[distudio]

They may not fully utilize the available gamut of the printer, but is that the same thing as the gamut of photographic paper?

The ICC describes the printer output gamut including paper.

Description Tag:
Gen_Sensatis_dlab_V1

ICC-Tag: BToA1 (Colorimetric)
Calculation: LUT with 35937 grid points

Header:
size = 646171 bytes
CMM = 'appl'
Version = 2.0.0
Device Class = Output
Color Space = RGB
Conn. Space = Lab
Date, Time = 4 Aug 2004, 16:11:01
Platform = Macintosh
Flags = Not Embedded Profile, Use anywhere
Dev. Mnfctr. = 0x0
Dev. Model = 0x0
Dev. Attrbts = Reflective, Glossy
Rndrng Intnt = Perceptual
Illuminant = 0, 1, 0 [Lab 100, -0, 0]
Creator = 0x0

Total volume of gamut is 425861 cubic colorspace units

\37777777650 2001-2002 Color Solutions Software\000\000\000

[attack11]

which service providers accept argb? all the ones available through zenfolio are srgb, KISS books are srgb; and that's where my printing ends, so i'm kinda curious.

does it matter if you're not printing at home or doing it for large publication?

[distudio]

which service providers accept argb? all the ones available through zenfolio are srgb, KISS books are srgb; and that's where my printing ends, so i'm kinda curious.

does it matter if you're not printing at home or doing it for large publication?

It's probably more important to consider a wide gamut work-flow if you are printing at home or via a specialist bureau, also it appears that many stock image agencies prefer or demand images in AdobeRGB colour space (it's easy to convert sRGB images to AdobeRGB it just serves no real benefit and could in fact cause IQ deterioration).

[GoremanX]

Now here's something I hadn't thought of before; since both sRGB and Adobe RGB are represented by the same 8 bit address space (I'm ignoring 16 bit for now), doesn't that mean that there are fewer steps between two colors? So a gradient that goes from blue to magenta, for example a sunset sky, would technically be smoother in sRGB than in Adobe RGB, since there are more colors available between blue and magenta in sRGB than in Adobe RGB. Changing from one color space to another doesn't change the way colors are represented, it's still a scale from 0-255.

[GoremanX]

I think this is one of the most telling diagrams I've seen so far. This is from the following web site:

Working Space Comparison: sRGB vs. Adobe RGB 1998

The image shows the color space of Adobe RGB outlined in black, sRGB outlined in white... and the typical WalMart printer (Fuji Frontier 390) outlined in gray.


...ouch...
maybe I won't use WalMart anymore.

[Kelarch]

Now here's something I hadn't thought of before; since both sRGB and Adobe RGB are represented by the same 8 bit address space (I'm ignoring 16 bit for now), doesn't that mean that there are fewer steps between two colors? So a gradient that goes from blue to magenta, for example a sunset sky, would technically be smoother in sRGB than in Adobe RGB, since there are more colors available between blue and magenta in sRGB than in Adobe RGB. Changing from one color space to another doesn't change the way colors are represented, it's still a scale from 0-255.

Yes, there are fewer steps in the gradiation which causes a risk of posterization when using wide gamut spaces in 8-bit mode. I use ProPhoto and anybody who uses ProPhoto will tell you don't even think about trying to use it in 8-bit. You run the risk of absolutely wrecking your image once you start pushing it around in an image editor. Thankfully Photoshop has really good 16-bit editing support. GIMP users are SOL, being stuck with 8-bit only.

[GoremanX]

GIMP users are SOL, being stuck with 8-bit only.

For now, anyways. 16 bit support is already mostly-implemented and can be enabled experimentally. It will be the default by either version 2.8 (in alpha) or 3.0.

[tlwyse]

Colourspace is assigned as part of raw conversion.
The OP has become aware that he doesn't know very much about colour management, but he doesn't yet understand how very little he understands, and so he is blaming the wrong thing.

Egg-zackly!

You can't reliably display the differences between sRGB and aRGB on a forum such as this....
1) They need to be tagged with the correct profile.
2) Your web browser has to be color-managed (Apple Safari is by default, FireFox is if you enable color management...and Internet Exploder isn't color-managed at all).

If you wanted to better show the differences, it would've been best to take your aRGB images and then *convert* (not assign) them back to sRGB. That would've given us a more reliable comparison regardless of browser platform.

Personally, I use ProPhotoRGB because....
A) I'm using a color-managed RIP(s) (GMG and ColorBurst) that, given the right paper, can actually exceed AdobeRGB.
B) While my display is "sRGB" (EIZO CG211), I can soft-proof in my printer space and get a reasonable approximation of how the print will reproduce...thus ProPhoto doesn't really hinder me all that much.
C) I'm not going to limit my colorspace to the weakest link (the display). IOW, I don't rely 100% on my display as the reference since the print is what I'm "selling".

Having said all THAT, if you're only goal is to view your prints via a web browser, then by all means stick with sRGB as that's the safest choice.

Regards,
Terry

[tlwyse]

Now here's something I hadn't thought of before; since both sRGB and Adobe RGB are represented by the same 8 bit address space (I'm ignoring 16 bit for now), doesn't that mean that there are fewer steps between two colors? So a gradient that goes from blue to magenta, for example a sunset sky, would technically be smoother in sRGB than in Adobe RGB, since there are more colors available between blue and magenta in sRGB than in Adobe RGB. Changing from one color space to another doesn't change the way colors are represented, it's still a scale from 0-255.

It would be more accurate to say that the spaces between each step is wider with aRGB vs. sRGB. IOW, any move you make, such as a curve adjustment, in aRGB will have a bigger effect on the image than the same move/edit in sRGB. Conversely, you have a bit finer control with edits in sRGB than you do in aRGB, assuming you're still in 8bpc (8 bits per channel). If you use a ruler analogy, AdobeRGB is a longer ruler than sRGB but still divided into the same number of increments. The difference between 8bpc and 16bpc is that the ruler is the same LENGTH but in 16bpc it's divided up into finer increments thus decreasing the space between each step/increment and giving you back finer control (and smoothness) than with 8bpc.

Terry

[tlwyse]

What's meant by "working" in Adobe RGB or sRGB is the workflow used to get from the source file to the final result, not necessarily just the processing you do to it. Sometimes this involves more than one application. If, at any step along your workflow you use anything but Adobe RGB (even as a final output result), then you're completely negating the use of Adobe RGB in the first place. This whole theory about "preserving the wider colorspace" doesn't make sense, because color management-aware applications (like Photoshop) process colors internally using the whole color space anyways.

Photoshop does NOT process colors "using the whole color space", if by "color space" you mean the widest available. If you've already "clipped" your image to sRGB and then do subsequent corrections, any color processing/editing will be restricted to sRGB. On the other hand, if convert from sRGB to Lab and then apply some corrections, you'll be editing "using the whole color space".

You also state "If, at any step along your workflow you use anything but Adobe RGB (even as a final output result), then you're completely negating the use of Adobe RGB in the first place." This is not true either. If the output/printer profile you're using is a good one, even if it doesn't have quite the gamut volume as aRGB (but larger than sRGB), the net result is that you've still gained some "color" that would not have been available had you clipped it to sRGB early in the workflow.

Don't get me wrong, I'm not at all saying that sRGB is going to result in "lower quality" images, I'm just saying that if you know exactly what your final output is going to be (photo lab prints, browser viewing, high-end inkjet printing), you'd be best served by working in a color space/profile that meets or slightly exceeds the final output. Frankly, I think most folks would be best served by sticking with sRGB, especially if they are using an outside photo printing service, using online photo galleries exclusivly and/or are not willing to invest in the gear necessary to custom profile their inkjet printer.

[GoremanX]

You can't reliably display the differences between sRGB and aRGB on a forum such as this....
1) They need to be tagged with the correct profile.
2) Your web browser has to be color-managed (Apple Safari is by default, FireFox is if you enable color management...and Internet Exploder isn't color-managed at all).

Right... except that JPG pictures straight out of the camera that are in AdobeRGB space will NOT display properly on your browser even if you have color management enabled (perhaps Safari excluded). This is because Pentax uses the DCF 2.0 format for specifying an optional color space in the EXIF metadata instead of embedding an actual color profile, and many applications are not configured to determine color space that way. Even Pentax's very own in-camera image viewer isn't color managed. Therefore, the original post stands: don't bother setting the camera to AdobeRGB, it gives bad results.

That's pretty much the conclusion that this whole thread came to.

[GoremanX]

Photoshop does NOT process colors "using the whole color space", if by "color space" you mean the widest available. If you've already "clipped" your image to sRGB and then do subsequent corrections, any color processing/editing will be restricted to sRGB. On the other hand, if convert from sRGB to Lab and then apply some corrections, you'll be editing "using the whole color space".

That was a poor statement on my part. What I meant was that Photoshop processes all color operations internally in 16bit, then interpolates the results back to the 8bit image. The entirety of Photoshop's color engine runs in 16bit internally.

[tlwyse]

Right... except that JPG pictures straight out of the camera that are in AdobeRGB space will NOT display properly on your browser even if you have color management enabled (perhaps Safari excluded). This is because Pentax uses the DCF 2.0 format for specifying an optional color space in the EXIF metadata instead of embedding an actual color profile, and many applications are not configured to determine color space that way. Even Pentax's very own in-camera image viewer isn't color managed. Therefore, the original post stands: don't bother setting the camera to AdobeRGB, it gives bad results..

But if you know that it's coming out of the camera as AdobeRGB, if you then opened the image in Photoshop and assigned/tagged it with AdobeRGB would it still produce "bad" results? I'm just asking as I never use JPEGs out of the camera except for "proof" images. Everything else is raw.

Regards,
Terry

[tlwyse]

That was a poor statement on my part. What I meant was that Photoshop processes all color operations internally in 16bit, then interpolates the results back to the 8bit image. The entirety of Photoshop's color engine runs in 16bit internally.

Are you implying that it's not necessary to save your image in 16bpc but to simply keep it 8bpc since Photoshop still processes it "as if" it's 16bit?

Regards,
Terry

[distudio]

But if you know that it's coming out of the camera as AdobeRGB, if you then opened the image in Photoshop and assigned/tagged it with AdobeRGB would it still produce "bad" results? I'm just asking as I never use JPEGs out of the camera except for "proof" images. Everything else is raw.

Terry, the out of camera AdobeRGB jpegs on import to PS are recognised as tagged AdobeRGB and will produce the correct results.