is to make a copy indistinguishable from the original when viewed side by side. The same process will produce rather unappealing prints in almost all other uses.

I don't agree with 'colorimetrically accurate prints are only useful for reproduction work'. There are many other fields where accuracy is very important, like biology, product & fashion photography, commercials...

As another aside, when I see PR stuff like claims of "Absolute Accuracy" from a camera profile I chalk the rest of the claims to marketing blather. Camera CFA responses are not linearly congruent to L/I or even very close.

I think you're taking the term absolute too literally. I guess what he means with absolute is as good as is close to perfect.
Logged
Hugo Rodriguez

Coloratti member | PhaseOne certified professional | BenQ Ambassador

hurodal

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 54
    • My website
Re: New color chart for high quality profiling in captures
« Reply #57 on: July 11, 2019, 05:13:05 pm »

Well. it would be interesting to test whether C1 CH's 'absolute accuracy' is indeed so by capturing Hugo's target with it :)

Yes, that would be a good test, but C1 CH is a completely different story and as far as I know, cannot be downloaded. And it can only be used with their own cameras.
Logged
Hugo Rodriguez

Coloratti member | PhaseOne certified professional | BenQ Ambassador

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: New color chart for high quality profiling in captures
« Reply #58 on: July 11, 2019, 08:38:46 pm »

I don't agree with that. Yes, with slight defocusing you avoid high resolution moire, but unless you're capturing the monitor at a very low res (i.e: just filling a small rectangle in the frame), you'll have other moires at different resolutions. Even focusing way behind or ahead the monitor, and even with a 16MP camera you have some sort of interference.
It might surprise you but even highly visible moire, with the exception of long wavelength moire, has little to no affect on the patch averages. The physics demands it. Even in the case of longer wavelength moire, using a Hann window is effective at measuring the light consistently. One just has to keep the moire at relatively high frequencies. Say, a wavelength of 10 pixels or less for a reasonably high re monitor. It's usually much smaller than that.
Quote

I've never heard about that. Is that related to film, flatbed, virtual drum? Can you elaborate, please?
Flatbed. Large area crosstalk is intrinsic to the design of virtually all of them.
Quote

I don't agree with 'colorimetrically accurate prints are only useful for reproduction work'. There are many other fields where accuracy is very important, like biology, product & fashion photography, commercials...
There are a few areas aside from repro that use scene referred profiles. Biology and scientific work, yes. Product and fashion photography, not so common.  By definition output referred profiles distort an image, typically increasing contrast in midrange while decreasing it at the high and low Ls. Scene referred profiles also have the intrinsic limitation of the output medium's dynamic range. And people are used to seeing photos that are made that way. To the point that a colorimetrically accurate print will look unattractive next to a standard print made using an output referred process. Just look around and see just how little scene referred profiles are used.
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #59 on: July 11, 2019, 08:42:03 pm »

There are a few areas aside from repro that use scene referred profiles. Biology and scientific work, yes. Product and fashion photography, not so common.  By definition output referred profiles distort an image, typically increasing contrast in midrange while decreasing it at the high and low Ls. Scene referred profiles also have the intrinsic limitation of the output medium's dynamic range. And people are used to seeing photos that are made that way. To the point that a colorimetrically accurate print will look unattractive next to a standard print made using an output referred process. Just look around and see just how little scene referred profiles are used.
Absolutely agree on all points. Hugo might wish to see what two of us on the ICC photography committee had to say on this in a white paper aimed at photographers:
http://www.color.org/ICC_white_paper_20_Digital_photography_color_management_basics.pdf
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".
Pages: 1 2 [3] 4 5 6   Go Up