Adobe RGB vs sRGB

[RobertBoire]

 OK, so I most admit the last several posts on UniWB were a bit esoteric (at least for me) but I''m back to one of my original questions. I understand that sRGB is preferred for the Web, but I am not clear why. Is because the monitors that most people use support only sRGB or is there something else.

[JeffKohn]

OK, so I most admit the last several posts on UniWB were a bit esoteric (at least for me) but I''m back to one of my original questions. I understand that sRGB is preferred for the Web, but I am not clear why. Is because the monitors that most people use support only sRGB or is there something else.

It's because many if not most people are surfing the web with browsers that treat all images as if they were sRGB.  When a browser displays a non-sRGB image as if it were sRGB, the result can look pretty bad (particularly for Pro Photo RGB).

[RobertBoire]

So it has nothing to do with the monitor but the application?

Meaning, for example, if I am showing a "slide" show on my monitor using an application that understands what color space the image is in (like Photoshop does), then the result should be ok?

[Slobodan Blagojevic]

A brief introduction to sRGB vs. Adobe RGB can be found here:

http://www.cambridgeincolour.com/tutorials/sRGB-AdobeRGB1998.htm

[ErikKaffehr]

Hi,

Yes, if you can find one and if your images are correctly tagged. The monitor should naturally be properly calibrated.

Adobe RGB and sRGB are pretty close, even if Adobe RGB primaries stretch farther in the blue greens. You can probably get away with showing Adobe RGB on an sRGB system even without correctly handling profiles. Prophoto RGB is another matter, it will be obviously bad.

That is actually an advantage with Prophoto RGB, it's obvious when wrong!
My 2 cents:

- If you have an Adobe RGB capable monitor use Adobe RGB
- If colors fall into sRGB than choice of RGB will not matter, but it may be that you get better tonal separation in sRGB
- Use ProphotoRGB as working space, this way you will not throw away colors that would be outside sRGB or Adobe RGB
- Export to sRGB if you are not absolutely sure that the recipient knows what he/she is doing

Best regards
Erik

So it has nothing to do with the monitor but the application?

Meaning, for example, if I am showing a "slide" show on my monitor using an application that understands what color space the image is in (like Photoshop does), then the result should be ok?

[elied]

So it has nothing to do with the monitor but the application?

No, it has to do with the monitor also. The sRGB space was created in 1996 by Microsoft and HP and designed to represent the native space of a theoretical "average" monitor in use at that time. There were deviations of course, but sRGB would be the "golden mean". Despite the fact that there are today many "wide gamut" monitors in the market, the large majority continue to be reasonably close to the sRGB model. Thus, if the image data is in sRGB, when it is passed through to most monitors without any alteration by a non-c.m. browser the display may not be accurate but it will be acceptable.

[Sheldon N]

As someone else said early on in the thread, it really boils down to individual need and usage.
A little goody I came across,  from the horses mouth ...


http://www.adobe.com/digitalimag/pdfs/phscs2ip_colspace.pdf

Are you calling Andrew Rodney a horse? LOL  

He's the author of that article, and has already posted to this thread several times (User ID digitaldog).

[digitaldog]

A little goody I came across,  from the horses mouth ...

From the dogs mouth <g>

[hjulenissen]

Did anyone mention that sRGB has an advantage over Adobe RGB when it comes to banding?

-h

[digitaldog]

Did anyone mention that sRGB has an advantage over Adobe RGB when it comes to banding?

An advantage, how so?

[Schewe]

Did anyone mention that sRGB has an advantage over Adobe RGB when it comes to banding?

If you have proof that the same image in sRGB bands less than the image in Adobe RGB with the same processing, I would like to see it...

A lot of people talk about the "efficiency" of a color space being important. To date, I've never actually seen proof. I've been using Pro Photo RGB since Bruce Fraser talked me into doing so in early 2001 (the only caveat is I also only use 16 bit). I've never seen banding that could be attributed to having used an inefficient color space. I have however seen banding introduced because 8 bits simply were not enough-regardless of the color space the image was in.

[hjulenissen]

An advantage, how so?

I did not investigate this my self, but the argument is often heard, and seems reasonable: if you stretch 8 bits over a larger distance, each step will be larger.

That is, of course, assuming that one is limited to 8/24bpp.
http://www.earthboundlight.com/phototips/srgb-versus-adobe-rgb-debate.html

[digitaldog]

if you stretch 8 bits over a larger distance, each step will be larger.

But I don’t stretch 8-bits per color nor recommend anyone do so.

[hjulenissen]

But I don’t stretch 8-bits per color nor recommend anyone do so.

It was my understanding that the threadstarter was asking about sRGB/Adobe RGB formats out of his Canon camera?

...
I know that Adobe has a wider gamut than sRGB. However are their situations when sRGB is in fact preferred?

I ask because my Canon user manual states "Adobe RGB is not recommended if you do not know about image processing, Adobe RGB and Design Rule for Camera File System 2.0", which in fact pretty much tells me nothing at all...

[digitaldog]

It was my understanding that the threadstarter was asking about sRGB/Adobe RGB formats out of his Canon camera?

His camera captures raw which is neither of those color spaces nor is it 8-bits per color. He wrote:

I know that Adobe has a wider gamut than sRGB. However are their situations when sRGB is in fact preferred?
I ask because my Canon user manual states "Adobe RGB is not recommended if you do not know about image processing, Adobe RGB and Design Rule for Camera File System 2.0", which in fact pretty much tells me nothing at all.

And then he stated:

In fact I have been shooting RAW recently.

[Schewe]

I did not investigate this my self, but the argument is often heard, and seems reasonable: if you stretch 8 bits over a larger distance, each step will be larger.

That is, of course, assuming that one is limited to 8/24bpp.
http://www.earthboundlight.com/phototips/srgb-versus-adobe-rgb-debate.html

I read the article...it only goes along with the myth and offers no PROOF that Adobe RGB is any more prone to banding than sRGB. So, until I see some actual proof that Adobe RGB is more prone to banding, I think I'll file this under "it's an old wives tale" rather than proven fact. Heck, I've even played with 8 bit ProPhoto RGB files to try to make them band...and you can if you apply outrageous adjustments to the file but the same adjustments will cause banding in sRGB and Adobe RGB as well. Working in 16 bit (actually, 15 bit plus 1 level) eliminates the need to worry about banding.

[hjulenissen]

Working in 16 bit (actually, 15 bit plus 1 level) eliminates the need to worry about banding.

Agreed, If that is possible.

However, if you want to use JPEGs out of a camera, you usually have the choice between 8-bit sRGB and 8-bit Adobe RGB (further losses introduced by matrixing, color subsampling and JPEG). Most people that are really into precision will probably use raw, but there are some scenarios where one might have to use JPEG (if maximum sustained fps is important, or maximum available storage)

-h

[jbrembat]

Abobe RGB 1998 is a  color space larger than sRGB.
Using the same bit depth, in Adobe RGB 1998 a color step is stronger.
That does'nt require a proof.
So, theoretically speaking, Adobe RGB 1998 is more prone to banding than sRGB.

But, in practice, you have to evaluate if banding is visible.
Banding visibility is related to noise level too.

Generally, a good editing (not an extreme one) doesn't generate banding in any color space.

Jacopo