sRGB or Adobe RGB?

Dear PN,

 

I have been using Miller labs and I have been very happy, however,

lately I am getting back images that I slightly under saturated and

bleak. After carefully going over those orders I have realized that

these were sent to them in the Adobe RGB format.

 

Please correct me if I am wrong, I thought that Adobe RGB has a larger

color gamut and is ideal for labs to print compared to sRGB.

 

Which mode profile do you use when sending digital files to your

favorite labs?

 

Sincerely,

Greg

my lab only accepts sRGB files.

I use the same lab and had the same problem. What happens, from my understanding, is they convert it to sRGB. That is the profile their printers are set up for. Also in 8 bit. I don't know if when they convert it there is a problem or not. They regularly have articles in their newsletter discussing the optimal way to send files to get the best possible prints. Just go to their customer only section and on the left hand side of the page is a newsletter link. Or else you could call and talk to one of the customer service people who should be able to steer you to a better explanation.

get their icc profile and profile your monitor. spyder's are so cheap now.

Eric,

Thank you for your advice, but my monitor is weekly profiled with the spyder, so clearly this is not the problem. As to the ICC profiles, this is something to look into. But even than you are still left with the same question what is better sRGB or Adobe for the amount of colors in the final print?

 

Greg

save in srgb if you are not working in their profile. millers should be able to give it to you though. i'm looking for some great threads in the digital darkroom for you right now.

<a href=" http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=009CG6">here's one</a>

<a href=" http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=009VhL">here's another</a>

Dry Creek probably has your Millers icc profile <a href=" http://www.drycreekphoto.com/Frontier/FrontierDatabase.htm">here</a> too.

Eric,

Thank you VERY much this link was increadibly helpful.

Greg

no worries man. what i lack in memebership payment, i make up in time with...

There are really 3 major color formats with subcategories with-in each: sRGB, RGB, and

CMYK.

 

sRGB is a truncated color format used for internet exchanges, and many photo printing

labs. The reason many labs use sRGB can vary, but chief among those reasons is that a

vast majority of digital cameras are set to sRGB as the default color space ... and ...

many labs offer internet transfer services where a less information dense format is more

efficient in terms of time and space... so a J-Peg in sRGB is desired. Most of the images

brought to Wall-Mart and other mass labs are from P&S family cameras default set to

sRGB. Even the D20 ships with sRGB set as default and has to be re-programed to RGB (at

least mine was). Journalist's cameras like the 1D, 1DMKII are usually set to sRGB for quick

wireless transfer to publishers for conversion to CMYK.

 

CMYK is the general color format for publication printing like magazine ads and editorial

images. CMYK is the designation because that corresponds to how printing presses work

... they lay down each of the 4 colors onto paper one at a time (four color printing). At our

ad agency, everything is calibrated to CMYK for that reason. the Quark design program, all

computer work spaces, and our Epson 3000 and 4000 ink-jet printers are all calibrated to

CMYK because the end proofs have to match what the printers are able to do. So, the color

profiles in each art director's photoshop proofing designation is CMYK.

 

RGB is a more expanded color format used primarily for photographic reproduction. Most

desktop printers (ink-jet and dye-sub) are equipped to deal with RGB.

 

The trouble arises because none of the three color formats match each other when dealing

with any given image on a computer screen. sRGB and CMYK tend to be closer, but RGB is

quite off in color when directly converted to either sRGB or CMYK.

 

But, you do know what the end means will be, (your RGB desktop printer, a lab supporting

sRGB, or CMYK if you are correcting an image for publication).

 

If your primary objective most of the time is to desktop print, you DO NOT want to

calibrate your screen to match what your desktop printer is doing. You want to do exactly

the opposite. You want to calibrate your screen and save that profile so you can select it as

your working space profile and proof profile when working in Photoshop. Then that

information is what is sent to the desktop printer so the images will match what you are

seeing on screen. This is how my system is set up because I print thousands of my own

images for weddings and advertising presentation layouts.

 

If you only are doing work for web viewing or wireless transfer, or to be sent to a lab

working in sRGB, then you want to select that as your default work space and PS proof

profile.

 

If, like our art directors at work, you prep images for publication, then CMYK becomes the

center piece of your color profile preferences.

 

ONE IMPORTANT THING FOR THOSE WHO SHOOT DIGITAL: check what your camera's color

space is set to. It is highly probable that it's set to sRGB unless you changed it. If you

primarily print on a desktop printer, change it to RGB. If you primarily post on the web or

send off images to a lab leave it on sRGB.

 

Hope this helps (if even a little bit) to understand what this is all about.

 

Here's an experiment: one image corrected in each of the three color formats, composited

into one and saved for web. Let's see if the sRGB one is the better of the three when

viewed on the internet. (I think this will work unless it all got transferred into one color

format when converting to web??? ... we'll see )<div></div>

oops, I forgot to convert the whole thing into sRGB for web upload:

 

here it is again...<div></div>

all 6 photos are exactly the same.

No they're not.

 

The lab I use for work prints on a frontier using sRGB and that is what I shoot in. Apparently if shooting RAW then it doesn't make a difference anyway though I would like that verified. The Lab I use for my landscape stuff uses Adobe RGB so I will work in RGB when doing that kind of stuff. Photoshop lets you convert from one to the other very easily, so unless the work is critical, it doesn't really matter if you work in one format and only change at the end depending on the usage of the file.

The shots are different, but only marginally.

 

My advice is to send files to the lab in the working space they are using vs complaining about it and adding another variable to the equation. The lab's either going to dump it down to sRGB, or I can do it for them and at least keep it under my control. I prefer to keep it under my control.

 

I shoot in Adobe with my 10D when I'm dealing with a strong gamut range, and can post more extreme and obvious examples of why AdobeRGB is the better color space. However, for general portraiture I rarely use AdobeRGB simply because lack of color saturation is a greater problem than too much.

 

If I'm working with a LightJet shop that knows what they are doing I'll be happy to keep files in the wider gamut profile.

<I>all 6 photos are exactly the same.</I><P>Not on my calibrated and profiled Samsung

SyncMaster 753df CRT monitor (that isn't an endorsement -- I'm eager to upgrade)! But as

Scott pointed out the differences are slight but they are there<P>

 

My way of working right now is to shoot Adobe RGB if I shoot a .JPG or if shooting RAW

files assign and save the processed TIF version in Adobe RGB (1998). If the the print is

being made by a lab that I know uses sRGB I'll duplicate the file and convert the duplicate

to sRGB (or to the lab's profile) and send over the converted duplicate.

Marc, I surely can count on you to give the most comprehensive overview. I would like to thank everyone who contributed to this thread.

 

SIncerely,

Greg

Marc's overview contains mostly good practical advice, but I suspect a couple things got slightly confused in quick writing.

 

You don't want to set your working space to your monitor profile. Calibrate and profile the monitor (creating what is called a "device profile") and then it should be set at the OS-level as your monitor profile. In PS (or wherever), your working space (as Marc said) should be a well-behaved editing space, such as Adobe(98), a.k.a. aRGB, or sRGB. [i recommend aRGB, but no one listens to me anyway.] From there, you can convert to another specific device profile (say, a printer) when necessary for output. So, do exactly what Mark is saying about not calibrating to your printer and for calibrating/profiling the monitor, just don't use that as your working space.

 

Many labs do indeed prefer sRGB, because it more closely matches the output range of their printers. However, I don't think sRGB is any more efficient in file size for transfer (8 bits is 8 bits, regardless of the color space defined in those bits). Also, as has been noted on these page by people smarter than I, no device exactly matches sRGB, so the ideal is for the lab to have a specific device profile for their specific device.

 

Mark's definitely right about most inkjets thinking in RGB. I think it was Bruce Fraser (or someone like that) who observed that these CMYK and CcMmYK devices behave so assiduously like RGB devices that we are forced to treat them as such.

 

Onward.

so...when do actually assign profiles? in the begining of your workflow or at the end? and do you soft proof or not? i really have a hard time with all the theories.

If the image comes with a profile assigned, you don't ever need to assign profiles. You make your edits and save a master file. Then, you can soft-proof in your output space (or Convert to it, if you prefer), make final adjustments for print, and save as a different filename so as not to throw away your master.

> There are really 3 major color formats with subcategories with-in each: sRGB,

> RGB, and CMYK.

 

Not really. There are at least three "major" color models; RGB, CMYK and Grayscale. Within

each color model there are literally thousands of different color spaces. In RGB, every

device that produces or output's an RGB file and is different from each other is a different

color space. Adobe RGB (1998) is a different color space from sRGB from Wide Gamut RGB,

from Epson 2200 Matt RGB etc.

 

> sRGB is a truncated color format used for internet exchanges, and many photo

> printing labs.

 

Not really. The sRGB color space is a synthetic color space that is supposed to mimic the

"typical" PC display. It's a very simple space constructed of nothing more then a specific

gamma, white point and three chromatisity values.

 

>The reason many labs use sRGB can vary, but chief among those

> reasons is that a vast majority of digital cameras are set to sRGB as the

> default color space ... and ... many labs offer internet transfer services

> where a less information dense format is more efficient in terms of time and

> space...

 

Most are lazy, don't want to handle specific output profiles (unique color spaces) or handle

color conversions on the fly for speed and simply ASSUME all the data sent to the device

begins in sRGB.

 

 

> so a J-Peg in sRGB is desired. Most of the images brought to

> Wall-Mart and other mass labs are from P&S family cameras default set to sRGB.

 

They produce a unique color space in RGB that's somewhat close to sRGB. Or to put it

another way, when you tell a color managed product like Photoshop (or a printer that

expects sRGB) that the numbers are in sRGB, you get a reasonable color preview or output

based on these assumptions.

 

> Even the D20 ships with sRGB set as default and has to be re-programed to RGB

> (at least mine was). Journalist's cameras like the 1D, 1DMKII are usually set

> to sRGB for quick wireless transfer to publishers for conversion to CMYK.

 

Bad idea considering the CMYK gamut of even something like SWOP exceeds the sRGB

gamut in Cyans and greens.

 

> CMYK is the general color format for publication printing like magazine ads

> and editorial images.

 

There are literally thousands of flavors of CMYK like RGB.

 

> CMYK is the designation because that corresponds to how

> printing presses work ... they lay down each of the 4 colors onto paper one at

> a time (four color printing).

 

Some CMYK color spaces with specific recipes for a specific press.

 

> At our ad agency, everything is calibrated to

> CMYK for that reason.

 

Which CMYK? There are as many different flavors of CMYK as there are CMYK devices.

There is a very specific recipe called TR001 which is a specific flavor of SWOP based upon

900 odd spectral readings of a press that the SWOP committee has said is producing SWOP

conditions.

 

 

> If your primary objective most of the time is to desktop print, you DO NOT

> want to calibrate your screen to match what your desktop printer is doing.

 

You never calibrate your display device to an output device (you can't). You calibrate it for

a specific aimpoint (eg 6500K, gamma 2,.2) and then you load an output profile so an

application like Photoshop can produce a soft proof based upon that device behavior. It's

simply impossible to "calibrate" a display to a CMYK device. They only deal with a specific

recipe of RGB.

 

> You want to calibrate your screen and save

> that profile so you can select it as your working space profile and proof

> profile when working in Photoshop.

 

Nope, you never want to use a display profile as a working space. That defeats the entire

way in which Photoshop divorces the actual display from how you edit your files. You pick

a Quasi-Device Independent working space like Adobe RGB (1998) (or sRGB) which is

based on NO real world device. They are synthetic color spaces for that very reason.

 

Andrew, I think Marc was trying to simplify :-)

image one color space, someday :)

If the 3 images in my second post look the same to you, I'd look into calibrating your

monitor.

 

A simplification? Yes, but I even went beyond the scope of the original question. Of course

there are a zillion iterations of CMYK for example, but the principle is based on 4 color

separations for press, which is enough info to get the idea. sRGB, RGB and CMYK are the

practical color areas photographers work appears in. No matter how anyone tries to tech it

up and make it seem mysteriously complex (often correctly), it's the basic categories of

practical applications we use.

 

No I didn't make an error in saying that I calibrate my monitor and then select that as my

working space ... then also select that as my PS proofing profile. Right or wrong, it's a

decision based on consistent results.

 

Every single device I use from a 2200 ink-jet, to an 8500 dye-sub, and all of my different

scanners are dead nuts the same in output... and exactly match what I see on my monitor.

Not only that, but the local lab I send my clients to, makes "as is" 4X6 prints for them @

.29 ea., that are also dead on in color match. The owner has told me he loves my files

because the clients never complain or return prints for reprinting.

 

 

I used to use Adobe RGB 1998, but since doing this have not once had a print that didn't

match what I saw on my monitor. I may be dead wrong (highly likely), but I'm sorry, with

these results I'd rather be wrong than be technically right and struggling with print

matching like I used to always be doing.