p.2 #1 · sRGB vs Adobe RGB...I thought I had it right!
There's litsle merit to the argument that web browsers expect sRGB images. Web browsers don't do proper colour management and won't give you the right colours regardless of what colour space or colour profile you use. The closest to getting it right is Apple Safari. The next Mozilla Firefox will do it a lot better. The rest do not do any colour management at all.
If you want to see the real colours then you have to transfer the posted images to a colour managed photo viewer. If the poster of the image wants you to see it properly then they need to include the colour space, colour profile and monitor calibration data with the images. However, most peple who post or view images on the internet do not have a clue about colour management and so you've got very little chance of seeing images as they were intended. Even the owners probably aren't seeing them correctly.
If you have a decent camera and a high-end monitor then I'm sure that you can see diffeences between adobeRGB and sRGB. Certainly not every colour in every picture will be different but there must be plenty of real life colours that fit into one gamut without fitting into the other. When it comes to printing, that difference can be disguised by the rendering intent used in the software. On top of that the prints are often severely limited in tonal range anyway compared with what was captured, and in that case the colour gamut may be a relatively minor consideration.
The most important lesson for Alex53 in all of this is to pay no heed to Ken Rockwell. I've seen plenty of criticism of his stuff here at FM from many people and never once a good comment. His articles seem to range somewhere between mischievous and wrong. Maybe pearlstreet is right about his motives.
p.2 #2 · sRGB vs Adobe RGB...I thought I had it right!
I love Ken Rockwell and the way he sees and describes things.
For the record:
You only need to know the color space in which the image is described. On an accurately calibrated/profiled display the image will look correct in an ICC-aware application. sRGB is the best option for an image posted on the Internet. The small gamut fits most displays gamut. A sRGB image will look good on most displays and outside an ICC-aware application, like IE.
I have to rewrite the Guidelines, it really needs an update, but it still gives some insight (see link at top of this page).
For me it isn’t relevant to know if you see differences or not. I try to optimize my workflow. The optimal workflow will be different for every photographer. I try to retain most image data and get the best image quality. Often I choose ProPhoto RGB as my working color space, but frequently I use sRGB. Adobe RGB is a between color space, strengths and weaknesses, I try to choose between large and small.
p.2 #3 · sRGB vs Adobe RGB...I thought I had it right!
Alan321 wrote:
There's little merit to the argument that web browsers expect sRGB images.
That is really not the rationale for editing web images in sRGB working space. Its not that the application expects anything, but rather that the average consumer grade monitor displays a gamut which is similar in size and shape to sRGB. So if the image is converted to sRGB and the target display is calibrated to D65 and 2.2 gamma the image should look similar if viewed side-by-side from within Photoshop and an unmanaged application.
If would seem to follow that if you open an sRGB image in an ICC aware browser on a very high-end monitor with a gamut wider than sRGB the image display will be constrained by the sRGB gamut of the file and thus not be displayed at the full color potential the monitor is capable of. You'd only benefit from the wider gamut of the monitor if an AdbobeRGB file was opened in the ICC aware browser, unless the monitor has some provision for expanding the sRGB gamut via perceptual rendering intent to the full monitor gamut. But that would change the contrast / color of the image creating a different dilemma: the color would be more vibrant on your screen but not match the author's.
Since the goal of color management is consistency device-to-device (not just making pretty colorful pictures) you'd usually want to cater to the lowest common denominator (i.e. D65 / 2.2 calibrated monitor space) and save files with sRGB space for web use.
As mentioned earlier Adobe1998 was apparently conceived as a working space for RGB > CYMK offset printing (based on the fact both device RGB and web press SWOP CYMK fit inside of it) but the gamut of 6, 8 and 9 color ink-jets has outgrown it, which would indicate the need for a larger editing space to take full advantage of the printer gamut from capture in camera-to-print.
p.2 #4 · sRGB vs Adobe RGB...I thought I had it right!
Alex53 wrote:
How do you spot 'out of gamut' issues? Posterising of areas with a particular colour? A colour blown out in print when its not blown out on screen?
Yup. I can think of exactly one image I've had to deal with that...and I was the only one to notice the issue, everyone else viewing the first round of printing thought I was nuts.
Once again, I have to agree with Christo's thoughts, which gets back to what I said earlier--look at your own images & use what space is needed for them.
p.2 #5 · sRGB vs Adobe RGB...I thought I had it right!
Hendrik wrote:
If you print B&W images, then the gamut is not relevant, … there is no color. You can only get good neutrals if you make (or let it make) profiles for your unique paper, printer, settings combination. Sometimes you can be lucky and the canned profiles are good enough. On my R2400 the Epson profiles are very good, but the Innova paper needed a custom-made profile. I print B&W as RGB images.
For perfectly neutral B&W on R2400 try using Advanced B&W settings in Epson driver instead of profiles. I use those most of the time when I do not want toned images.
p.2 #6 · sRGB vs Adobe RGB...I thought I had it right!
cgardner wrote:
That is really not the rationale for editing web images in sRGB working space. Its not that the application expects anything, but rather that the average consumer grade monitor displays a gamut which is similar in size and shape to sRGB. So if the image is converted to sRGB and the target display is calibrated to D65 and 2.2 gamma the image should look similar if viewed side-by-side from within Photoshop and an unmanaged application.
Well, at least under Windows, there is truly a reason why sRGB is required, and yes, the applications do "expect" sRGB, though "expect" is of course, a slightly odd concept for dumb old software. A more proper way to say it is that a lot of software is written assuming the RGB pixel values are to be interpreted as sRGB. It has to do with the Windows bitmap graphics API's. I the bad old days, they had no built-in support for interpreting a color profile in a bitmap image. You loaded the RGB pixel values, and did your bit-blitting. Windows had the RAMDAC in the video card setup to interpret those RGB values into analog color signals using the sRGB color space. Software like PS had to jump through hoops to do proper color management. Things are a lot better now, under XP the Win32 API includes full support for RGB value interpretation and translation between ICC profiles, but the legacy calls exist, and a lot of software out there that displays bitmaps still bypasses all color management and assumes the RGB pixel values in the bitmap data are based on sRGB.
I agree that witth the average uncalibrated consumer display that what actually gets rendered on-screen is something unique of no standard color space, but even there an image displayed with RGB values based on AdobeRGB usually looks noticeably washed out when spit out in Internet Explorer.
p.2 #7 · sRGB vs Adobe RGB...I thought I had it right!
Hendrik wrote:
You only need to know the color space in which the image is described. On an accurately calibrated/profiled display the image will look correct in an ICC-aware application
I'm not convinced. Doesn't the calibration affect the display of the image by offering a different white point and gamma ? The profile is only working within the constraints of the gamma and white point specification of the calibration. If our calibrations are different then the image is then being displayed differently despite profiling and you are not seeing what I was seeing and the whole point of the exercise is lost. We need to share a common colour space and also the same calibration in order for our individual profiles to do the right thing and render the image as identically as the hardware allows.
p.2 #9 · sRGB vs Adobe RGB...I thought I had it right!
Peano: Good illustration, thanks for taking the time!
Alan: You make a good point. There are other factors beside just calibration and color space. Gamma is certainly one. That is why most people recommend Windows folks stick with Gamma of 2.2. Is it the ideal? Ahhh, probably not, but it is standard.
Blackpoint and whitepoint do make a difference, for sure. To make things look best on screen, a blackpoint of zero is wonderful, and there is no such thing as too high a whitepoint. The problem with that is that is makes judging print match about impossible as prints don't offer the same dmax. That, by the way is why there are such outlandish claims on "contrast ratio" these days with the newer LCD's -- the blackpoint can be set very low. Even a dim display with a blackpoint of zero has an infinite contrast ratio.
Anyway, there is really nothing you can do about the fact that different people's displays are calibrated differently. All you can do is calibrate yours using the most commonly accepted standards and let the chips fall. That basically boils down to using a gamma of 2.2 (on Windows, anyway) and using 6500K target color temp. Judging prints to be displayed in tungsten light is easier at 5000K, but it isn't too bad at 6500K, and most people's uncalibrated displays have an even higher target color temp as the standard in the PC industry is to use a very high color temperature (8-9.5K) to improve the contrast of black text on white window backgrounds. If you calibrate your monitor to 5000K, the images will look very different to people using uncalibrated or "factor sort-of calibrated" displays.
A lesser issue is the rendering engine used. Adobe has their own "ACE" rendering engine, and you can set PS to use it (default) or use the Windows engine. In some images, there can be noticeable differences between the two. I've found it's particularly noticeable in images with a lot of finely graduated shadows. For screen presentation, if you want the images to look to other people as close to the way they look to you, you should switch PS to use the Windows engine.
p.2 #10 · sRGB vs Adobe RGB...I thought I had it right!
The whole point of profiling and to use ICC-aware applications is to get colorimetric true representation of your image. It is true that monitor calibration differences will be a factor, but should be minimized after correct calibration/profiling on a reasonable display. Calibration parameters are highly dependent of other factors, like room luminance. My optimal calibration parameters are different compared to yours. The colorimetric representation should be close though if we use a good profile on a good display and within a practical range.
If I would like to know how your monitor behaves, I would need your monitor profile, but this would be useless, since it’s not how all other displays behave. In the ideal world we all work on the same displays (with each a slightly different behavior), under the same room luminance, on the same desk and room and we all wear color neutral clothes. Then maybe we can use calibration parameters more closely to each other.
The goal of calibration is to use the most of the displays capabilities, without clipping. A blackpoint too low, or a whitepoint (brightness) too high will result in clipping.
Btw, I see clearly differences between the photos.
p.2 #11 · sRGB vs Adobe RGB...I thought I had it right!
pearlstreet wrote:
Somewhere on Ken Rockwell's site he says that his site is a spoof, like the Onion. He's deliberately making misstatements and laughing when they provoke discussions.
p.2 #12 · sRGB vs Adobe RGB...I thought I had it right!
Understand that Ken has no use for RAW. Instead he describes methods to get the best JPEGs right out of the camera. Ken also suggest that his readers send their JPEGs out to CostCo etc for printing on their hi Quality, super expensive, chemical machines.
Obviously this leaves out so many of us that shoot RAW, often PS a lot and use our home printers to complete the custom circle to control our output quality. So, obviously, Ken's recommendation for sRGB instead of aRGB (or other profiles) is consistent with his preferences (and retail printers who use JEPG/sRGB files) even though a lot of us have gone past this stage of digital imaging.
p.2 #13 · sRGB vs Adobe RGB...I thought I had it right!
What stage of digital imaging is that which you have gone past, SoundHound? It all depends on what you are doing. Thinking that you are getting superior results by ignoring camera and printer abilities, always using RAW, and always printing yourself is a pretty narrow band of the whole span...
p.2 #14 · sRGB vs Adobe RGB...I thought I had it right!
Thier are many forms of photography today. That is what makes it so great. We can now do things at home that were not possible a few years ago. But it does not mean you have to. Shooting in Raw and trying to get the best possible file,enhancing this file to the best of our abilities. Then trying to make the best Fine Art Print on our own home printer is what many of us enjoy doing. This is not to say it is the only way to go with your photography. Thier are many ways to go and they are all legitimate. But if this is what you prefer sRGB is waisting your time.
p.2 #15 · sRGB vs Adobe RGB...I thought I had it right!
Just to help someone who is new to all this color management stuff, there is a logical reason to use sRGB vs AdobeRGB (1998, like the cameras offer): there simply isn't enough advantage in AdobeRGB to make much difference, and if one is concerned about processing in a wider color space, there is absolutely no reason to not use a wider one than that. No offense to Adobe -- this is their 1998 standard, and that is ancient history in this world.
