RAW - aRGB vs sRGB

I was told today that even though I shoot RAW (no JPG), I still need to to use aRGB and then convert to sRGB JPGs before I get them printed online; being that aRGB will give me a better color space to work from and thus better color prints. Is that true?

My Canon 20D and 40D are set to capture RAW and use sRGB (if that matters). When I get home, I convert the RAW using DPP to JPG for online prints (Smugmug or dotPhoto). Would having aRGB be any better? Or was I told a bunch of hog wash?

Thanks.

RAW files have no color space, they're only given a color space when they're converted to PSD/jpeg. The only difference the color space makes is to the preview embeded jpeg and the filename as an underscore as the first character with Adove RGB. Any conversion loses information, so you're better off going directly to sRgb.

You shouldn't take any advice from the person who told you this ever again, they're obviously full of s...

Tims answer is forthright but not very informative!

Your informants advise has a grain of truth in it, but whether you use sRGB or aRGB, the difference is so marginal it is all but impossible to see.

However, in theory, if you generally print your own work at home and don't upload images or send them out for printing, then there are advantages to aRGB. If you tend to upload most of your work and farm out all your printing, then sRGB is the better choice.

So the real answer is "IT DEPENDS"!

Lowner, my post says to use sRgb rather than convert, amongst the bluntness. Your suggestion about "printing at home" while true probably isn't so useful since the OP "print online".

The real answer for anyone who has to ask this question is "use sRgb", since it's far more difficult to mess things up with an sRgb workflow.

Tim,

Like a lot of others here, you do tend to be a little too "abrupt" in your responses. I was not disagreeing with you, but trying to explain a little further. I'm sorry if you thought it was anything else.

I disagree with the oft repeated "Use sRGB" that is aired here frequently, everything depends on how the eventual image is to be used. sRGB might be the right call in one set of circumstances but wrong in others and I believe the OP deserves a fuller explanation.

I think this could be somewhat related to what has been told to the op: I noticed that in the dpreview test of the original 5D they noticed a difference in color saturation between sRGB and sRGB converted from Adobe RGB.

This is the link: http://www.dpreview.co​m/reviews/CanonEOS5D/p​age13.asp

They don't explain how they did the test, I presume shooting jpeg with sRGB first and then with aRGB. They should have converted the profile in photoshop, right?

Why there is such a difference in color saturation?

Fede

Fede,

All I can say is I recently discovered that numbers on a homemade greyscale were different in sRGB to those I'd given them in aRGB when it was created. This was a pure greyscale and I can only imagine that the same is true for the "real" colours as well.

The effect was at the black end of the scale, not the white, and the blocks of dark actually became slightly more difficult to tell apart on screen than one I susequently created from scratch in sRGB as an experiment The change occured when I converted one gamut to the other to post here.

Until then, I'd have told anyone that I'd never seen any difference between sRGB and aRGB. The visible differences are tiny and only noticable because we were discussing monitor dynamic ranges and calibrating. I've still not seen differences in colours but that does not mean they are not there.

Fedone1275 wrote in post #7161510
I think this could be somewhat related to what has been told to the op: I noticed that in the dpreview test of the original 5D they noticed a difference in color saturation between sRGB and sRGB converted from Adobe RGB.

This is the link: http://www.dpreview.co​m/reviews/CanonEOS5D/p​age13.asp

They don't explain how they did the test, I presume shooting jpeg with sRGB first and then with aRGB. They should have converted the profile in photoshop, right?

Why there is such a difference in color saturation?

Fede

Note they say:

(Not quite corect by the way: Both FF3 and PSCS4 will recognise the sRGB and AdobeRGB tag in the image).

On the difference: I can only presume Canon goofed around with the saturation of the sRGB image to prevent clipping (without success I might add).

Both images are clipping *big time*, so that *will* cause a color / saturation shift.

I'd be curious what a well done test would conclude, as well as whether a conversion (trough Canon software) would show the same differences.


If the camera isn't altering (faking) things, and no colors are outside gamut, sRGB and AdobeRGB should look identical.

If colors are ouside sRGB but within AdobeRGB, then they'll clip in sRGB, but not in AdobeRGB. > More accurate and "saturated" colors in AdobeRGB in that case.

Lowner wrote in post #7161430
Tim,

Like a lot of others here, you do tend to be a little too "abrupt" in your responses. I was not disagreeing with you, but trying to explain a little further. I'm sorry if you thought it was anything else.

I disagree with the oft repeated "Use sRGB" that is aired here frequently, everything depends on how the eventual image is to be used. sRGB might be the right call in one set of circumstances but wrong in others and I believe the OP deserves a fuller explanation.

Just a little Using other color profiles has benefits, but if people don't understand color profiles they're likely to mess it up and have an inferior end result, which is why I recommend sRgb exclusively for color n00bs.

René Damkot wrote in post #7163609
Note they say:

(Not quite corect by the way: Both FF3 and PSCS4 will recognise the sRGB and AdobeRGB tag in the image).

On the difference: I can only presume Canon goofed around with the saturation of the sRGB image to prevent clipping (without success I might add).

Both images are clipping *big time*, so that *will* cause a color / saturation shift.

I'd be curious what a well done test would conclude, as well as whether a conversion (trough Canon software) would show the same differences.

If the camera isn't altering (faking) things, and no colors are outside gamut, sRGB and AdobeRGB should look identical.

If colors are ouside sRGB but within AdobeRGB, then they'll clip in sRGB, but not in AdobeRGB. > More accurate and "saturated" colors in AdobeRGB in that case.

Rene' so there is no explanation why the ColorChecker chart shots show different saturation? Both pictures are sRGB, but one has been converted from aRGB. How did they do it then?

Has anyone printed a JPG with aRGB ... OUTSIDE of Photoshop?
Doesn't most home/office inkjet printers only "see" sRGB?
I was told if I was to print from Adobe Photoshop, the aRGB is really "converted" to sRGB and then sent to the printer driver (all behind the scene).
Is that true?

No. I don't know where you've been getting your information but you really need to stop using that source.

Get this book, it will teach you color from the basics up to everything you need to know, rather than relying on others who know think they know something. Color's quite a complex topic, quite technical, but that book makes it a bit easier. Once you read it you won't be asking questions, you'll be answering them.

JChin,

My new Epson R2880 works in sRGB, but the printer that it replaced (again an Epson) used aRGB. However I read in the i-photo review that the printer is able to print a larger gamut than either. If someone would care to explain how thats possible when applying a certain colour space I for one would learn something. Because it sounds impossible.

I certainly have never bothered to convert the image from aRGB to sRGB for printing. As I've already noted here, the difference in the two is very marginal, but aRGB has a slightly larger gamut than sRGB and that's why we see the suptle differences when converting aRGB to sRGB - that larger gamut is squeezed into a slightly smaller one and something has to give.

Richard, Epson pro inkjet printers have a larger gamut than commercial printers, larger than sRgb. The only time you see it is times when sRgb will clip, which is just with some of the more vivid colors.

Like most things in photography the best answer is "try it and see". Go take photos of something with vivid colors, in RAW, as that captures as much information as possible. Convert to sRgb, Adobe RGB, and ProFoto RGB, then print them. Compare the prints. Make your decisions based on these prints. At this point you'll have a soap box you can stand on to debunk the theory that most people go by, practical experience is more important.

Lowner wrote in post #7169094
My new Epson R2880 works in sRGB, but the printer that it replaced (again an Epson) used aRGB.

Both printers worked in their own individual and unique color spaces as determined by the physical characteristics of the machine, the inks and the paper. Profiling a printer measures and records that space. The printer and/or paper companies supply generic profiles that are usually close but not perfect. Color management, either in the editing/printing application or in the printer driver, remaps the image data from the working space to the printer space.

I believe that statement should have been "...the printer is able to print a larger gamut provided the source image is in an equal or larger space..." When remapping from a smaller gamut to a larger one the CM will maintain all the colors present in the data but will not invent new ones. The brightest red in sRGB is 255/0/0. That same red might be 245/0/0 in AdobeRGB and 235/0/0 in the printer's space (numbers not exact). Print an sRGB image that contains 255/0/0 and by the time the data gets to the printer's processor it will have been changed so that the desired red will be represented as 235/0/0 and the range from 235 to 255 will not be utilized.