Pages: [1]   Go Down

Author Topic: Struck me as strange ...  (Read 5160 times)

Rand47

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1882
Struck me as strange ...
« on: December 25, 2012, 06:41:24 pm »

This comment from another web site - discussing color space preferences:

"Capturing in Adobe RGB and then converting to sRGB to print loses colors that would have been there if the original had been captured in sRGB in the first place."

Comments?  I don't know enough about the intricacies of these things, but this struck me as odd & counterintuitive.
Logged
Rand Scott Adams

Czornyj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1948
    • zarzadzaniebarwa.pl
Re: Struck me as strange ...
« Reply #1 on: December 25, 2012, 06:58:40 pm »

It loses tones rather than colors, like any other color conversion. The question is - why would anyone pre-convert to sRGB "to print"? Unless the controller of the minilab is not color managed and expects sRGB numbers it's a complete nonsense.
Logged
Marcin Kałuża | [URL=http://zarzadzaniebarwa

Rand47

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1882
Re: Struck me as strange ...
« Reply #2 on: December 25, 2012, 07:17:13 pm »

It loses tones rather than colors, like any other color conversion. The question is - why would anyone pre-convert to sRGB "to print"? Unless the controller of the minilab is not color managed and expects sRGB numbers it's a complete nonsense.

Right.
Logged
Rand Scott Adams

Tony Jay

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2965
Re: Struck me as strange ...
« Reply #3 on: December 25, 2012, 07:30:02 pm »

This comment from another web site - discussing color space preferences:

"Capturing in Adobe RGB and then converting to sRGB to print loses colors that would have been there if the original had been captured in sRGB in the first place."

Comments?  I don't know enough about the intricacies of these things, but this struck me as odd & counterintuitive.

I don't know if there is a broader context to the statement but as it stands I can't agree with this.
I can certainly see an image shot and processed as an sRGB, and an identical image shot and processed as AdobeRGB and then converted to sRGB, looking slightly different.
However, the key thing is that the gamut of colours incorporated by the sRGB colourspace is fixed and so the colours in each image, sRGB straight versus AdobeRGB converted to sRGB, can still be indentical.
To say that colours are 'lost', in the context mentioned in the quote, is a distortion.

Also as mentioned, the reason for converting to sRGB is important. If it is for printing purposes then that is a nonsense. Modern photographic quality printers either approximate or exceed the AdobeRGB colourspace. If a Minilab wants images in the sRGB colourspace then find another Minilab!
sRGB is for digital projection, not for printing.

Hopefully Andrew Rodney or one of the other colour management gurus will lend an opinion.
Also it may be an idea to provide a link to the article so we can read it in its fullest context.

Tony Jay
« Last Edit: December 25, 2012, 07:42:48 pm by Tony Jay »
Logged

Rand47

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1882
Re: Struck me as strange ...
« Reply #4 on: December 25, 2012, 11:11:08 pm »

I don't know if there is a broader context to the statement but as it stands I can't agree with this.
I can certainly see an image shot and processed as an sRGB, and an identical image shot and processed as AdobeRGB and then converted to sRGB, looking slightly different.
However, the key thing is that the gamut of colours incorporated by the sRGB colourspace is fixed and so the colours in each image, sRGB straight versus AdobeRGB converted to sRGB, can still be indentical.
To say that colours are 'lost', in the context mentioned in the quote, is a distortion.

Also as mentioned, the reason for converting to sRGB is important. If it is for printing purposes then that is a nonsense. Modern photographic quality printers either approximate or exceed the AdobeRGB colourspace. If a Minilab wants images in the sRGB colourspace then find another Minilab!
sRGB is for digital projection, not for printing.

Hopefully Andrew Rodney or one of the other colour management gurus will lend an opinion.
Also it may be an idea to provide a link to the article so we can read it in its fullest context.

Tony Jay

Tony,

Thank you.  Your perspective comports well with what I've come to understand through my reading, watching tutorials, and experience. 

Rand
Logged
Rand Scott Adams

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Re: Struck me as strange ...
« Reply #5 on: December 25, 2012, 11:34:09 pm »

"Capturing in Adobe RGB and then converting to sRGB to print loses colors that would have been there if the original had been captured in sRGB in the first place."

Bullshyte...what web site? Who said it?
Logged

Rhossydd

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3368
    • http://www.paulholman.com
Re: Struck me as strange ...
« Reply #6 on: December 26, 2012, 04:54:22 am »

why would anyone pre-convert to sRGB "to print"? Unless the controller of the minilab is not color managed and expects sRGB numbers it's a complete nonsense.
Lots of printing labs expect sRGB format, as do most .Print on demand. book printing services.
Whether you'd see any difference between originating in AdobeRGB or sRGB with this type of service is unlikely though.
Logged

Czornyj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1948
    • zarzadzaniebarwa.pl
Re: Struck me as strange ...
« Reply #7 on: December 26, 2012, 05:34:41 am »

Whether you'd see any difference between originating in AdobeRGB or sRGB with this type of service is unlikely though.

I saw significant difference in saturated photos printed on digital presses, traditional and dry minilabs. Here's Kodak Endura Premier profiled in Frontier 350's PD mode:


From latest Photokina I also brought a few super-vivid test images printed on new Canon DreamLabo 5000 (7 dye based inks)  - sRGB would just ruin the effect.
« Last Edit: December 26, 2012, 05:39:40 am by Czornyj »
Logged
Marcin Kałuża | [URL=http://zarzadzaniebarwa

Rhossydd

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3368
    • http://www.paulholman.com
Re: Struck me as strange ...
« Reply #8 on: December 26, 2012, 05:51:46 am »

I saw significant difference in saturated photos printed on digital presses, traditional and dry minilabs. Here's Kodak Endura Premier profiled in Frontier 350's PD mode:
...
From latest Photokina I also brought a few super-vivid test images printed on new Canon DreamLabo 5000 (7 dye based inks)  - sRGB would just ruin the effect.
You're talking about the differences between printing wide gamut images and sRGB, that goes without saying.
The OP's quote is suggesting that images captured in sRGB contain more colours than images captured in AdobeRGB space then converted to sRGB. If there is a difference, I'd be surprised if it would be noticeable.
Logged

Tony Jay

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2965
Re: Struck me as strange ...
« Reply #9 on: December 26, 2012, 06:12:40 am »

You're talking about the differences between printing wide gamut images and sRGB, that goes without saying.
The OP's quote is suggesting that images captured in sRGB contain more colours than images captured in AdobeRGB space then converted to sRGB. If there is a difference, I'd be surprised if it would be noticeable.

I agree, but the, as yet unsubstantiated, article that was quoted suggested there was.

As for the printing issue, that arose, it seems partly, as speculation about why there was this fuss about sRGB and conversion to sRGB in the first place. If it was for the purpose of printing some of us suggested that sRGB-based printing seems redundant, if still widespread.

Tony Jay
Logged

Rhossydd

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3368
    • http://www.paulholman.com
Re: Struck me as strange ...
« Reply #10 on: December 26, 2012, 06:26:50 am »

article that was quoted
<Being pedantic> There's no source for this 'quote' and it's quite possible it's just a simple typo.
Logged

Tony Jay

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2965
Re: Struck me as strange ...
« Reply #11 on: December 26, 2012, 06:33:21 am »

<Being pedantic> There's no source for this 'quote' and it's quite possible it's just a simple typo.

No argument there.
If you look back to my first post in the thread I ask for the link to the article to be provided so that we could all read for ourselves what was really meant.

Tony Jay
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20614
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Struck me as strange ...
« Reply #12 on: December 26, 2012, 11:36:08 am »

Without seeing the full context of the post in question, let's just say it doesn't sound like it washes.

A 24 bit document in sRGB and a 24 bit document in Adobe RGB have the same encoding. The author may believe that with a smaller color gamut (sRGB), there are more bits to define 'colors' in a smaller gamut than a larger gamut (AdobeRGB)?

A 24 bit image in either sRGB or Adobe RGB (1998) of a gray card will have less colors than an image of a colorful field of flowers. Scene gamut should not be ignored! Gamut and encoding of data should not be ignored! We can talk about color encoding where the math results in billions of color despite the number of colors we can see are nothing approaching this value.

If I take a Granger Rainbow I built in Lab, high bit, convert one to sRGB and one to sRGB FROM Adobe RGB (1998) in Photoshop, using V2 profiles and no dither, they both have the same number of unique colors according to ColorThink! So I think the original author is discussing something else or is just wrong.

The other issue is, we often have to convert color spaces. The only reason I can think of converting from Adobe RGB to sRGB is if I needed sRGB! For the web or other emissive non color managed use. There is no such thing as an sRGB printer (or Adobe RGB printer for that matter). While some silly labs may demand sRGB it is very easy to prove such devices do not follow anything like sRGB in their behavior. Even if someone forced me to send sRGB for output to a printer, I'd still want to start with Adobe RGB (1998) for other output needs.

 Actually I'd prefer ProPhoto but that's a story summed up in this long video:

Everything you thought you wanted to know about color gamut

A pretty exhaustive 37 minute video examining the color gamut of RGB working spaces, images and output color spaces. All plotted in 2D and 3D to illustrate color gamut.

High resolution:http: //digitaldog.net/files/ColorGamut.mov
Low Res (YouTube): http://www.youtube.com/watch?v=n0bxSD-Xx-Q
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Rand47

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1882
Re: Struck me as strange ...
« Reply #13 on: December 28, 2012, 02:38:38 pm »

Without seeing the full context of the post in question, let's just say it doesn't sound like it washes.

A 24 bit document in sRGB and a 24 bit document in Adobe RGB have the same encoding. The author may believe that with a smaller color gamut (sRGB), there are more bits to define 'colors' in a smaller gamut than a larger gamut (AdobeRGB)?

A 24 bit image in either sRGB or Adobe RGB (1998) of a gray card will have less colors than an image of a colorful field of flowers. Scene gamut should not be ignored! Gamut and encoding of data should not be ignored! We can talk about color encoding where the math results in billions of color despite the number of colors we can see are nothing approaching this value.

If I take a Granger Rainbow I built in Lab, high bit, convert one to sRGB and one to sRGB FROM Adobe RGB (1998) in Photoshop, using V2 profiles and no dither, they both have the same number of unique colors according to ColorThink! So I think the original author is discussing something else or is just wrong.

The other issue is, we often have to convert color spaces. The only reason I can think of converting from Adobe RGB to sRGB is if I needed sRGB! For the web or other emissive non color managed use. There is no such thing as an sRGB printer (or Adobe RGB printer for that matter). While some silly labs may demand sRGB it is very easy to prove such devices do not follow anything like sRGB in their behavior. Even if someone forced me to send sRGB for output to a printer, I'd still want to start with Adobe RGB (1998) for other output needs.

 Actually I'd prefer ProPhoto but that's a story summed up in this long video:

Everything you thought you wanted to know about color gamut

A pretty exhaustive 37 minute video examining the color gamut of RGB working spaces, images and output color spaces. All plotted in 2D and 3D to illustrate color gamut.

High resolution:http: //digitaldog.net/files/ColorGamut.mov
Low Res (YouTube): http://www.youtube.com/watch?v=n0bxSD-Xx-Q

Gentlemen, thanks for all the feedback.  I did not intend to cause a rukus!  And it wasn't an "article" but a post / thread on color spaces posted on the notorious web site for ignorance on all things photographic that starts with DP.  I'm somewhat proficient with LR and have watched the LULA tutorials, From Camera to Print and Screen, and the Advanced LR tutorials, and "thought" I had a good handle from Jeff Schewe on colorspaces and what they're for etc., and what LR uses as a default and why that's a good thing.  

In trying to be "helpful" (which for the most part was pearls before swine) on that web site, there was one fellow who seemed a little more advanced (Ryan) who was responding in a way that put me off and made me wonder if I was missing something.  Hence the quote above.  I'll post the link to that thread here:  http://forums.dpreview.com/forums/thread/3356941   in case any of you would like to weigh in - though it is probably a waste of your time and effort, I suspect.  What I would be interested in, however is your thoughts on one response I received - for my own education, if I'm missing something relevant.

The post I'm referring to is below: (I'm alsoRand47 in the post below)

Quote
Rand 47 wrote:


Ryan Williams wrote:

It still matters as you can't have an image without it existing in some kind of colour space, however the RAW itself contains no such info — only once it's brought into a RAW processor will it be assigned a profile, so in Lightroom that'd be sRGB.

Other processors may vary in what profile they use, and it's even possible the X-mount cameras could embed a meta/EXIF into the RAW file which tells the RAW processor which profile to use if it supports that option, though I doubt this is the case.

If you are saying that the native colors space for LR is sRGB, that's incorrect. LR uses ProPhoto RGB when working with RAW files. You may, of course export w/ whatever color space your image needs for its intended display mode.

A good reason to work in a large color space with RAW files is that you retain flexibility and don't throw away color information that you can never get back. Your display needs may not be able to display it now, but who knows what the future holds. The Apple retina display is close to Adobe RGB now, so sRGB is already an obsolete and too small color space for retina devices. And if you work in sRGB and someone ends up wanting a nice print of your image, well it won't be as good as it could have been, potentially, because you trashed color information that current printers can print.

Best,

Rand

No, I said for all intents and purposes it's sRGB. To be precise, it uses Melissa RGB which is a modified version of ProPhoto RGB that has the same tonality as sRGB.

Something that I'm not sure everyone understands is that colour profiles go well beyond simply allowing more colours — they can completely change the tonality and colour signature of a photo. For example, all of the X-series film simulations are colour profiles that're baked directly into the JPEGs. So using different colour profiles can have a DRAMATIC impact on how your photo looks.

In a perfect world this wouldn't be an issue because profiles are meant to embedded into files and then viewing software would show the photos using that profile. But this takes us back to my previous post: much software ignores colour profiles altogether and shows in sRGB. This includes many web browsers — the very software most of our online audiences will be using to view our photos.

So this brings me back to my original point: use sRGB or you risk people seeing them with wildly different tonality to what you intended. A wider gamut is only really useful for printing as the existence of things like 'retina displays' is completely negated by the fact that much software won't use the profiles anyway.

So what do you think?  I don't want to be a bother here, as I suspect this discussion is below a threshold of interest for most - but I am interested in always being open to learning.  I've learned so much from Jeff Schewe and Michael and all of you here who are genuinely knowledgable and generous with that knowledge.  So let me apologize again for a tempest in a tea-pot, and express my appreciation for your help.

Best regards,
Rand
« Last Edit: December 28, 2012, 02:58:42 pm by Rand47 »
Logged
Rand Scott Adams

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20614
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Struck me as strange ...
« Reply #14 on: December 28, 2012, 02:48:13 pm »

About 50% +/- of the posts are correct and the rest isn't worth commenting on. Anytime a post on color management starts with Ken Rockwell, time to move on, you're dealing with flat earth viewpoints at some point!

Quote
Lightroom technically uses Melissa RGB but this is for all intents and purposes sRGB.

This fellow is very confused...
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Re: Struck me as strange ...
« Reply #15 on: December 28, 2012, 02:57:33 pm »

So what do you think?

Considering the OP of the thread mentions Ken Rockwell, I think the whole thread is a waste of time...

Logged

Rand47

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1882
Re: Struck me as strange ...
« Reply #16 on: December 28, 2012, 03:01:36 pm »

Considering the OP of the thread mentions Ken Rockwell, I think the whole thread is a waste of time...



Jeff,

Yes, and that's why I am somewhat embarassed to even have you see it!  I was trying to be "helpful" over there (I should have known better).

Rand
Logged
Rand Scott Adams
Pages: [1]   Go Up