• I want to thank all the members that have upgraded your accounts. I truly appreciate your support of the site monetarily. Supporting the site keeps this site up and running as a lot of work daily goes on behind the scenes. Click to Support Signs101 ...

Confused About Color Management

grafixemporium

New Member
Can someone explain to me why I can create a beautiful bright lime green in an RGB Photoshop file, send that file to my CMYK Roland Printer and the green looks just as beautiful and bright. However, with a CMYK Photoshop file I cannot get anywhere close to that shade of green? Obviously the CMYK gamut is able to physically reproduce the color... because my printer does it.

Well just print RGB files, right?

Sure, that's what we do. But the problem comes when a customer wants business cards or brochures to match the wrap we just did and we can't get anywhere close to the same color with the CMYK files we have to send out.
 

randya

New Member
The CMYK color gamut is smaller than RGB.
The ink set you are using is larger than CMYK.
So designing and printing in RGB takes advantage of your ink set gamut.

Plus may be your rendering intents settings.
 

grafixemporium

New Member
Interesting. So, how can the ink set I'm using have a larger gamut than CMYK when there are 4 ink colors... C, M, Y and K?
 

bob

It's better to have two hands than one glove.
Can someone explain to me why I can create a beautiful bright lime green in an RGB Photoshop file, send that file to my CMYK Roland Printer and the green looks just as beautiful and bright. However, with a CMYK Photoshop file I cannot get anywhere close to that shade of green? Obviously the CMYK gamut is able to physically reproduce the color... because my printer does it.

Well just print RGB files, right?

Sure, that's what we do. But the problem comes when a customer wants business cards or brochures to match the wrap we just did and we can't get anywhere close to the same color with the CMYK files we have to send out.

Don't send CMYK files out to be printed, always send RGB. If whoever is doing your printing throws a fit, change printers. Any good printer will let his process do the conversion just like you do when you send an RGB file to your RIP.

If your RIP does a far better job mapping RGB to CMYK than your other software, why do you assume that your print shop's process is any different?

I've been sending RGB files to print shops for lo these many years with the understanding being that this is the master and what you're trying to reproduce, you're the printer, you have to hit the color not me. They manage to do so without any complaint.
 

HaroldDesign

New Member
Always an interesting topic. I work in cmyk to hit G7 standards. I have to match lithograph printing digitally. RGB has the "pop" gamut, but cmyk delivers proper tonal values.
 

bbeens

New Member
Interesting. So, how can the ink set I'm using have a larger gamut than CMYK when there are 4 ink colors... C, M, Y and K?

The answer to this question is, there are many different CMYK. Your Photoshop settings are probably configured to some SWOP CMYK. SWOP is typically a small gamut CMYK. If you used the output profile from your printer as the CMYK working space in Photoshop you would see different results.

Short answer, your question is wrong. There is no one 'CMYK'. You are comparing oranges to apples.

Bryan
 

Rooster

New Member
SWOP inks are a different animal. Yes they both use the same basic pigment colors and make up the colors in your images the same way, but the inkjet inks are brighter and increase the gamut size over the SWOP inks.

What's accounting for your lack of a decent green is probably your selection of a smaller CMYK gamut. Different offset presses produce different quality printing. A non-heatset web has a tiny gamut compared to a sheetfed press or even a heatset web.

For instance using the US web coated swop v2 profile will only provide about 76% of the gamut of the US sheetfed coated swop v2 profile. Going from one extreme to the other you find that the US web uncoated profile designed for CMYK newsprint only provides 37% of the gamut that the Coated Gracol 2006 provides. Yet both are offset press profiles using CMYK inks.

If you have an understanding of the different print processes you can choose the appropriate profile. Ensuring your inkjet prints won't be more vibrant than the rest of the project that gets printed with an offset press. Yet still retaining the maximum gamut possible that allows for the different production methods to match.
 

eye4clr

New Member
There can be two different goals when printing, color accuracy OR color vibrancy. If the goal is accuracy, you use a conventional color managed workflow of respecting and using embedded profiles or assuming common working space profiles like sRGB and US Web Coated SWOP.

If the goal becomes "let's get the biggest, brightest color my printer/ink/media combination is capable of", you print CMYK files and do not apply an output ICC. This makes the CMYK number print literally and gives the most saturated color possible. There are some downsides you have to watch for like ink loads in blacks, a black dot being used too early in the lightness scale, and inaccurate color. But if you want big gamut, this is the way.
 

Rooster

New Member
If the goal becomes "let's get the biggest, brightest color my printer/ink/media combination is capable of", you print CMYK files and do not apply an output ICC. This makes the CMYK number print literally and gives the most saturated color possible. There are some downsides you have to watch for like ink loads in blacks, a black dot being used too early in the lightness scale, and inaccurate color. But if you want big gamut, this is the way.

How does printing a CMYK file without a profile produce a higher gamut than a properly profiled media? You're restricted by the media's white point and ink handling capabilities in either case. Your total ink limits define your gamut boundaries and your media can't hold any more ink in an unprofiled configuration than with a profile attached.

If you want to produce predictable results you're better off soft-proofing to the printer profile and jacking up the saturation settings until you're producing the maximum vibrancy the gamut can produce. At least you'll have decent grayscale reproduction and won't be printing purple when you wanted an indigo blue.
 

eye4clr

New Member
Your total ink limits define your gamut boundaries and your media can't hold any more ink in an unprofiled configuration than with a profile attached.
You're absolutely right. The difference is printing w/o the icc as opposed to printing properly color managed US Web Coated SWOP (or sheet fed for that matter) means the literal CMYK numbers will max out the saturation based on those ink limits/restrictions you set when you make your profile instead of being boxed in by the gamut of the CMYK working space.

Comparing my suggested non-icc workflow against a properly run Adobe RGB or ProPhoto RGB workflow starts to loose steam. The big RGB workflows can take advantage of most if not all of the output gamut if done correctly. Problem here is the majority of us do not see many AdobeRGB or other big gamut RGB files. Especially from a source that would want poppy color at the expense of accurate color. And of course converting CMYK files to big RGB spaces doesn't gain you anything but more work.

If you want to produce predictable results you're better off soft-proofing to the printer profile and jacking up the saturation settings until you're producing the maximum vibrancy the gamut can produce. At least you'll have decent grayscale reproduction and won't be printing purple when you wanted an indigo blue.
This is why you strive to set yourself up for this non-icc option when you make your own profile by making some extra effort to have proper grey balance in your linearization before profiling. I'm the last guy to suggest this non-icc output is predictable or accurate, that's the trade off.

I'm also not suggesting a non-icc worfklow for day to day use. Only as a tool in the toolbox of workflows to give big gamut when that matters more than accuracy.
 

Bly

New Member
We use a profiled CMYK workflow nearly all the time.
Onyx RGB input profiles seem to give weird posterisation in certain files.
If we get a signage job where matching anything is not important we send CMYK art and run without ICC profiles as Eye4 has suggested.

Listen to eye4clr people - he knows his stuff.
 

Rooster

New Member
Onyx RGB input profiles seem to give weird posterisation in certain files.

This is why I'll softproof to the printer profile when I'm "punching up" the color of an image. You can see on the screen when you reach the limits of the gamut and it starts to posterize.

Using 16 bit images and 16 bit profiles helps prevent this from occurring. Even if you have to switch from 8 bit to 16 bit you'll still see a difference when you're pushing the images to the gamut limits. Even though the image starts as an 8 bit image the adjustments you make after switching to 16 bit will spread the tonal range over the greater number of shades to fill in the gaps.

Your tone compression curves can also cause this to happen if they are set incorrectly or for the wrong tonal range. They tend to be pretty hardwired into the profiling apps these days though. It's been a long time since I've seen a profiling application that gives you any real control over their adjustments. Adjusting the tone compression is always a trade off and usually set to maintain the most mid-tone detail, but can be adjusted to create more tonal range (detail) in the highlight or shadow areas at the expense of the other end of the scale. Handy for times where you have an image that's mainly highlight or shadow detail. Say a shot of a girl in a white lace wedding dress or a night scene photograph with mainly shadow detail and specular highlights. It's one of those tricks that's nice to have in your bag when you need it.

As for running any sort of an unprofiled workflow, I just can't bring myself to do it. Crossing my fingers when I hit print and hoping for the best isn't something I can bring myself to do. Except for my most aggressive profiles, the printer is set to have a neutral linearization, but a swop CMY gray and an inkjet CMY gray are usually two completely different sets of numbers. What are neutral color numbers on one printer are casted on the other.

What's been suggested is merely a quick and dirty way to get punchy color from an existing CMYK file that's been saved with a small gamut profile. It doesn't produce any higher gamut at the printer and it sacrifices all control over color accuracy. I work too hard to reduce wastage at the print stage, so I'd rather take the small amount of time to make the kind of adjustments where I know I'll get predictable results. It takes less time to punch up the color the right way than selling the client on a print of questionable color accuracy after it's been printed.

I should add that I'm pretty anal about color if it's not readily apparent from the above. Most people aren't and still make good money in this industry. I've simply used my skills to carve out a niche for others like me and can charge a little more to those people. I'm never the lowest quote, but I stay busy because of that niche I've created.
 

eye4clr

New Member
What's been suggested is merely a quick and dirty way to get punchy color
Right again.

I do this maybe once every 5 or 6 months or for very unique situations. Deviating from the normal icc workflow, or any established workflow is likely to go wrong just because it isn't what you do day to day.

Again, in no way is this a suggested method when color accuracy is needed.

As far as grey balance, my method is to get close to SWOP grey balance output when testing the linearization with SWOP based CMY balance numbers in the test file. That way I'm close to the desired grey balance before profiling and it sets up the system to potentially get sell-able results without the icc if I choose.
 

Bly

New Member
You have to lookout for black and greys in a non ICC job too - these will print using only K giving you thin, banded blacks.
 

Rooster

New Member
As far as grey balance, my method is to get close to SWOP grey balance output when testing the linearization with SWOP based CMY balance numbers in the test file. That way I'm close to the desired grey balance before profiling and it sets up the system to potentially get sell-able results without the icc if I choose.

That's where our workflows differ. In order to create a swop gray balance I would have to reduce the ink limits to balance the CMY and I like to keep them as high as possible to get the best color depth.

Straying from my workflow would also remove the additional GCR that I apply to prevent/reduce metameric shifts in the neutrals.

I'm not knocking your suggestion. It does what you says it does, it's just cheating a little. Even I will break the rules and resort to some sneaky tactics to get a job out the door from time to time. I just wanted to open the discussion up a little to show that while this method works, it's far from optimal.
 
Top