ICC question

Started by baker7, August 23, 2019, 06:50:12 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

baker7

Good morning~

Coming back to printing after a hiatus I'm amazed at how little I knew about what I was doing. Yet I'm amazed at how little I still don't know and why my boss would call me back.  Anyway I'm in sublimation and I just learned there's something called an ICC.

I think that's what's making our colors come out all wacky because we don't use the embedded profile from our clients' images when proofing on digital printer. And when it goes out to our offset colors are off. Did that make any sense?

We just paid our ink supplier $500 to calibrate a CTP curve to match our offset but it's still off. So I have to go back to the file in Photoshop and tweek curves, burn plates and hope for the best.

My question still is is this normal? Seems for every job I have to tweek the curves a different way, thus no consistency. I'm guessing ICC is the culprit?  It would make life so much easier to have the colors come out the same as the file and eliminate the tweeking part of it!

Sometimes it takes us 3-5 hours on a single job just color matching and production comes to a halt! All the paper, ink and time wasted!

JohnO

#1
For our 4-color proofing and the 4-color Komori press, we calibrate those using the GRACoLCoated1_ISO12647-7_ControlStrip2009. The plates are curved for the press to match that GRACoL industry standard. We strip out any profiles in the customer files.

This way, our jobs should print looking the same as they would at any other shop in the country using that standard.

Make sure your platesetter and press aren't "drifting" which makes them a moving target to try to hit. Chemicals in the developer, blankets or different/inconsitent ink densities at the press, are different ways in which your target could drift.

Joe

I guess my first question is what are you trying to match your press with? A proof supplied by your customer? A proof you make? Or heaven forbid someone's monitor? Basically a hard copy proof (printer) has to be calibrated to match your press and not the other way around. So if it is a customer proof good luck with that as you have no control over their printer. If it is your proof (printer) at least you have control over that. But you certainly aren't going to get that done by someone else for $500. Basically the steps needed to match a press to a hard copy proof is to run a test page such as the G7 test files on the press. Then you have to take readings from the test run and use the printer calibration software to input the data into the software and make a hard copy proof and compare it to what was run on press. You would probably need to do this more than once to get the printer dialed in to match the press. This is kind of over-simplified here and you have to know what your are doing and your company has to be willing to take the time, money, and materials to make it happen. IMO it is better left to hiring a color professional to do this.

As far as "$500 to calibrate a CTP curve to match our offset". Usually what you want is a linear plate, though not a must, where the values of the input data is matched by the output of the plate. Then I think what you mean by the CTP curve is a calibration curve used to allow for dot gain as an offset press can have dot gain anywhere from 10% gain to 25%+ gain depending on the paper being used. A coated high glossy paper will have less dot gain than for, example, uncoated newsprint which acts like a sponge.

(And I agree with JohnO...strip out any ICC profiles used by your customers. Customers have no idea what they are doing and for sure any profiles they use are not going to make your press magically match some proof from somewhere)
Mac OS Sonoma 14.2.1 (c) | (retired)

The seven ages of man: spills, drills, thrills, bills, ills, pills and wills.

baker7

Thanks for the reply guys.  We do strip the ICC profiles from the customers' files, but we have to get the final product to match their colors exactly that come out of their printers.  (We get a sample printout from them that we heat press on a poly fabric - even this causes variation but negligible.)

So from the jpg file I get from them I print from our printer, heat press on fabric and compare it to the customer's sample.  Then I go into photoshop and tweek the curves to try and match exactly.  Then I make the plates and pressman tries to get the colors up to match yet again.  During this process I go back into photoshop and tweek curves and replace certain plates when there's no way to match with ink. 

I guess the problem lies with calibrating the curves.  The $500 guy tweeked the CTP curve using this color reading gadget from the test file printed from our printer.  And our CTP curve is definitely not linear.  I guess that explains why for example cyan darks are too dark and lights are too light on some files and flip flopped on other files. 

Another problem is lot of our customers are using the badly tweeked printer calibration profile in their digital printers.  Adding to that they don't use the same amount of ink as us.  They crank it up like 200% because their customers want bright vibrant colors and there's no way we can go that high from the files they give us.  We have to do it in offset by pouring tons of ink which screws up all the other colors so that I have to bring the colors way down so that it lines up.

SO FRUSTRATING

Anyway thanks for the refresh on where our problem lies.  So in short we're screwed... until we get a linear curve from our printer to match the CTP.  We've been trying to match the CTP to our printer   https://www.b4print.com/Smileys/default/facepalm.png:facepalm:


Joe

Quote from: baker7 on August 23, 2019, 10:52:50 AM
Thanks for the reply guys.  We do strip the ICC profiles from the customers' files, but we have to get the final product to match their colors exactly that come out of their printers.  (We get a sample printout from them that we heat press on a poly fabric - even this causes variation but negligible.)

So from the jpg file I get from them I print from our printer, heat press on fabric and compare it to the customer's sample.  Then I go into photoshop and tweek the curves to try and match exactly.  Then I make the plates and pressman tries to get the colors up to match yet again.  During this process I go back into photoshop and tweek curves and replace certain plates when there's no way to match with ink. 

I guess the problem lies with calibrating the curves.  The $500 guy tweeked the CTP curve using this color reading gadget from the test file printed from our printer.  And our CTP curve is definitely not linear.  I guess that explains why for example cyan darks are too dark and lights are too light on some files and flip flopped on other files. 

Another problem is lot of our customers are using the badly tweeked printer calibration profile in their digital printers.  Adding to that they don't use the same amount of ink as us.  They crank it up like 200% because their customers want bright vibrant colors and there's no way we can go that high from the files they give us.  We have to do it in offset by pouring tons of ink which screws up all the other colors so that I have to bring the colors way down so that it lines up.

SO FRUSTRATING

Anyway thanks for the refresh on where our problem lies.  So in short we're screwed... until we get a linear curve from our printer to match the CTP.  We've been trying to match the CTP to our printer   https://www.b4print.com/Smileys/default/facepalm.png:facepalm:

Getting a linear plate is the least of your problems. It gives me a headache just reading your post. :D Your boss should write a book on how to NOT do color management.
Mac OS Sonoma 14.2.1 (c) | (retired)

The seven ages of man: spills, drills, thrills, bills, ills, pills and wills.

Tracy

It's time for boss to bring in a professional, he will save money if he does

baker7

Joe - yeah, something's screwy but then again the boss says sublimation printing is different than other commercial printing.

Tracy - hey~ long time! Yeah I agree, but he has no money. I started a month ago and he already owes me $3000 for paying his electricity bill lol. I'm laughing at him and myself. I realized I shouldn't have done that after not paying me my last paycheck lol

mc hristel

I was in a similar situation to you baker a while back and unfortunately, getting color right takes a lot of time and resources and your boss won't want to waste all that paper and press time to do it right. Back then I was getting pressure to calibrate our proofer to match the customers files. "Ok, but it's pointless if the press can't actually reproduce that color" would be my response and I would move on. It didn't help that the prepress area had these terrible yellowish safety lights that screwed up any ability to see accurate color.

Anyway, ideally, what you want is to match your proofer to your press/final output device.  The person that was brought in sort of attempted to get you closer there, maybe but probably just messed things up more.

To expound on what Joe said a bit (sorry if this gets a bit long):

You want to run test sheets on your press (or whatever) with the device settings (ink flow, curves, etc.) neutral, or in the middle of the presses range. That will give you a baseline for what the press can actually print. Hopefully your test print has a series of tint boxes that go from a 0-100% screen (5-10% increments) which you need to measure with a densitometer. You are checking to see if the expected % matches the printed %. So then, if the 70% doesn't read 70% on press, that's where you need to use the CTP curve to bring the actual and expected densities in line. That process is repeated until the press and expected densities match. If you print on different materials, you may want to have separate profiles which means you would have to go through the whole process for each different thing you print on.

Once that is all done, then you adjust the proofer output to match the calibrated press sheet. In your situation, it probably doesn't need to be too exact so you can just do this by eye. Output a proof and check it to the press sheet. +Y, -C, etc. Again repeat until the proof is in line with the press. You want to do this in good light (under a full spectrum light bulb, or even outside can work).

Short of making your customers match their files to what they actually want, there is unfortunately, no way to get around all the color work you are doing in the files to get them to reproduce as needed, but you will know that once your proof matches the customers, it can be reproduced on your press without having to jump through hoops.

Hopefully all your prep time is getting charged back to the customer. If not, I can understand why your employer has no money.

baker7

Thanks for even more detailed response. Yeah I need to try that eventually. Problem is.. there are a lot of problems.. too tired to get into the details. This sublimation business in LA downtown is tanking and these customers are doing whatever they feel like. Apparently they act like they're my boss's owner because they know without them he would go under.

It's a sad situation and sad to see him scrounging to just keep his business afloat. His wife started working 2 jobs cuz he hasn't brought money home in over a year. Guess we gotta just manage and try and tweek the plate curve as we go.

Thanks again for the explanation