Question for edgework and anyone else...
Newbie2006
Registered Users Posts: 5 Beginner grinner
Hi there, I'm new here. Been lurking for a couple of days and I'm learning a lot thanks to you all.
I have a few very basic questions to do with colour spaces, photoshop setup and basic tonal adjustments. I'd be interested to know edgework's opinions on this since (as far as I know) he works in the pre-press industry and has a thorough understanding of CMYK and preparing for print.
1. Which is the best colour space to work in assuming images will end up printed in magazines, newspapers, advertising use?
2. When correcting colour/tonal range in RGB, do you view your working CMYK proof setup?
3. When targeting your image in RGB, do you set your highlight/shadow points (eg: C5 M3 Y3) first or do you drag sliders until detail begins (alt-drag method in levels) and then compress tonal range at the end?
I think I have more questions... I'll post more here when I remember what they are.
Thanks in advance, I really appreciate any comments from everyone.
Cheers
I have a few very basic questions to do with colour spaces, photoshop setup and basic tonal adjustments. I'd be interested to know edgework's opinions on this since (as far as I know) he works in the pre-press industry and has a thorough understanding of CMYK and preparing for print.
1. Which is the best colour space to work in assuming images will end up printed in magazines, newspapers, advertising use?
2. When correcting colour/tonal range in RGB, do you view your working CMYK proof setup?
3. When targeting your image in RGB, do you set your highlight/shadow points (eg: C5 M3 Y3) first or do you drag sliders until detail begins (alt-drag method in levels) and then compress tonal range at the end?
I think I have more questions... I'll post more here when I remember what they are.
Thanks in advance, I really appreciate any comments from everyone.
Cheers
0
Comments
As to what output format to save your document as, I would think you should ask the magazine/newspaper/advertiser how they would prefer the file.
2. If you know the specific ICC profile your magazine/newspaper/advertiser is going to use, this would be a very helpful step, I think. If you don't, you might be spinning your wheels.
3. Use whatever method makes your image BETTER (by whatever metric you decide is appropriate) and that you can integrate into your workflow.
In addition to the issues covered in that post, Photoshop will also try to make black conform to the specs that are programmed into it's profile. If your image is decidedly warm (intentionally, that is) with shadows that lean towards red and yellow, they will become much colder in comparison with a straight conversion to CMYK. It is often necessary to use a selective color adjustment, adding magenta and yellow back into the blacks to create the effect.
And, of course, if your file is going to a newspaper, you have a totally different animal, one that brings a whole host of additional problems that have to be dealt with. Newsprint can't take as much ink, the ink that you do use will spread and make everything darker than it should, you have to restrict the use of black and most highlights blow out to nothing below a certain range (5% - 15%, depending on who is writing the specs.).
Often, but that's just because I still think in CMYK. It's what I learned. I've known designers who have never strayed beyond RGB and they are quite comfortable with RGB numbers. It's what you're used to.
I find that for skin work, CMYK is a more useful reading, in that so much of the shadow information has been shuttled off to the black plate. The CMY plates are much purer in the areas that count most, highlights to three-quartertones. When I look at RGB numbers, I have to run through a couple of translation algorithms in my head. CMYK values tell me immediately if something is high or low.
Actually, I'm not quite sure about the method you are describing. I avoid Levels whenever I can, which is always. Lately I've been doing most of my cast removal in LAB, though that can be dangerous due to LAB's ability to add tone to areas that in RGB or CMYK would remain neutral. If I'm in RGB and the task is to specifically balance highlight and shadow values, I'll glance at a CMYK sampler, but it's much more accurate to bring the RGB values into sync. If my shadow point is below 10 or 11 and the three values are matched, that will work. When converting Photoshop will separate the colors according to dMax and black plate generation values that are set. Hightlights are a little trickier. If you use the standard CMYK highlight, eiither 4C, 2M, 2Y, or 5C, 3M, 2Y, or even 5C, 3M, 3Y (which tends to print warm), your RGB values won't be neutral. It will be something along the lines of 238R, 240G, 242B. On the other hand, setting values in RGB that are dead neutral will translate into CMYK that have Cyan slightly elevated as opposed to the M and Y, which will be equal, more in line with what you want.
The point is, using CMYK readings to set highlight points in RGB isn't reliable. Keep it somewhere in the range of 240-245 per channel and keep all the channels equal. You do better trusting Photoshop to do the conversion, but you may still need to tweak after you are in CMYK, depending on whether the highlights are running warm or cool in relation to the rest of the image.
—Korzybski
Let me explain a bit more about my situation. I studied Desk Top Publishing a while ago (1999) concentrating on the usual packages (Quark, Illustrator, Freehand, Photoshop). After finishing up at college I went on to work as an apprentice at a Repro house learning bits and pieces from all the various departments - mostly page layout, typography and scanning/retouching.
In that time we were using Photoshop 3,4 and maybe 5.
Anyway, to cut a long story short, we worked in CMYK only and I was taught to find the highlight and shadow areas and assign them (with curves) a particular dot percentage eg: C5 M3 Y3.
Also, we didn't use colour management ie. colour spaces like Adobe98, sRGB etc.
Consequently, I didn't learn the colour management side of things.
Now I'm in the situation where I want to work with RGB images where the final output is unknown ie. magazines, newspapers etc.
The conversion from RGB to CMYK doesn't take place on my side, I simply supply images to various agencies.
In that case, is it wise to work in RGB yet read/use CMYK numbers when colour balancing?
Should I be working in Adobe98, sRGB, ProPhoto when the final print method is unknown?
The myth of color management is that there exists, somewhere out there, a device-independent color space to which we can refer for all our conversions. Nice wish. I still hope Santa Claus will drop down the chimney on Christmas. I've got as much chance of experiencing one as the other. The only way we have access to this device-independent space is through a DEVICE!!!! It's a chimera, a ghost, one more grail-quest that will remain unfinished.
Color management is possible within a given work flow. Once you send your file out of your immediate oversight, it's up for grabs. If separating it into CMYK is not your responsibility, then spend no more time worrying about color management. Do everything you can to work in a color compatible environment: calibrate your monitor, don't import any profiles built by a guy named "Slick" and if you do, don't use them. If you know your file is going to the web, you might go ahead and convert it to sRGB. But if you aren't handling the print end, it's pointless to worry about it since there's nothiing you can do about it anyway.
—Korzybski
I would add to Edgework's comments that even if your printer wants sRGB (as many do these days), you can still get a profile for their printers and learn how to use soft proofing in CS. That will help you see when your colors are out of gamut for the destination printer and/or see how it's likely to get converted when it prints.
Whether to work in aRGB vs. sRGB is a matter or personal preference with advantages to each. I personally have experimented with aRGB but haven't found the advantages of aRGB to outweigh the extra hassles of always having to convert to sRGB and checking to see if I'm still in Gamut for sRGB anytime before I send to a printer or put it on the web or send it in email, etc... There just are so many non-color managed displayers of images that if an image you share isn't in sRGB, it's probably going to look bad outside of your environment. I respect those who have made the aRGB choice, it just hasn't looked like the best for me.
Homepage • Popular
JFriend's javascript customizations • Secrets for getting fast answers on Dgrin
Always include a link to your site when posting a question
Why do you avoid Levels?
jfriend: I've heard so many arguments for and against Adobe98 vs sRGB that I'm still not sure either way but I've been using the bigger colour space just to be safe
Curves can do everything Levels can, but Levels can't do all the things that Curves can. Therefore many Photoshoppers just use Curves in their workflow as it is a more powerful tool. In addition, it is easier, IMO, to accidently throw away important pixels in Levels-- especially in clipping the highlights. Having said that, Levels are a fast and easy tool to use, and can be a lifesaver when you just need to get something "good enough."
If you have red highlights and green shadows, Levels will be hard pressed to move in different directions in different ranges. It's what curves are meant to do. Pulling anchor points in curves approximates the effect of a levels move. But most corrections aren't that straightforward. You might need to raise the midtones and lower the three-quarter tones and really lower the highlights... weird stuff like that.
For the same reason I don't use Color Balance, Brightness/Contrast either. They're all subsets of curves.
Selective Color, and Hue/Saturation have very definite uses, and can often achieve effects that are more complicated in curves.
Levels is useful for making masks from a channel. That's about it.
—Korzybski