SRGB vs RGB 1998

I think you mean sRGB -v- aRGB?

To cut a long story short, aRGB is capable of handling more colours than sRGB, but thinks will look darker if you're using a program that doesn't understand aRGB to view them.

If you're using something like Photoshop (which understands aRGB) then it displays it just fine.

"...(which understands aRGB) then it displays it just fine"

Actually, if your using a color-managed application then an sRGB and Adobe RGB image should look the same. They will look different in non-color-managed applications (like most Windows web-browsers). In the latter case, sRGB is closer to the run-of-the-mill PC display, so it's usually better to save anything that you're making for the web as sRGB.

seriously i'm ready to fall on a knife or stick needles in my eyes when it comes to "color management"....i'm definitely talking about "srgb" vs "adobe rgb 1998" when using lightroom i have these options when i export....one lab i use wants them in "srgb" so i noticed that srgb has more color than adobe rgb 1998 and according to what i read srgb is a smaller color space. so i'm very confused - can someone please break this down for me???????????

Oh Boy ... here we go ... :)

RGB stands for Red, Green, and Blue - the three primary colours that we use to make up our images.

So - if you were to peek inside a photo file you'd find that a particular pixel might have a R(ed) value of 123; a G(reen) value of 231, and a B(lue) value of 23 - unfortunately, although we know "how much red", "how much green" and "how much blue" to use for this pixel, we don't know what shade of Red, green, or Blue that we're talking about - if the shades of red, green, and blue that your camera uses is different to the red, green, and blue that my monitor uses then what I'm going to see isn't what the image is supposed to look like.

So ... we need standards - standards that define "what shades of red, green, and blue" were talking about - and that's what sRSB and aRGB 1998 are - different "shades" of colour. So to correctly describe an image you need a couple of things - firstly you need "the numbers" then you need a "standard" (called a colourspace) that defines what the numbers mean.

It's a bit like me typing an eMail in English - and then sending it to someone in Italy who assumes that what I've written is in Italian - and because of this the meaning of the eMail changes. If I'd written the eMail in English and sent it with a big post-it note attached that said "written in English" then all would be sweet.

So - if you capture an image and save it in aRGB format, and then dislay it or print it on a device that's expecting sRGB then the image won't display correctly - in general it will be darker with more subdued colors.

To explain why it looks darker when aRGB has more colours ... think of a single red pixel - that pixel under sRGB might have a value of 200 (out of 254) - but because aRGB has to allow for more colours it has to be more conservative in it's numbering (because 254 is still the limit) - so in aRGB that same red pixel - at the same level of brightness might only be given a number of 175. Now when you display that (aRGB) pixel in a prgram that thinks it's a sRGB pixel all it sees is a pixel that's 175 (not 200) - so it displays it dimmer.

To be honest unless you're a glutten for punishment, don't worry too much about it - just work in sRGB and you won't go too far wrong.

You might be wondering why they have these different colourspaces (there's even bigger ones like pro-photo and LAB) - the basic problem is that the human eye can see a certain range of colours - cameras capture a different range - and printers print get a different range again - sticking to a small colourspace like sRGB keeps you out of trouble (it's pretty much just the lowest common denominator), but cuts down on what might have been possible - the flip side of the coin is that other spaces like LAB are cabable of describing colours that we can neither print nor display - and it's possible to get oneself into all kinds of trouble if one doesn't know what they're doing.

Hope this helps - just let me know if you need more help and I'll try to expand on it some more for you.

Cheers,

Colin

Meg, I also got sick of all the geek talk about color management. It's all true, but I really don't give a darn ... I just want to get what I'm seeing on the monitor on a print. So, for my pea brain I simplified it all ...

1) calibrate the monitor. I use the Mac calibration for my 23" cinemas. At the end of the calibration process I can assign a name to it, send that calibration to be saved in the list of color profiles.

2) In Photoshop I go to >Edit in the top menu and scroll to >Color Settings, then select that same color profile that I named as my Monitor working space from among all the profiles in the list.

3) with a image open in PS, go to > View in the top menu and select >Proof Set Up > Custom ... and select that same profile from the drop down menu. Once you select that, it'll stay until you change it.

4) When sending prints to my home printer, I select that same profile in the printing dialog box and let Photoshop determine color NOT the printer.

5) If sending for external prints or for on-line proofs, I batch CONVERT to sRGB. Convert attempts to keep the same relative color appearance. If a printing lab has a specific profile they want you to use, it gets loaded into the profiles, and you CONVERT to that profile instead of sRGB.

The key is what you are seeing on the monitor is what you get in the print ... so it just made sense to me to use the monitor profile in every place there's a choice. The CONVERT option is suppose to do exactly that ... interpret the colors you see to a different color space, not change them.

I know I'll get ripped apart by the geeks, but it works for me whether printing on my home printer or from a lab. Even when I've done a match print at home and brought the same image to a lab, it usually matches.

Hi Marc :)

You were doing well until you got to #3 - this is where you're setting up the soft-proofing so that you can simulate how it's going print, but on your screen rather than waste paper and ink.

For this bit you should select the profile for your output device (ie printer). If you get away with it most times then great, but if your trying to soft-proof something that has vastly different tone characteristics (like canvas) it'll bite you in the bum if you make changes to get the softproof "looking right" (like dark greys of a grooms suit) - if you don't soft-proof (and usually there's no need to) then it doesn't make any difference.

In step 4 you should also be selecting the correct profile for your output device - by selecting a monitor profile as your output profile the degree to which it matches is proportional to the degree to which your printer behaves like a monitor (from a colour management perspective).

Unless you are using a stupidly expensive screen, and as far as I know there are only 2 of them, you won't be able to see more than sRGB on screen anyway. I hope this answers your original question.

Colin, I don't select the output device profile, I tell the print dialog to let Photoshop select the colors in the "Color Handling" drop down menu. Since I selected my Monitor profile as the proof profile, it automatically is that profile that PS is using. Under "Print" both the Document and Proof options are the same profile... so it doesn't matter which is selected, it's the Monitor profile that is converted to the printer color space. Since I'm sending prints to two different printers (Epson 2400 and 3800 ) it's always the same ... the same photo sent to both printers matches ... which I tested to be sure. Seems to consistantly work. BUT I do select which paper I'm using when printing.

Profiles fall into two categroies - device independent and device dependent.

Examples of device independent are sRGB, Adobe RGB, Pro Photo RGB, etc - these are also referd to as working spaces for images. Some of these are able to represent a wider gamut of colors, and each has a fixed definition of how to represent colors.

Examples of device dependent profiles are monitor, printer, scanner, etc. These are profiles for a specific piece of hardware (either an input or an output device). Printer profiles are specific for the printer with a specific ink and paper combination - you need one for each paper and ink conbination for that printer. Monitor profiles are specific to your monitor and are not adequate for any other monitor (if you want accurate color).

You should never assign a device dependent profile (monitor or others) as your working space for image editing software. You should select a device independent profile (sRGB, Adobe RGB, etc) for your working space. Photoshop automatically selects the current active monitor profile and uses it to modify data sent to the monitor so it displays color correctly. Some software requires that you select a monitor profile so it can use it to modify data sent to the monitor to display color correctly.

Software that displays images can be either color managed or not color managed. Color managed software reads the embedded profile of an image (assuming it was embedded) and uses it to modify data so it displays color correctly. Photoshop is color managed, so an images with sRGB, Adobe RGB, Pro Photo RGB, etc will all appear the same on the monitor. Non color managed software (Windows Picture and Fax Viewer, the internet, and most others) will only display sRGB correctly - others will appear flat and low in saturation.

That is interesting info Robert. I'll have to read up on that more , (although I'm reluctant to alter anything that already works for me, I most certainly am willing to change to something that works better or more consistantly across all devices).

What I did not know is that Photoshop automatically selects the active monitor profile to alter the image on the monitor so it is correct.

Thanks.

Robert's information is dead on. What you're doing, Marc, is creating a closed-loop system that may well work for you, but is hard to generalize either to new labs or to other people. I completely agree that changing something that doesn't work is a questionable tactic, but I'd be reluctant to recommend that path for other folks, because it may not work consistently, and if their monitors are weaker in some way, the Monitor space may create image problems.

Not that I'm anyone, but... if I were to attempt to simplify, I'd end up here:

1. Calibrate/Profile the monitor with a hardware/software setup like the I1 Display 2. The first time, it takes 20 minutes. After that, it really takes like five minutes and not a lot of thought.

2. Work in a device-independent space. sRGB is a lowest-common-denominator space that can be dealt with by all cameras, most labs, and it contains most of the colors you need for many purposes. Adobe (98) is the next-larger commonly-used space and it is also a very reasonable choice - some printers already exceed aRGB, and some real-world scenes contain those colors, but it's not necessary for every purpose. Larger spaces, such as ProPhoto, have some real advantages, but also shouldn't be used in 8-bit editing (for reasons that are pretty geeky but can have real-world impact). [Yes, I use Prophoto, but I don't shoot weddings professionally as y'all do.]

3. Before printing, Convert to the output space (which may be sRGB, a printer-specific space, or a space provided by a pro lab). This part sounds like it's complicated and adds a step to the process. However, once you know what space you're going to, it's doesn't really take any more thought. It can be done in batch.

4. Anything that gets sent to the web should be in sRGB. When changing to sRGB from something else, always use "Convert", and not "Assign".

I have no idea if that helps, but it sure beats going back to the *sigh* work I'm about to go back to...

Too add to Marshall's comment and tie into Lightroom, Lightroom uses the ProPhoto color space. When you export Lightroom gives you the option of which color space to assign. Also, when editing in an external application from Lightroom, such as PS CS3, you can set the color space that will be assigned when it is opened in that application. It is set in Lightroom's Preferences > External Editing.

Also, one more semi-geeky note: the fact that sRGB is often used as an "output space" is confusing. It really isn't an output space (no device exactly matches sRGB). However, because it is an effective lowest common denominator, it often gets treated as such.

Hi Meg!

I spent five years in color science for a major printer manufacturer. Let me explain the basics.

The reason the colors look better on your screen under sRGB is because you are not using color management to display the Adobe RGB files. If color management was activated properly, the two files should look quite similar. If not, the Adobe RGB file will look desaturated, particularly in the reds.

The sad truth is that although Adobe RGB has a much wider gamut space, most customers won't be able to tell the difference if both files are printed properly. So if you are having troubles, stick to sRGB. MPIX, and many other print services, only accept sRGB anyway.

Later,

Paulsky

More and more printing services are now accepting aRGB. That's a fact.

Files saved in aRGB tend to look towards the future. If you want your files to be viable (working with more color) into future technologies that can handle aRGB (and more) then I'd suggest saving a set in aRGB.

That may be important or not ... depending on your view of your photography and longevity for work and for the work of your customers.

Howdy!

William, when you say aRGB, do you mean Apple RGB or Adobe RGB (1998)? They're different animals entirely.

Apple RGB has roughly the same gamut size as sRGB, but the primaries are slightly different. Adobe RGB has a much wider gamut than sRGB.

William has a valid point, but if you want to be really forward looking (as I am), save the RAW files, or better yet, convert them to DNG with RAW pixel information and save that. Color space is a VERY minor aspect of image quality. There's also noise reduction, sharpening, color balance, etc. Future generations of tools might do a better job at these things as well.

Ideally, the best editable archival format other than RAW or DNG is sixteen bit TIFF in the Pro Photo color space. But those files eat disk space because sixteen bit TIFF's don't compress very well. For example, an eight megapixel 16 bit TIFF is 44 megabytes, which can use up a lot of gigs in a very short period of time.

If you examine the content of most natural scenes (which is what wedding photographers shoot) you will find that they fit nicely into the sRGB space, and there is little to no gamut clipping. Most of the time, you'll lose a lot more gamut printing the picture with standard RA4 processes on archival photo paper that you would by converting from a large gamut space to sRGB. In order to take advantage of the Adobe RGB color space, you need to use something like a six ink (or greater) inkjet printer which can pop the colors, or another wide gamut printer.

Therefore, my editable archival format is sRGB JPEG, because it's what my print service wants, and it's more than adequate for the current state of the art in mass printing. I keep the RAW files in case customers want something special, or the state of the art changes.

Later,

Paulsky

aRGB(1998)

might be overkill but I save them all: sRGB, aRGB (not apple, lol) and RAW

Looks like you've got the bases covered. ;<)>

I tend to be pretty geeky about this stuff, but isn't there an argument to be made for wedding shooters to prioritize workflow at high quality over squeezing the last 0.5% out of the files? The last 1000+ shot thing I did had some intense colors so I wanted to start with the widest space possible, but for most weddings (correct me if I'm wrong) that isn't as much of the concern. Am I off-base here?

Marc,

If you're keen to learn more about it (and it's something I think EVERY photographer needs to know), then pickup a copy of Real World Color Management by Fraser, Bunting, and Murphy - it covers all of this in great detail (it really is the gold standard / bible of colour management).

Cheers,

Colin

Marshall Goff wrote: "Also, one more semi-geeky note: the fact that sRGB is often used as an "output space" is confusing. It really isn't an output space (no device exactly matches sRGB). However, because it is an effective lowest common denominator, it often gets treated as such."

sRGB came to be more because of a need for some kind of colour standard on the web - it's definately the LCD, but it "kind of" works!

Further reading at ...

http://www.w3.org/Graphics/Color/sRGB

I loved Real World Color Management, but should we really recommend a 500-page tome deep with the science of color management to people who get turned off by it. Might be overkill, good as it is...

I know what you mean Marshall. After thinking about it long and deep, I personally think it's something that every photographer needs to know, but having said that I don't necessarily think that they need to know the whole book.

In particular I thought that the first couple of chapters lay an excellent foundation - the last few are somewhat program specific, so many could be skipped - and as for the middle bit, I felt that if someone were to "dig deep" and plough through it then it's knowledge that will carry them a long long way. I'd suggest that it's because of books like that that the likes of you and I can help others with questions like this, and ultimately help ourselves - it's definately a tough read in places, but it also made me a much better processor and printer, so in my opinion it's worth the pain. I've got other books too (like Color Confidence), but they just don't quite seem the same.

thanks so much for everyone's time and effort to answer my questions - i really appreciate this. i'm still trying to wrap my head around this....so i'm going to post anther related question now in the main forum...