Color Management From Camera to Print

I need some quality advice about the colour management process after investing

in a new HP w2207 monitor, and having to get prints made from a lab.

 

My story is: I have been using photobox.co.uk for some printing...however I

found after doing some high-key work that the prints were nothing like my screen

[old screen]...and photobox didnt help me one bit. So i ended up using a spyder2

[and also powerstrip] to calibrate my monitor but I changed all the settings to

match my "screen" with the print...thus having an eyeballed near enough view on

screen...the prints were still a bit red and saturated though and i could never

truly match the screen to the photo.

 

However...with the HP W2207 I have decided to do some research and would like to

know if I have understood the process correctly:

 

Firstly.

 

I calibrated my lcd monitor with Spyder2express [it appears slightly blue and a

little contrasty]. Dont change anything else and turn off adobe gamma in startup

[it wasnt in there to be honest].

 

In Camera [D70, D200, D2hs, and D2x]:

Set my colour workspace to Adobe RGB [so all are same color space...and Adobe is

a wider Gamut.

 

 

In RAW:

Make sure my workspace is using Adobe RGB?

 

In Photoshop CS2:

I use the "spyder2express" colour space to sort my photos out and work on them

to how i want them.

Then after they have been completed...I duplicate the file, but this time change

the colour assigned to it to the "icc" of the lab i will be using to print my

images...they use Fuji frontier 370s. This is to see how my picture will look

after printing by them...

Now the image changes quite drastically in this new colour space...so do I

re-sort [levels, saturation etc etc] the image to match the original

[spyder2express assigned colour space] image?

I can never eally get the 2 images [done in the different colour assignments] to

match fully though!

 

So is anything in what I say wrong? and especially is the last part correct? or

have I misunderstood what I am meant to do...as I am sick of getting prints back

which do not match my screen pics!

 

Thanks

From reading your post, I think you are on the right track in general, but my interpreation of what you wrote indicates it's not all right yet, so I hope the following is helpful.

 

An easy way to conceptualise profiling and calibrating is to think of languange translation. Imagine that you have something to say to someone, but you don't speak their language. If you use a trnaslator who can speak their language, then the correct message will get through.

 

This is how calibration and profiling work. Screens speak a different language to printers, so you need calibrate/profile your monitor and have a profile for the printer in order to make sure that the message gets through correctly.

 

So, you have a message (your image file) and you need to display it on screen. You profile your LCD screen and use the profile to help the screen to understand exactly what the RGB values mean.

 

Then, you adjust the image and you want to output it to the printer, and get the same color. In this case, you now need a profile that tells the printer in it's language, what mix of pigments/ink to use in order to keep the message correct.

 

I hope that is understandable. Basically, profiles are just trnaslators that let each piece of hardware know how it should output color.

 

So the process should be to profile your monitor and afterwards, the profile should, by default, be loaded when you switch the computer on. This should be automatic and you don't need to change it until next time you profile your monitor.

 

Since you are shooting RAW, it doesn't matter what colorspace you tag the file with, you can select any colorspace when you open the file in your RAW converter.

 

So assuming that Adobe RGB works for you, then select Adobe RGB in your RAW converter and then in any program that you do your editing.

 

At this point, you'll have a file where the RGB numbers refer to Adobe RGB values and your profile will convert those numbers into the correct values for the screen to output consistent color.

 

After you finish your edits in Adobe RGB, then you need to see how the file will look in print. The best thing to do is to soft-proof the image using the profile for the printer.

 

This will effectively translate the new RGB values (since editing) into numbers that the printer can understand and then show you on screen the result.

 

Soft-proofing (ie. stay in Adobe RGB, but softproof with the printer profile) is generally a better way to operate IMHO, than to convert to the output profile. Keeping the file in Adobe RGB (or sRGB for many commercial printers) will allow you to easily translate the numbers again down the track if you want to print to a different printer (which will speak a different language again and will need a different profile).

 

So, in CS2, under the view menu, setup your softproof to use the icc profile for the printer/paper combination and then before output press Ctrl-Y (Windows shortcut key) or enter softproof through the view menu and this will translate between the printer language and the screen language and show you onscreen how the print will look.

 

At this point, you'll usually need to apply a couple of adjustment layers. Normally for me this would be a curves adjustment (to apply an S-Curve) and a saturation adjustment to increase saturation. These two usually are enough, although if the print is going to be too dark, then I sometimes also use a levels adjustment layer.

 

After that, the file can be sent to the printer and if your color management is good, the print will closely match what you are seeing on screen.

 

I hope that has been useful.

 

Regards,

 

Peter

Sorry I forgot to mention...what ever i do in the actual "icc" print colour assignment i.e the Fuji frontier 370 colour space in photoshop will be how it is printed and not what I do to the actual "spyder2" assignment.

 

For example when I have the lab assigned image up and heavily change the colours hue from red to blue...or if I up the white level and make it really over exposed....will be what happens to the "print" and will be exactly how it comes back to me in the print?

 

And is the newly assigned image [saved] the one that I send to the lab?

So I don't use the spyder2 colour assignment...but use Adobe rgb? Im confused by that part.

The spider2 profile is only a translator between the RGB values in your file, and the output for the screen.

 

You don't use it in Photoshop at all, it's your operating system that needs it in order to display color correctly.

 

By default, your monitor is loaded with a factory profile when you first get it, and the colorimeter allows you to keep the color on screen the same as the screen ages and drifts away from the factory performance.

 

So, the spider profile isn't sued in Photoshop on your file. Just let the operating system use it and be happy knowing that the operating system is correctly translating the Adobe RGB values into the numbers for the screen.

 

Hope that helps.

 

Regards,

 

Peter

Just to add, you need to assign a profile to the image (in your case Adobe RGB) so that the computer knows what language you are trying to represent by the RGB values.

 

By definition, translators must translate between two languages. So you are saying, the file is speaking Adobe RGB and then the translator (spider2 profile) says, OK, I can convert that language into the screens language.

 

So, use Adobe RGB for the file and let the operating system translate those numbers for output.

 

Regards,

 

Peter

Brilliant...I think im getting the hang of this now.

 

And im right in saying then that once the [ctrl Y] convert to my printer [lab] colour space, I do adjustments also in there to how i need it to be "printed"....and thus save this file as a "for print file" and send this one to the lab?

 

The other pre [ctrl Y] is simply for viewing on the screen which has been correctly RGBed?

That last bit (ie. final conversion) depends entirely on the lab.

 

If your lab accepts Adobe RGB embedded files, then I would leave the file in Adobe RGB. When you softproof using the printer profile, the adjustments you then make to the file will change the RGB values so that even if the file is left in Adobe RGB, it will print properly when the printer RIP converts the image to pigment/ink on paper.

 

Some pro labs will accept Adobe RGB tagged files, many consumer printlabs won't and will expect a file in sRGB.

 

So, the best way is to check with your ptinter (most have the information available online).

 

If you are printing yourself, I'd leave it is AdobeRGB and send the data as is to the printer.

 

Regards,

 

Peter

You see....this is why iam confused. When I soft proof...or view the print version it is completely different [sometimes they arent so different] so if I just send the Adobe rgb file then surely it will be different? I mean the print will come back different...looking like the soft proofed veiw we had when we ctrl y the adobe rgb version!

 

This is my major misunderstanding of what is happening! If i make changes to the soft proof [the printers version of the pic] then am I not affecting the picture to how I want it to look on a print...or does the change not take place? E.g if the soft proofed veiw shows us that our pic has blown out cyan and no other saturation elsewhere then if i send the Adobe rgb version it will be printed like this...and not how I am seeing it on my screen [the adobe rgb version]?!?

 

My lab suggests using the fuji frontier 370 icc.

This is what it says...http://www.photobox.co.uk/static/colour.html...so how do I use the things it says to modify my workflow...and in order to get the perfect prints?

 

As what ever I do the prints never come back the same as how they are on screen, even using adobe rgb mode.

 

Very confusing.

http://www.photobox.co.uk/static/colour.html this woks!

Martin,

 

Do yourself a BIG favour and use my patent-pending 3-step process:

 

1. Browse to www.amazon.com

 

2. Search for Real World Colour Management by Fraser, Murphy, and Bunting

 

3. Click "Buy Now"

 

Within a month I guarantee that it will have saved you more than the purchase price - and saved you a LOT of wasted time too. The short answer is that none of us here are professional educators, and we don't have the time (or ability) to write a 582 page structured post that lays all the foundations and theory that you REALLY need to know.

 

Cheers,

 

Colin

What? With all due respect I don't want to know the theories of colour management...and I've spent enough money on the pieces...whats and where fors...now I am seeking "advice" on understanding the simply principles of a specific workflow in order to correctly match my screen to my printed image! I don't want the bloody doomsday book which wastes my valuable time, and money...does your book advise me on "my" specific problem...if not then go tout somebody else Colin...cheers

 

and ps, you can guess from my spelling that I wouldn't be able to use amazon.com...I live in England!

Martin,

 

I must admit to being somewhat surprised and taken-aback by your attitude. Not a particularly nice way to thank someone who's trying to help in my opinion.

 

It's not "my book" - and there's nothing in the recommendation for me. It just happens to be THE industry standard text on colour management. When I purchased it and worked through it - and applied it to my business, I went from being one of the ones asking colour management questions to being one of the ones answering them. However, after a while I realised that in many cases - without a proper foundation the advice given by myself any many others was often of little use. My preference these days is to encourage someone to "learn how to fish" rather than "giving him a fish that feeds him for a day".

 

You say you don't want to learn the theories of colour management - and yet you have a colour management problem. And when you solve it you'll have other colour management problems. Eventually you'll get things "kind of working" (most of the time) - but you still won't understand things like "why shadow detail disappears with Relative Colorimetric rendering" when it's fine for everything else. What I'm trying to suggest is taking a step back - get some of the fundamentals under your belt - and make much better colour management decisions because if it.

 

With regards to Amazon.com - I'm a full british citizen - living in New Zealand - and it works just fine for me (they ship books world-wide). If you have another preferred provider use them if you're so inclined - no skin off my nose one way or the other.

 

Just seems a little illogical to spend "enough money on all the pieces" and nothing on one of the few texts that can give you the help you're asking for in getting them all to play nicely together.

 

My apologies for whatever it was I said that's obviously offended you - it was my sincere intention to try and help, not offend.

Well, I do apologise sincerely, however your answer seemed a tad presumptuous. Im not just some chump who has just happened to pick up a compact camera and decided to talk to the big boys. I've been a professional for over 10 years and earn my living from it. I know about colour management, I know all aspects of photography and am capable of understanding pretty much everything that is written in said book. What I don't understand is a tiny specific rule which has illuded me thus far...and caused me no end of expense and bother. What I don't need is someone who appears to be snobbishly pointing me to their 'patent pending' 3 step plan...in order to do myself a BIG favour!

 

Next time perhaps just explain who you are and tell me that there is a chapter in a certain book. Your comment was quite arrogant. I now see that this was not your intention so please accept my apologies.

 

I would rather have people giving me some advice that I need in this specific case...and not some general throw of paint on a wall...I need a fine brush to paint the details!

 

I understand colour...I understand light...etc...these are my everyday tools. However this thread I started is asking about profiles and why such a profile in photoshop on a certain monitor and a lab would be in so much conflict. It is a thread seeking advice on why a colour does this or a mathmatical ideal is placed on my digital instructions so-on. It is finding the wheat....and I don't need a 580 page read to seperate the info from the chap!

 

I heartedly apologise to you but when I read your comment it did appear to me as though you were talking down to me. I realise now that you weren't, so thanks for the advice, truly.

 

I'm a practical person...I work things out through knowledge and understanding, it's just in this case I need the quick solution to my headache...as i have a client waiting for photos!

 

Once again sorry about my ignorance.

 

Martin

If I understand your flow correctly it appears you are converting your image in Adobe RGB to the printer profile and that is what you are sending to the printer.

 

You should use the printer profile to soft proof the image and make any required changes to the image (still in Adobe RGB profile) and then save a copy with a different file name (still Adobe RGB profile) to supply to the printer. This image is only for that specific lab and if you need to print later using a different printer start with the unmodified image and do the same using the profile for that printer.

 

This is assuming your lab wants the image in Adobe RGB profile - the professional lab I use wants Adobe RGB for their Lambda printer, but most want sRGB for the Fuji Frontier 370.

Martin,

 

Reading through all of the posts a couple of things haven't been mentioned that will make a big difference to how the soft proof displays ...

 

1. SOFT-PROOF DISPLAY OPTIONS (RWCM Page 369/370)

 

I normally get the best results by turning BPC (Black Point Compensation) on, but leaving "Simulate Paper Colour" and "Simulate Black Ink" Off

 

2. RENDERING INTENTS (RWCM pages 88-92)

 

You'll get big chages depending on which you use. Saturation is inappropriate - Absolute Colorimetric would be an unlikely choice - leaving Perceptual & Relative Colorimetric. "Normally" perceptual will do a better job of maintaining shadow detail if needed (say, perhaps detail of a grooms suit) - otherwise Relative Colorimetric is my usual first choice.

 

Forgive me if you've resolved the profile question already (it's 1:34am here and my eyes are a little weary), but following the link you gave earlier, Photobox appear to be saying that they want prints in sRGB, and that custom profiles are unavailable at this time:

 

http://www.photobox.co.uk/quality.html

 

If so then it's important to convert to sRGB as your final step - if you send in prints tagged or embedded adobe RGB 1998 (and the printer strips - ie "ignores" them) then you'll get prints looking darker and with more subdued colours (because aRGB is a bigger colourspace, lower absolute numbers will be assigned to RGB values for colours of equivalent saturation - when these lower numbers are misinterpreted as sRGB instead of aRGB lower saturations and intensities result).

 

If they are only printing to an sRGB gamut then you could probably save yourself a few steps by assigning sRGB to your images during RAW conversion - the few extra colours that aRGB includes are not particularly memory colours and as such the rendering intents will swap them for something so close no one will notice (our eyes are not particularly sensitive to yellow anyway, which is one of the main ones).

 

Hope this helps.

Martin, I'm in the UK also, the book Colin mentions is available 1/2 price on amazon UK http://www.amazon.co.uk/exec/obidos/ASIN/0321267222/ref=ord_cart_shr/203-3291206-6842334?%5Fencoding=UTF8&m=A3P5ROKL5A1OLE, I expect it may also be available in your local library - I think I've seen it in mine and I'm down in rural Devon.

"I know about colour management...

You may know about color management but you aren't implementing it well.

 

For starters it's good that you are actually calibrating and profiling your monitor. That's basic. Whether you've got the best tool in that price range is debatable but I'm not going to go there.

 

And While Bruce Fraser's book is terrific, it was really written for people printing commercial jobs on a CMYK printing press. For photographers I think Andre W Rodney's Color Management f may be more accessible.

 

Back to your problem

 

Once you've created your monitor's profile and it has been stored in the correct location for your ICC aware programs to access, you don't do anything with it except let it run in the background and update it every couple of weeks or so or as you start a big job.

 

Since you are shooting using your camera's raw format, just be aware that the only things setting a color space on your camera has on are what you see on the cameras LCD, if you are creating in camera JPEGS, or if you are using your camera manufacturer's raw processing software, telling it what settings you used as an incoming baseline. In all raw imaging software is where you actually assign color space and bit depth, sharpening and other processing parameters. These are only applied once you decide to output them as a JPEG, TIFF or PSD format. And still, your original data is kept pristine in case you want to later revisit those choices.

 

<I>So I ended up using a spyder2 [and also powerstrip] to calibrate my monitor but I changed all the settings to match my "screen" with the print...thus having an eyeballed near enough view on screen...the prints were still a bit red and saturated though and i could never truly match the screen to the photo. </I><By changing the monitor settings you very effectively invalidate the profile. Changing those settings makes everything else you do color mismanaged. So if you are still doing that stop it! <P><I>In RAW: Make sure my workspace is using Adobe RGB (1998)? </I><P> if that is the color space you want to work in, yes. <P><I>In Photoshop CS2: I use the "spyder2express" colour space to sort my photos out and work on them to how I want them. </I><P>What happens by doing that is you are applying the corrections made for your monitor?s performance (the monitor's profile) on top of those corrections which are already and automatically being applied in the background. Leave your Photoshop working space set to the color space your TIFFS, JPEGS or PSDs were created in.<P><I>

Then after they have been completed...I duplicate the file, but this time change the colour assigned to it to the "icc" of the lab I will be using to print my images...</I>Almost perfect but instead of Assigning the profile the lab provided, Convert to there profile. In the conversion stage you have a couple of options> Since you are converting fro ma working space to an output profile, choose Perceptual as the rendering intent and turn on black point compensation. <P>

Try the above as your workflow and see if yields better basic results.

Awesome help...thank you so much for taking the time. It is all very much appreciated. I will buy the book I think as you have mentioned a few things Colin which do concern my workflow and need attention.

 

I am struggling to understand the soft proofing thing...I feel so retarded for not getting it. What I can't get is if I change the soft proofed [duplicate] or the original ...so when soft proofed it appears correct. I also have to send a disk of images to the client which again is another conundrum :(

 

Martin

For the disk you send to the client:

 

First find out how color management savvy they are and what the images are to be used for. My general solution is to send level 10 or Level 12 JPEGS in the sRGB color space if they aren't really ready to handle color management. that one of the beauties of raw capture: you take the original and target it for different users and purposes and output methods with out losing anything.

 

Listen, you aren't the only who finds this stuff hard to get. It's a bit of an arcane science and a bit of an art as well.

I think i'm getting it!

 

So I am wrong in thinking that once I convert to my lab's profile [fuji frontier icc] I just upload to their website and get them printed...but what I have to do after converting is then adjust everything again [levels saturation etc] till the image is how I want it...save it as a [print] version of the image and then give them that to print?

 

This is the remaining question I can't seem to understand.

 

[My general solution is to send level 10 or Level 12 JPEGS in the sRGB color space] I don't really know what you mean by level 10-level 12?

Soft-proofing is a way of combining the profile conversion and the post-tweaking, so all you have to do is edit the picture with soft proof enabled, then convert it.

 

JPEG levels (this is Photoshop nomenclature) indicate the amount of compression that is applied to the image. The lower the number the greater the compression. Level 10 is a good compromise between image quality and file size.

Why is it not the case that you process the picture with the srgb or argb, then when you have the final image, why does it not work that you assign the lab's icc...and that information tells the lab how your picture looks as? It seems to me, if I've understood this correctly that assigning/convert to the lab space just means another picture to process!?!

If you meant to ask why it does not look the same after you convert from one profile to another, the answer is that different devices have different capabilities (gamuts) so compromises have to be made.

 

You want to convert to, not assign, the lab's profile.