Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
02-06-2020, 01:19 AM   #1
Pentaxian




Join Date: Feb 2015
Photos: Gallery
Posts: 12,177
Adobe RGB controversial?

Looking through old JPEG from when I started with the K200D, I've found some sunset images out of gamut.

Out of sRGB gamut? How was it possible? Simple, in camera saturation was too much and there is no way to check for out of gamut with the built in camera JPEG conversion, and at that time I wasn't well ware about what color space really meant (and now it is not possible to recover the out of gamut from JPEGs).

At the time, should I have used the Adobe RGB color space instead of sRGB? Not so sure about that, because it seems Adobe RGB JPEG (8bits) comes with a loss of tone resolution for everything inside the gamut. It seems the 16bits depth of ProPhoto RGB addresses that limitation.

Not only Adobe RGB is still 8bits deep per channel, but the Adobe color gamut isn't that much larger than sRGB, saturation still can easily clip pure reds in images. It seems sRGB is the standard requested by all photo development/printing labs, although recent printers are able to exceed the color gamut of Adobe RGB.

Was Adobe RGB a pseudo standard that came out of the blue from Adobe for the sake of strengthening their business?

02-06-2020, 01:55 AM - 2 Likes   #2
Veteran Member
Silent Street's Avatar

Join Date: Nov 2011
Location: Castlemaine, Victoria, AUS
Photos: Albums
Posts: 1,149
When scanning transparencies, ProPhoto or ColourMatch (most common for this specific task) is profiled, right through to the output device (screen, printer, be it RA-4 or giclee). In some unique cases, combometric profiling (custom profiles that are blend of typically two, sometimes three) are used to gamut expansion. All of this also assumes rigorous attention and quality control of device profiling from source to end-point (monitors, especially).

As things stand, sRGB and AdobeRGB are provided for convenience and as defacto lazy industry standards, often specified for file prep and output. There is no obligation to continuously use either or both with the limitations known when others do the job better. 16bit depth is fine if the printer is capable of it, and certainly not with JPEG (tif or png).

I understand the ICC (International Color Consortium) is revising the sRGB profile. Their page on Profiles provides a lot of information and publications and is worth reading.
02-06-2020, 03:12 AM - 1 Like   #3
Digitiser of Film
Loyal Site Supporter
BigMackCam's Avatar

Join Date: Mar 2010
Location: North East of England
Posts: 20,574
QuoteOriginally posted by biz-engineer Quote
Looking through old JPEG from when I started with the K200D, I've found some sunset images out of gamut.

Out of sRGB gamut? How was it possible? Simple, in camera saturation was too much and there is no way to check for out of gamut with the built in camera JPEG conversion, and at that time I wasn't well ware about what color space really meant (and now it is not possible to recover the out of gamut from JPEGs).
How have you determined there are "out of gamut" colours in your JPEG photos? Presumably, it's by loading them into a photo editing suite and toggling the out of gamut warnings? Which leads me to ask, what output profile have you selected within the software?

So far as I understand - and I'm sure someone will correct me if I'm mistaken - the out of gamut warnings are telling us that there are colours in the currently loaded image that aren't covered by the output gamut; so, if your photo uses the AdobeRGB colour space, you load it into your photo editing software and the output profile is set to sRGB, there's a very strong possibility that areas of the photo will show as out of gamut. But if you set the output profile to AdobeRGB, none of the photo's colours should be out of gamut.

QuoteOriginally posted by biz-engineer Quote
At the time, should I have used the Adobe RGB color space instead of sRGB? Not so sure about that, because it seems Adobe RGB JPEG (8bits) comes with a loss of tone resolution for everything inside the gamut. It seems the 16bits depth of ProPhoto RGB addresses that limitation.
Arguably, if you're intending to post process files for a variety of existing and future output media, you should shoot using the widest available gamut - on the basis that you can always convert to a smaller gamut after the fact. In reality, sRGB is the de facto standard for on-screen viewing, and most mass-market commercial printers either expect - or will at least accept - sRGB files. I would guess professional printing services accept wider gamut images.

But, really, you should have been shooting raw rather than JPEG - in which case, colour space doesn't matter... the limitations are those of the sensor and your exposure settings. Then, when processing your raw files, you can select the output colour space as required for the target medium or service.

QuoteOriginally posted by biz-engineer Quote
Not only Adobe RGB is still 8bits deep per channel, but the Adobe color gamut isn't that much larger than sRGB, saturation still can easily clip pure reds in images. It seems sRGB is the standard requested by all photo development/printing labs, although recent printers are able to exceed the color gamut of Adobe RGB.
AdobeRGB is is approximately 35% larger, which isn't insignificant. But whether this is even represented in a JPEG file using the AdobeRGB colour space, I honestly don't know.

QuoteOriginally posted by biz-engineer Quote
Was Adobe RGB a pseudo standard that came out of the blue from Adobe for the sake of strengthening their business?
Not sure, but I'm confident someone here will know

Last edited by BigMackCam; 02-06-2020 at 04:03 AM.
02-06-2020, 03:16 AM - 2 Likes   #4
Moderator
Loyal Site Supporter
Loyal Site Supporter
pschlute's Avatar

Join Date: Mar 2007
Location: Surrey, UK
Photos: Gallery
Posts: 8,115
QuoteOriginally posted by biz-engineer Quote
because it seems Adobe RGB JPEG (8bits) comes with a loss of tone resolution for everything inside the gamut. It seems the 16bits depth of ProPhoto RGB addresses that limitation
I am no expert in this field (minefield !) but my understanding was that anytime you process an image to jpeg it becomes 8-bit. Irrespective of colour space. That is the nature of the jpeg format.

If i do extensive editing I will usually create a 16-bit TIFF file from the raw file. This is in Adobe RGB colour space, but I could use proPhoto if I wished. I then convert to sRGB colour space before saving as jpeg. Either way it is the creation of the jpeg file that changes the image from 16 to 8 bit.

02-06-2020, 05:56 AM   #5
Pentaxian
Prohibitiory's Avatar

Join Date: Jan 2020
Location: Buckinghamshire, UK
Posts: 471
QuoteOriginally posted by pschlute Quote
I am no expert in this field (minefield !) but my understanding was that anytime you process an image to jpeg it becomes 8-bit. Irrespective of colour space. That is the nature of the jpeg format.

If i do extensive editing I will usually create a 16-bit TIFF file from the raw file. This is in Adobe RGB colour space, but I could use proPhoto if I wished. I then convert to sRGB colour space before saving as jpeg. Either way it is the creation of the jpeg file that changes the image from 16 to 8 bit.
thank you for good explanation!
02-06-2020, 06:10 AM   #6
Pentaxian




Join Date: Feb 2015
Photos: Gallery
Posts: 12,177
Original Poster
QuoteOriginally posted by BigMackCam Quote
How have you determined there are "out of gamut" colours in your JPEG photos?
Yes. It's not all photos, only some photos containing red apples, oranges (the fruit) using the bright picture style of Pentax, sunset at its max level of red/orange using CTE white balance.

QuoteOriginally posted by BigMackCam Quote
Presumably, it's by loading them into a photo editing suite and toggling the out of gamut warnings?
yes

QuoteOriginally posted by BigMackCam Quote
Which leads me to ask, what output profile have you selected within the software?
sRGB, as described in the exif metadata.

QuoteOriginally posted by BigMackCam Quote
if your photo uses the AdobeRGB colour space, you load it into your photo editing software and the output profile is set to sRGB, there's a very strong possibility that areas of the photo will show as out of gamut. But if you set the output profile to AdobeRGB, none of the photo's colours should be out of gamut.
Yes and no. JPEG sRGB file (sRGB in exift), displayed as sRGB with color manager set to monitor calibration profile shows out of gamut areas. I compared the monitor color volume to sRGB official volume using ICCview free tool (Home - ICCView, I found interesting stuff here: Blog - The Ninth), and it happens that my monitor color gamut is pretty good relative to sRGB, it covers slightly more than sRGB (unless the monitor.icc file is acutally the reverse of the color volume coverage by the monitor?). Of course if I set output color space setting to Adobe RGB (input space is sRGB from the exif of the JPEG), the out of gamut warnings are gone, I understand that. Now if my JPEG is mapped to sRGB and my tool set to sRGB output color space there should not be out of gamut areas, unless the out of color gamut tool has a conservatif warning threshold.

Anyway, I can see with my own eyes that if I decrease saturation on a sunset from 100% to 90%, the banding on oranges is gone, letting place to a smooth orange tone gradation on my display. So I guess the saturation of Pentax style color is too much for certain images already rich in colors coded within the sRGB space.
02-06-2020, 07:54 AM   #7
Site Supporter
Site Supporter
JensE's Avatar

Join Date: Dec 2015
Location: Leipzig
Photos: Gallery | Albums
Posts: 1,942
QuoteOriginally posted by biz-engineer Quote
Anyway, I can see with my own eyes that if I decrease saturation on a sunset from 100% to 90%, the banding on oranges is gone, letting place to a smooth orange tone gradation on my display. So I guess the saturation of Pentax style color is too much for certain images already rich in colors coded within the sRGB space.
There can be a lot at play to cause banding. Monitors/graphics cards/software are traditionally limited to 8bits/color channel. You need a capable combination of all three to get to 10bits/channel. That 8bit limitation easily alone leads to banding in bright areas, even when working with full-depth raw data, e.g. 14bit on the K-5/K-3/K-1. The exact way colors are calibrated (graphics card LUT use) can also add to the problems ...

There is a fundamental limitation though: The closer you get to full brightness (255,255,255) in RGB triples (8 bit in this example), the smaller the resolution of colors. Your greatest palette is in the mid-tones. Approaching black (0,0,0), you available color values shrink again. This is inherent in the RGB representation of color. Your choice of color space (sRGB/ AdobeRGB) influences, how those limited number of values are getting used: sRGB has a bit finer differentiation at the price of limited Gamut, AdobeRGB is more 'coarse' but has slightly more saturated primaries, giving you more space in the saturated mid-tones. Either way, if you want any of finely differentiated colors, or smooth gradients or intense saturation, you need to stay somewhat in the middle tones in terms of brightness.

02-06-2020, 08:44 AM   #8
Pentaxian




Join Date: Feb 2015
Photos: Gallery
Posts: 12,177
Original Poster
QuoteOriginally posted by JensE Quote
There is a fundamental limitation though: The closer you get to full brightness (255,255,255) in RGB triples (8 bit in this example), the smaller the resolution of colors.
I tested this hypothesis. A change of level doesn't affect the out of gamut warning, nor the banding I see on my display in orange area. Modifying contrast doesn't affect out of gamut warning in the area. Increasing color saturation makes the warning worse, decreasing color saturation makes the warning and banding disappear.
02-06-2020, 09:18 AM   #9
Site Supporter
Site Supporter
JensE's Avatar

Join Date: Dec 2015
Location: Leipzig
Photos: Gallery | Albums
Posts: 1,942
QuoteOriginally posted by biz-engineer Quote
I tested this hypothesis.
Did you start from the raw data? If not, you're already working with quantized data, where reducing saturation makes the steps smaller and thus look smoother. Seeing your example might help to understand what's going on. Just found this article visualizing and explaining those concepts likely better than I did.
02-06-2020, 09:37 AM - 2 Likes   #10
Veteran Member




Join Date: Feb 2016
Posts: 706
QuoteOriginally posted by biz-engineer Quote
Looking through old JPEG from when I started with the K200D, I've found some sunset images out of gamut.

Out of sRGB gamut? How was it possible? Simple, in camera saturation was too much and there is no way to check for out of gamut with the built in camera JPEG conversion, and at that time I wasn't well ware about what color space really meant (and now it is not possible to recover the out of gamut from JPEGs).
It is not possible for a JPEG to be out of gamut (OOG) unless you or your editor modifies it). A JPEG is a baked file either from camera or your editing of a raw or other image. Anything OOG has already been discarded to fit neatly (or not!) into the JPEG container limited to 8 bit. If you are seeing OOG then your image is exceeding the gamut of either monitor or your chosen output profile. The OOG warning in PS and LR is rather buggy and very old and gives no degree of warning just how far OOG you may be.


QuoteQuote:
Not only Adobe RGB is still 8bits deep per channel, but the Adobe color gamut isn't that much larger than sRGB, saturation still can easily clip pure reds in images. It seems sRGB is the standard requested by all photo development/printing labs, although recent printers are able to exceed the color gamut of Adobe RGB.
Adobe RGB has considerably larger gamut than sRGB in certain areas such as greens, blues, oranges etc - many of the colours found in nature. Looking at a 2d representation of colour spaces ommits a lot of info - view in 3D to get a better feel for differences (see attached sRGB sat inside Adobe RGB)

Photo Labs demanding sRGB are doing this for their own convenience rather than follow a colour managed workflow. Printing needs an ICC profile for the paper in use and as you say many can exceed the colour gamut of Adobe RGB

QuoteQuote:
Was Adobe RGB a pseudo standard that came out of the blue from Adobe for the sake of strengthening their business?
Adobe RGB was a happy accident by Adobe engineers. They were copying coordinates relating to SMPTE RGB colour space and a typo was made. They found even with the error the colour space had value
Attached Images
View Picture EXIF
 Photo 
02-06-2020, 09:38 AM   #11
Site Supporter
Site Supporter
JensE's Avatar

Join Date: Dec 2015
Location: Leipzig
Photos: Gallery | Albums
Posts: 1,942
The 'in print' diagrams for a 'High-End Inkjet' in Working Space Comparison: sRGB vs. Adobe RGB 1998 may actually answer your initial question: The AdobeRGB space looks like it is designed to exactly cover the additional saturated green-cyan tones of such a printer (High End back in the time), without getting excessively big and thus make best of the 8-bit RGB value resolution in JPEGs.
02-06-2020, 10:09 AM   #12
Pentaxian




Join Date: Feb 2015
Photos: Gallery
Posts: 12,177
Original Poster
QuoteOriginally posted by JensE Quote
Just found this article visualizing and explaining those concepts likely better than I did.
Yes, that's a good one. I like the summary which says "My advice is to know which colors your image uses, and whether these can benefit from the additional colors afforded by Adobe RGB 1998. Ask yourself: do you really need the richer cyan-green midtones, orange-magenta highlights, or green shadows? Will these colors also be visible in the final print? Will these differences even be noticeable? If you've answered "no" to any of these questions, then you would be better served using sRGB."

QuoteOriginally posted by JensE Quote
Seeing your example might help to understand what's going on.
I used the "granger rainbow" found here : Check your monitor's gamut - DamienSymonds.net

QuoteOriginally posted by TonyW Quote
It is not possible for a JPEG to be out of gamut (OOG) unless you or your editor modifies it).... Anything OOG has already been discarded to fit neatly (or not!) into the JPEG container limited to 8 bit.
That was what I thought initially, that's why when I see out of gamut area on the JPEG , I'm puzzled. I know for high light and shadow warning there is a margin that can be set (less then 255 and more than 0 thresholds for the warnings), but I didn't find anything about how what the OOG threshold is.


QuoteOriginally posted by TonyW Quote
If you are seeing OOG then your image is exceeding the gamut of either monitor or your chosen output profile.
If I enable color management with monitor profile or sRGB profile, the OOG warning is the same.
If I enable color management for the second monitor and move the window to the second monitor (second monitor gamut volume is 60% of sRGB, pretty bad), the warning is the same.

If I disable color management, the warning is the same. The only way I can reduce the warning is by decreasing the saturation slider from 1 to 0.9 (the scale of the slider in SP Dev 9 is from 0 to 3).

---------- Post added 06-02-20 at 18:10 ----------

QuoteOriginally posted by TonyW Quote
Looking at a 2d representation of colour spaces ommits a lot of info
Yes you are right. At first I looked at 2D, the difference between sRGB and aRGB is small , but in 3D the difference is more significant.
02-06-2020, 11:25 AM - 1 Like   #13
Otis Memorial Pentaxian
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 42,007
QuoteOriginally posted by biz-engineer Quote
Adobe RGB controversial?
Nope...

Crummy results with gamut coercion are partially caused by crummy software and partially caused by shooting JPEG in one colorspace and attempting coercion to a more restricted gamut having only eight bits to work with. Lightroom's soft-proof feature may be helpful in that it will show gamut issues for both the current display and the target colorspace. It also provides option for rendering intent (important).

Shooting RAW makes this so much easier since the capture data is full 14-bit with no colorspace assigned. LR's Develop module operates in broad gamut ProPhoto RGB (Melissa variant)* and soft proofing makes gamut coercion to sRGB or whatever much more predictable.

BTW...bit depth and colorspace are fully independent concepts.


Steve

* The rest of LR uses Adobe RGB.

Last edited by stevebrot; 02-06-2020 at 11:32 AM.
02-06-2020, 11:34 AM   #14
Pentaxian




Join Date: Feb 2015
Photos: Gallery
Posts: 12,177
Original Poster
QuoteOriginally posted by stevebrot Quote
Crummy results with gamut coercion are partially caused by crummy software and partially caused by shooting JPEG in one colorspace and attempting coercion to a more restricted gamut having only eight bits to work with.
QuoteOriginally posted by stevebrot Quote
Shooting RAW makes this so much easier since the capture data is full 14-bit with no colorspace assigned.
For sure. It's never too late to learn (and from issues and mistakes). I didn't like Adobe pricing, and still not like it. I'm trying to work around the software I already paid by trying to understand what's going on, because sometimes understanding avoids expenses.
02-06-2020, 11:18 PM   #15
Pentaxian




Join Date: Feb 2015
Photos: Gallery
Posts: 12,177
Original Poster
QuoteOriginally posted by stevebrot Quote
Lightroom's soft-proof feature may be helpful in that it will show gamut issues for both the current display and the target colorspace. It also provides option for rendering intent (important).
Actually, by researching again the documentation of Silkypix Dev 9, I found that all this is also possible, I just didn't know it was in the print module since I don't have a printer at home (not printing directly from SPD9).
The contributions here helped me to refine my search. The "out of gamut warning" in SPD9 also warn when parts of the image color is at the edge of the color space. When there is no OCG warning, all of the image colors are inside the target color space. The tolerance between warning and no warning is about one to two step of the saturation slider.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
adobe, adobe rgb, camera, color, colors, default, display, files, gamut, icc, image, intent, jpeg, lab, management, monitor, paper, photography, photoshop, printer, profile, prophoto, rgb, space, srgb
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
sRGB vs. Adobe RGB viewfinder Troubleshooting and Beginner Help 26 08-26-2020 12:22 PM
Question Adobe RGB in albums / gallery jido Site Suggestions and Help 7 12-15-2018 05:33 PM
Phillips Adobe RGB monitor- cheap AutumnRowan Digital Processing, Software, and Printing 14 07-28-2016 06:10 PM
Regarding Adobe RGB GoremanX Pentax DSLR Discussion 105 02-07-2010 05:04 PM
Adobe-RGB -vs- sRGB Ed in GA Photographic Technique 8 01-26-2007 04:08 PM



All times are GMT -7. The time now is 04:21 PM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top