Pages: [1] 2 3 ... 6   Go Down

Author Topic: New color chart for high quality profiling in captures  (Read 7983 times)

hurodal

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 54
    • My website
New color chart for high quality profiling in captures
« on: July 09, 2019, 08:07:33 am »

Hello all,

My name is Hugo and I work as a color management professor, technician and consultant at Barcelona, Spain.

I want to introduce you a new color chart I've been slowly working on for the last 4 years. I've been making wide gamut IT8 since 2006 and one of my main interests is achieving a true high fildelity in captures, specially for those areas where highly saturated/difficult colors sometimes appear: commercials, art reproduction and, specially, cloth wear e-commerce.

I designed this chart using my experience making IT8s, taking some ideas frrm other charts and adding new ideas of my own. This is a color chart aimed to produce the very best camera profiling out there, with no compromise.

It's not as easy to use as a matte chart (CC classic, spyderchecker) but with it's 999 color patches it offers really high precision. I've been working with it and at the same time testing it in some really hard scenarios like some of the biggest cloth brands here in Spain: ZARA, Mango, Tendam, Desigual, as well as in the National spanish Biblioteque and other institutions. The results have been impressive so far.

You can see all the information about it here:
HR-1 SuperChroma

I'd love to share and discuss technical details and the like with anyone interested here. I hope you find it interesting.



Best regards,

Hugo
Logged
Hugo Rodriguez

Coloratti member | PhaseOne certified professional | BenQ Ambassador

Czornyj

  • Sr. Member
  • ****
  • Online Online
  • Posts: 1950
    • zarzadzaniebarwa.pl
Re: New color chart for high quality profiling in captures
« Reply #1 on: July 09, 2019, 12:33:47 pm »

Sorry, but you can predict in advance that this is actually a bad idea, which is proved by the samples that you providied.
Logged
Marcin Kałuża | [URL=http://zarzadzaniebarwa

hurodal

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 54
    • My website
Re: New color chart for high quality profiling in captures
« Reply #2 on: July 09, 2019, 01:00:45 pm »

Hi Czornyj,

I wasn't expecting warm compliments in the very first replies but at least some curiosity and questions about the chart, rather than negativity and rejection.
OK, let me know what you already predicted and concluded just by looking at the samples (I guess you also got the reference spectral file and did some calculations and research...).

Regards,

Hugo
« Last Edit: July 09, 2019, 01:39:33 pm by hurodal »
Logged
Hugo Rodriguez

Coloratti member | PhaseOne certified professional | BenQ Ambassador

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20652
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #3 on: July 09, 2019, 02:24:57 pm »

You can see all the information about it here:
HR-1 SuperChroma

I'd love to share and discuss technical details and the like with anyone interested here. I hope you find it interesting.
So can you explain what the before's ("Standard: without profile") implies?
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

MHMG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1285
Re: New color chart for high quality profiling in captures
« Reply #4 on: July 09, 2019, 02:47:07 pm »

Hard to judge just by looking at your photos of your actual target, but it looks as if you are generating the color patches the same way IT8 targets are made.

If so, then that's a deal breaker. The color patches would be made of dyes generated by wet process (RA-4) chemistry. The cyan, magenta, yellow layers are stacked in three layers to form the colors, and there will be lots of color constancy (metameric mismatches) between how the camera records those colors and how the camera records real world spectra. A camera target is a different application all together from the use of IT8 targets to profile scanners for scanning film or color prints. IT8 targets are "training" the scanner to look at the very same dyes which will be scanned during the reproduction, i.e., either wet process color prints or color transparency films. It's why a different IT8 target is ideally required for every film type, definitely Kodachrome versus EktaChrome, because the color dyes are so spectrally different between those two film types.

The 24 patch Macbeth color checker chart uses specially formulated color pigments designed to precisely mimic real world color spectra (e.g., skin tones, blue skies, green foliage, etc), and this technical achievement by C.S. McCamy, et al, back in 1976 is a critical reason why the Macbeth Colorchecker chart functions as well as it does for digital camera profiling in the digital age.
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20652
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #5 on: July 09, 2019, 02:51:28 pm »

Great history of the MacBeth found here:
http://www.babelcolor.com/colorchecker.htm


I agree with Mark about potential issues using a process built for profiling scanners with cameras.
One issue with the MacBeth is only one path falls outside sRGB (Cyan). So we shouldn't expect the gamut of the resulting profile to be any larger but gamut and cameras are problematic for oh so many reasons.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

MHMG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1285
Re: New color chart for high quality profiling in captures
« Reply #6 on: July 09, 2019, 03:03:27 pm »

Great history of the MacBeth found here:
http://www.babelcolor.com/colorchecker.htm


I agree with Mark about potential issues using a process built for profiling scanners with cameras.
One issue with the MacBeth is only one path falls outside sRGB (Cyan). So we shouldn't expect the gamut of the resulting profile to be any larger but gamut and cameras are problematic for oh so many reasons.

I'm no expert in camera profiling (and don't use it much, if truth be told), but my understanding is that the camera profiling programs are using the 24 Macbeth patches to help scale the RGB channels. They aren't clipping nor are they constrained to those values by any means. It would be an epic fail if they did. So, some interpolation and extrapolation is happening in the profiling app to scale colors different than the 24 chosen patches and also with higher chroma than those 24 patches. The camera primaries are also generally well known, so as long as the result generates a good match when photographing the ColorChecker, other interpolate and/or extrapolated values should be more accurately reproduced as well. Maybe not perfect but better than nothing.

I prefer to season all of my color work to taste from the RAW files. I use Adobe ACR, and find the built in  camera calibration flavors are often just as suitable a starting point for my work as the custom camera profiles I have made with apps like ColorChecker Passport, etc.
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20652
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #7 on: July 09, 2019, 03:12:58 pm »

The 24 patch target is just fine to build .DCP profiles which are vastly different than ICC profiles especially with respect to how and where in the processing chain the operation. For ICC camera profiles, it's similar to making a scanner profiles as you outlined. It's 'finger-printing' as best it can, the rendered (output referred) image. So along with too few color patches and a limited target color gamut, GMB (now X-rite) produced two actual 'camera' profile targets. The SG had a wider gamut using glossy patches but it was a bitch and a half to shoot (no reflections please). Then they came up with a newer target without the glossy patches but smaller color gamut. Frankly I found camera profiling for the creation of ICC profiles hugely hit or miss. Gave up on em. And since my raw processor of choice doesn't even allow the use of ICC camera profiles, even better. In very controlled studio like conditions (think copy work), I was able to produce good ICC camera profiles. But shoot outside that condition, not so much.


Since at least you and I use an Adobe raw processor, a custom .DCP profile with the Macbeth gets the job done nicely. KISS.  ;) [size=78%] [/size]
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 #8 on: July 09, 2019, 03:49:58 pm »

So can you explain what the before's ("Standard: without profile") implies?
Hi,
Yes, it means with the generic profile (Capture one's profile in this case).
Logged
Hugo Rodriguez

Coloratti member | PhaseOne certified professional | BenQ Ambassador

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20652
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #9 on: July 09, 2019, 03:51:39 pm »

Hi,
Yes, it means with the generic profile (Capture one's profile in this case).
Begs the question, how would the 'before's' look with another raw converter?
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 #10 on: July 09, 2019, 03:54:10 pm »

Hi Hugo,

Good stuff, I'll probably be a customer.  I am curious on how you chose those specific patches and the kind of improvement you would expect from using this hi-res chart vs a lower resolution one like the CC24.  Where would you expect the most improvements?  Are there any comparisons of its performance vs CC24 once profiled, on standard targets representative of specific situations (like landscape, portraits etc.) all other things being equal?

Thanks!

Jack
PS I kind of agree with others that the before and after images look kind of hokey and are not going to pass muster here: were those done by your marketing department? ;)
Logged

hurodal

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 54
    • My website
Re: New color chart for high quality profiling in captures
« Reply #11 on: July 09, 2019, 04:30:00 pm »

Hard to judge just by looking at your photos of your actual target, but it looks as if you are generating the color patches the same way IT8 targets are made.

If so, then that's a deal breaker. The color patches would be made of dyes generated by wet process (RA-4) chemistry. The cyan, magenta, yellow layers are stacked in three layers to form the colors, and there will be lots of color constancy (metameric mismatches) between how the camera records those colors and how the camera records real world spectra. A camera target is a different application all together from the use of IT8 targets to profile scanners for scanning film or color prints. IT8 targets are "training" the scanner to look at the very same dyes which will be scanned during the reproduction, i.e., either wet process color prints or color transparency films. It's why a different IT8 target is ideally required for every film type, definitely Kodachrome versus EktaChrome, because the color dyes are so spectrally different between those two film types.

The 24 patch Macbeth color checker chart uses specially formulated color pigments designed to precisely mimic real world color spectra (e.g., skin tones, blue skies, green foliage, etc), and this technical achievement by C.S. McCamy, et al, back in 1976 is a critical reason why the Macbeth Colorchecker chart functions as well as it does for digital camera profiling in the digital age.

The chart is printed in an Epson inkjet plotter, not by chemical process. Chemical process is being removed everywhere. For printing color charts, chemical process presents many variables and it's hard to keep constancy. Inkjet is much more adequate, and it offers a wide gamut.

There's not a big difference between how the spectra is mixed in a chemical paper and inkjet print, because in chemical, dyes are layed one over another, while in an inkjet, pigments are placed one beside another.

I've read many times about the argument that the spectra of the real world vs the pigments but for this chart is not an issue, because:
- It's aimed for art reproduction, which uses systhetic pigments, just like inkjet.
- it's aimed for product photography, which (except food) is always printed used synthetic pigments. Cloths, packaging... all use these pigments.

But, anyway, I've tested in real world, landscape and it gives extraordinary results. Ornitologhists, i.e. had told me that photos profiled with the Superchrome are the best they've seen, ever. And birds use natural pigments, of course.

Regarding the CC classic, I know it very well; In fact I have 5-6 in almost all versions.
I've read babelcolor site and I'm also in contact with Danny pascale, the man behind Babelcolor. In fact, he helped me in some stages of the development.

Yes, Macbeth developed pigments that are similar to the real world, back in 1976, which is more than 40 years ago. Since then, there's a company in Japan that had invested zillions of Yens to improve their pigment formulation: Seiko-Epson.
Don't you think that with the money they invested in research (which I guess we all agree that is WAY more than what Macbeth invested 40 yrs ago) they achieved extraordinary pigments with excellent spectra nowadays?

the main problem with CC is that it only has 24 different patches (actually only 18 in color). I've used it for camera profiling and while with skin and some low saturation color works well, with highly saturated or difficult colors, the results are far from 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 #12 on: July 09, 2019, 04:54:59 pm »

Great history of the MacBeth found here:
http://www.babelcolor.com/colorchecker.htm


I agree with Mark about potential issues using a process built for profiling scanners with cameras.
One issue with the MacBeth is only one path falls outside sRGB (Cyan). So we shouldn't expect the gamut of the resulting profile to be any larger but gamut and cameras are problematic for oh so many reasons.

Thanks for the link. I already knew that article.

Well, a limited gamut chart can actually build wider gamut profiles for linear capture devices. The problem is that cameras are not that linear, and there's also another problem: the interpolation between primaries is actually very complicated if the chart has not high resolution.
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 #13 on: July 09, 2019, 05:07:22 pm »

Begs the question, how would the 'before's' look with another raw converter?

Quite different, actually. That's one of the main differences of the several raw converters. ACR, i.e., IMHO renders dirt colors, just by opening the file with no setting at all.
Capture one shines on that: in the same scenario, the colors are quite cleaner.

Not to mention how bad is ACR color engine when you use the sliders of HSL or calibration tab, which causes a lot of strange artifacts and posterization, something that never happens with Capture One.
Thus I only recommend Capture One for the purposes that chart is aimed to.

Please take a look at this photo I've taken, developed with ACR and C1. Both with no special settings, other than a small under exposure. Both with generic profiles (provided by Adobe/Phase one)
Logged
Hugo Rodriguez

Coloratti member | PhaseOne certified professional | BenQ Ambassador

MHMG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1285
Re: New color chart for high quality profiling in captures
« Reply #14 on: July 09, 2019, 05:11:49 pm »

The chart is printed in an Epson inkjet plotter, not by chemical process. Chemical process is being removed everywhere. For printing color charts, chemical process presents many variables and it's hard to keep constancy. Inkjet is much more adequate, and it offers a wide gamut.

There's not a big difference between how the spectra is mixed in a chemical paper and inkjet print, because in chemical, dyes are layed one over another, while in an inkjet, pigments are placed one beside another.

I've read many times about the argument that the spectra of the real world vs the pigments but for this chart is not an issue, because:
- It's aimed for art reproduction, which uses systhetic pigments, just like inkjet.
- it's aimed for product photography, which (except food) is always printed used synthetic pigments. Cloths, packaging... all use these pigments.

But, anyway, I've tested in real world, landscape and it gives extraordinary results. Ornitologhists, i.e. had told me that photos profiled with the Superchrome are the best they've seen, ever. And birds use natural pigments, of course.

Regarding the CC classic, I know it very well; In fact I have 5-6 in almost all versions.
I've read babelcolor site and I'm also in contact with Danny pascale, the man behind Babelcolor. In fact, he helped me in some stages of the development.

Yes, Macbeth developed pigments that are similar to the real world, back in 1976, which is more than 40 years ago. Since then, there's a company in Japan that had invested zillions of Yens to improve their pigment formulation: Seiko-Epson.
Don't you think that with the money they invested in research (which I guess we all agree that is WAY more than what Macbeth invested 40 yrs ago) they achieved extraordinary pigments with excellent spectra nowadays?

the main problem with CC is that it only has 24 different patches (actually only 18 in color). I've used it for camera profiling and while with skin and some low saturation color works well, with highly saturated or difficult colors, the results are far from perfect.

I can easily debate some of the conclusions you have drawn and stated above, but I won't. Good luck with your project. I'm happy you are happy :)
Logged

hurodal

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 54
    • My website
Re: New color chart for high quality profiling in captures
« Reply #15 on: July 09, 2019, 06:01:01 pm »

Hi Hugo,

Good stuff, I'll probably be a customer.  I am curious on how you chose those specific patches and the kind of improvement you would expect from using this hi-res chart vs a lower resolution one like the CC24.  Where would you expect the most improvements?  Are there any comparisons of its performance vs CC24 once profiled, on standard targets representative of specific situations (like landscape, portraits etc.) all other things being equal?

Thanks!

Jack
PS I kind of agree with others that the before and after images look kind of hokey and are not going to pass muster here: were those done by your marketing department? ;)

Thanks Jack!

The approach I've taken for this chart is quite different than any other chart in the market. Most charts uses gradients. They put gradients for colors, going from black/white/grey to the most saturated color. This looks good at first glance, but it causes poor distribution in three areas:
  • near black
  • near white
  • between gradients


The superchroma uses perceptually evenly distributed patches that doesn't follow gradients. this means that there's the same amount of patches almost everywhere inside the gamut -> constant quality through the entire gamut.
Other charts provide good quality near/around the gradients of patches but poor (or very poor) quality far from (or between) them.

Yes, I have lots of examples that I'm preparing for an article to be published here in the main website, with the kind help ot Kevin, main editor of this site.
The biggest improvement I've seen so far was with a turquoise color of a women bag, made with colombian wood. A pro photographer that was doing a job for the customer was in big problems because even doing a good profile with CC24 in ACR the result was way off the path, because it was rendered as blue.
With my wide-gamut IT8 and Capture One it was rendered as green, much closer, although not yet perfect. This photo was the reason I've started to design this new chart.

Please look at the attached examples. One (the blueish one) is the generic profile in ACR. Another is also from ACR but with the custom profile with CC 24. I also attach the color sample of the wood for reference.

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 #16 on: July 09, 2019, 06:12:25 pm »

I can easily debate some of the conclusions you have drawn and stated above, but I won't. Good luck with your project. I'm happy you are happy :)

I'm open to discussion, sir, as long as you talk with respect as I do. As you wish :-)

As we're discussing with our own opinion, let me show some data, that is out of discussion.
You've mentioned blues and greens. Well, I've searched for the closest blue and green to the primaries on the CC24. There are no exact colors that match in the Superchroma but these two are the closest. Beside you can see the spectra information for both samples.
It has to be taken in count that the ones in the HR1 are actually mixed inkjet primaries, but as you can see the spectra curves are both very smooth and look so close to the other.
I don't think metamerism is an issue here.
Logged
Hugo Rodriguez

Coloratti member | PhaseOne certified professional | BenQ Ambassador

elliot_n

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1219
Re: New color chart for high quality profiling in captures
« Reply #17 on: July 09, 2019, 06:15:26 pm »

Interesting product. Thanks for posting.
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20652
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: New color chart for high quality profiling in captures
« Reply #18 on: July 09, 2019, 06:19:32 pm »

Couple points:
1. This comment about linear or non linear is moot. ICC profiles take place on rendered data that might be either, NOT the raw data.
2. There is no such thing as ‘no setting’ in ACR/LR. You might not like the defaults (dirt rendering?) but they can easily be changed and since these products CANNOT use ICC camera profiles, for lots of good reasons, that comment is moot too.
3. Your site shows a before image with a profile. Is this a custom profile you built from a ‘standard’ available target and if so, which target using what software? IF it’s C1’s canned profile, your examples are rather worthless sorry. You can’t compare a canned profile (built using any target) to your custom profile with your target. So hopefully you’ll tell us that the before is a custom profile in C1 using a differing (and inferior) target.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

samueljohnchia

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 498
Re: New color chart for high quality profiling in captures
« Reply #19 on: July 09, 2019, 06:37:55 pm »

I can easily debate some of the conclusions you have drawn and stated above, but I won't. Good luck with your project. I'm happy you are happy :)

Mark, for the sake of other forums members like myself who are curious and eager to learn, please share your insights!
Logged
Pages: [1] 2 3 ... 6   Go Up