No matter what your workflow, you still need the right conversion and file setup. Turning off CM, as your printer suggests, won’t affect your end at all, only make it more difficult for you to get your work done. Printers are afraid of CM because it threatens their proofing cashflow.
If you do turn CM off, PS will still use whatever you have loaded for your RGB and CMYK working spaces and use those as the basis for any color conversion. You can still use Convert to Profile to select an alternative destination profile. So turning CM off doesn’t really turn it off, it only makes your job more difficult.Haveing CM on or off has nothing to do with setting highlight and shadow dot, which have to be measured with the Info Palette no matter what your overall workflow.
turn off colour management in the PS colour settings
Unless your printer confirmed your working spaces match his working spaces, he is "very confused" (to put it kindly) about HOW Adobe color management works…
Thanks Peter,
how do I know what my working space settings should be, do I need to obtain information for the cmyk settings from my printer, like paper stock, ink limit, positive or negative plates,separation methods- ucr or gcr, black generation etc.as he is asking for composite pdf files and has provided pdf settings which don’t mention anything about the separation settings needed
I’m at the stage where I’m optimising my images for press. I’m colour correcting rgb tiffs, restting the end points etc (do I need to know the specific highlight and shadow dot reproduceable by his press) or can I assume a 90 percent shadow and an 8 percent highlight for example
It’s confusing, I just want to make sure I prepare the files correcly in PS before I bring them into Quark
Any advise from a more experience head would be most welcome as this is my first attempt at anything this big.
Sorry it’s so long Rob
Your printer should supply you with a custom profile that you can Image> Mode> Convert to Profile, but if he is recommending CM Off he likely won’t understand what you’re asking for.
The best I know how to do in such cases, is find out if the press is sheetfed or web, if the paper is coated or uncoated, then Image> Mode> Convert to Profile (US Sheetfed Coated if the job is going to sheetfed press on coated stock, for example).
The general theory is work/archive in a higher-gamut RGB colorspace (like Adobe RGB) — BEFORE ever going to CMYK — then as the last possible step: Image> Mode> Convert to Profile the Target CMYK…
If you are not sure about PS and color-managed workflow, I would start you here (with steps 1-4): <
http://www.gballard.net/nca.html#getagoodscreen>
You can be pretty sure that this printer is not going to have a profile. You can be pretty sure that the majority of printers don’t. You can, however, usually get pretty good information regarding ink limits and sometimes even how small a dot they can reliably hold. If it’s DTP, the your 8 percent highlight is probably too high, but it depends on the paper.
It also appears from your spelling of the word color, that you might be in the UK, which also changes the equation. The chances that your printer might possibly conform to any of the profile standards included in Photoshop are even less than they are in the US, where those odds are mighty slim.
I think that in the absence of have a good profile, or having one made for you, your best bet is going to be to output a couple of tests with the printer to see where his proofing system is at. If it were me, I would just send him a profile target and make a custom profile and be done with it, but you may not have that luxury.
Printers are afraid of CM because it threatens their proofing cashflow.
Not to mention that it "reduces" colour matching from an Art to a mere science…
One thing I often do with peopl;e who want me to turn CM off is to ask them to tell me what their colour settings are and then convert to the CMYK space they use.
I recently peeked into the Color Settings of a printer that a client was sending my RGB scans to. The printer’s conversions yielded proofs that were so dark and yellow-green that they were for the most part unasable and unfixable. I made a custom profile of their proofing system. Their Color Settings were of course set to Off, with AppleRGB for RGB and PS4CMYK Defaults for CMYK. This meant that every RGB was assumed to be AppleRGB no matter what it really was, and that their conversion, which was starting with a false assumption and then going to a CMYK profile that was totally inappropriate for the proofing system. No wonder they couldn’t get anything close to decent color.
The custom profile I made also revealed that their new proofing standard, and everyone has a different standard when it comes to DTP, was nowhere in the vicinity of US Sheetfed v2, or anyone elses profile that I have in my ColorSync folder. A simple Assign Profile told me how far off they were.
So, at least in this case, getting the Color Settings was a help, but not the whole story. They were instrumental in troubleshooting the situation.After the book is printed I’ll probably have a meeting with the printer to shed some light on why they are having such a problem, but I really don’t know if anyone there understands enough to matter.
and the printers name in LA?
this is the big problem with adobe’s approach to CM, in that by forcing it on the creation side, it puts the onus on the creation side to ensure that CM is obeyed all the way through…but thats not always possible, and when thats the case there is no feedback to that effect except bad colour accuracy on the final job or proof. The other areas it fails to address is that all presses have working tolerances which vary which can be addressed by the printer to maintain consistancy, but at what point do the profiles take their stance, especially where stock etc is concerned..? Thats where the science breaks down and leaves the art in the printers hands to make the press behave well.
In the abscence of profiles, proof first, ensuring that their proofs match their press output. Where you go from there is really guesswork…but really all you want to change is your preview rather than any ink settings etc (as you have already run a proof)…so really that takes us back to bending the monitors look to suit…which takes us back to the old knoll gamma cp which allowed this to be done. Once the job is dead you can switch back to how it was, but this is really the only route i have found that will allow this to happen.
The other option is to run a proof through a profiled output bureau and tell the printer to hit that, so he has a visual target to aim for. ie you get it how you want, supply the profiled files that produced that and hand that over to the printer. That way you have proved that the file can do what you want it to do, now all the printer has to do is match it on their system.
If you are in the UK/europe forget the US prepress defaults etc, the inks are way off over here…odd hues and uglyness abound.
Heres my uk settings, that i find are broadly consistant in terms of no ugly shift in colour hue and good representation of ink density/gamma.
RGB:Apple RGB (sometimes sRGB depending on what im doing) CYMK:Photoshop 4 Default
engine:Adobe ACE using relative colourmetric
Profiles always set to save with the file.
My monitor is hardware calibrated and then "tweaked" a little in the brightness/contrast to match repro output. We’re quite happy with our output now, having been in the dark for a while finding a profile that suits our client’s broad range of outputs whilst not actually knowing if they obey CM or what they may even be outputting to.
Mike,
It was a specific printer that doesn’t need to be named, but it could easily be ANY of dozens or even hundreds out there. You already know that.
Progress,
I don’t agree with you about Adobe’s CM implementation. Would you rather not have the tools that we now have and use? It’s not my fault (or yours) that the printing industry in general acts the way they do. I’d much rather have the tools to circumvent the process and get the results my clients demand.
As far as your recommendations regarding UK printing. If you are using the Color Settings you refer to and still are resorting to tweaking the monitor, it only means that your settings are off and your are invoking a huge kludge to compensate. The RGB space doesn’t matter nearly as much as the correct CMYK setup. It’s the latter that really makes or breaks the job. If, after you’ve run the proof, your colors or tonality is off, and you want to tweak your screen, it should be done by either choosing a different profile that provides a closer visual match or by tweaking the Printing Inks setup, which generally requires a spectrophotometer for best accuracy but can be done visually in a pinch. And if you are dealing with a vendor that you repeatedly use, it’s certainly worth the hundred buck or so that it will cost to get a good custom profile that will address with much more finesse, the peculiarities of that proofing system.
Progress,
What preview do you get using View Proof with PS 4 default settings?
There must’ve been some big changes between PS 4’s inkset and ink limit definitions and the new US Prepress defaults. I wish there was a visual example of prints showing the difference between these two on the web. The previews seem more accurate to press using relative colormetric, paper white, ink black and US Prepress defaults over PS 4’s.
Your method of bending the monitor to match the print is surprisingly included instruction in the PS 5 manual but I couldn’t find it in PS 7’s. I believe it might be in online Help, though.
Pete,
If it’s my company, I want to know.
Tim,
All you have to do to see the difference is convert a file (the test file I sent you earlier is good) with one profile and then Assign the other profile. Toggle them preview on and off. If you convert to U.S Sheetfedv2 and assign PS4Default, the new preview is very blown out. Try it, you’ll see instantly. As far as your previews seeming more accurate with the later settings, all that means is that your printer’s proofing system is more closely aligned to these settings than the earlier ones. There is so much variation these days that without a custom profile, it is a complete crapshoot.
Mike,
The people that run prepress at almost all printing companies have no idea what I’m doing. They humor me and think it’s quaint when they print out the targets. The people that own the companies are further out of the loop. I WILL go to this vendor, but only after the book is printed. Their egos are just too sensitive to risk offending someone and affecting the job. I think one of the reasons the egos are so sensitive is that, after asking a few innocent questions, it becomes readily apparent that they don’t quite get the concepts.
Pete,
you ain’t telling me anything new here dude.
but you have to understand that printers, at least the bigger shops ARE starting to get into this color stuff and ARE becoming pro-active.
and yes, big babies work for these companies with fat wallets.
so if you want some of that, you have to deal with the crap.
Mike,
I know all of that. I know that some shops are progressing. I also know that the vast majority are not. You would think that that might be the case with all the DTP, but it seems like an outside (RIP vendor?) is hired to come in and set everything up without educating the people who are stuck working with the new proofing systems.They are given a turnkey solution without understanding what’s going on under the hood.
Pete,
If you’re speaking of the PDI file, I do get a very accurate preview on my display compared to your Epson 2200 print you sent when viewed in standard RGB US Prepress Default view without View Proof.
However, the preview is too pretty in it’s current RGB view and I have to resort to View Proof, using the settings mentioned earlier, to get the narrowed gamut and slightly yellowed paper print look of PS’s SWOP v2 profiles. The VP preview now shows the predominant yellows (8C,30M,35Y) in the fleshtones that are in the CMYK data. If I use the sheetfed version, the PW shows a slightly bluer (more neutral?) paper white.
I’ve seen this PDI file printed in the CM section of a digital imaging book in a bookstore I frequent and it looks off (to much magenta) from your Epson print and my RGB preview. The CMYK View Proof with PW and Ink Black simulations on is closer but doesn’t show the magenta. Of course the file has been shrunk down to about 2 to 3 inches tall in the book and that might have something to do with the hue shift.
The PS 4 defaults choice in the VP won’t let me use the simulations check boxes. The settings are grayed out. This was my point to Progress on the major differences between PS 4 and the new PS US Prepress settings. I was wondering if he was aware of the preview limitations between the two. I was just curious what made them different.
And in addition, I do get the blowout affect mentioned by Peter with PS 4 according to the convert and then assign instructions.
Mighty clammy looking previews by comparison to PS US prepress defaults. It’s almost as if PS 4 injects a 1.8 gamma curve as apposed to the seemingly perceptual curve of PS’s US Prepress default.
I remember this look when I restored photos in PS 4 space before CM back in the day and when I output to a nonCM service bureau, I got that flattened, washout clammy look in my portraits. Talk about a waist of time and energy.
Now I see the difference. Boy! Adobe must of worked really hard to make such improvements in a profile and conversion engine. And thank you SuperCal for the perceptual curve for my monitor profile.
Tim,
I dont know, I think for some reason im not on PS 4 defaults because the setup is now labelled custom. PS 4 CYMK as a profile though seems to work for me as the best out of all of them. I work in RGB/CYMK preview with PS 4 CYMK as the target. Works for me 🙂 I dont have any shift with convert/assign, and im not sure you should see any. US prepress just looks guady, hue bent and unnatural. If i dump CYMK red into the document it looks nothing like repro output. PS 4 does as it does with a lot of other colours that need to be right. As far as im concerned, it previews naturally, my files output naturally, so it works for me.
Surely though one should convert to a profile and work/preview using that same profile? Im not sure how seeing the differences in profiles helps, surely the most important thing is how closely it matches your output if your looking for a suitable profile?
If your getting clammy previews and clammy outputs then perhaps your profiles were ok but your monitor was out so you were working skewiff to start with. Dont judge a profile by its on screen look if you werent working calibrated or didnt have an output to compare.
Of course if your seeing how a file may behave under your profile enviroment compared to its intended then you should see a difference surely and adjust the file to compensate. This will could obviously degrade the image but then i think this is the argument behind supplying RGB files and letting the printer do the conversion, but I think thats a real leap for some, who may not understand the basics anyway.
Peter,
I am aware of what your saying. I realise the caveats but sometimes the kludge was necessary, but I dont do it now because adobe gamma cant do what knoll gamma can and im on osx now, so i float with the profile boat and not hope for a storm.
If i was dealing with a few print houses then I would go to the lengths of profiling if i knew they were in the profile party as well. However im not, i work blind, blinded by several layers of fat between me and the clients output, who may print the same image in different parts of the world depending on where its going to be used. Basically i dont know if CM gets broken down the line, but i stick my little tag on it and hope.
I think CM breaks down at the moment because it only implies intent to those blessed to understand CM, and there is no stopping the chain being broken, or even guranteeing that the originator knew what they were doing in the first place. It works within the finite world of the blessed, but isnt solid or reliable in the world of the misinformed, the uniformed or the wont be informed. Thats the problem, but its magnified by the amount of people who fall into those catagories. I think an Adobe CM survey would provide a shocking result. I’d say it was probably 1/10 houses that know about it and implement it.
What I meant about the kludge was that if you have generated the file and proofed to it, keeping as many things constant, the easiest thing to change is your monitor representation so you can see what that route gives and compensate the file if necessary using the temporary screen recalibration. However that only applies to that particular output and that should be remembered, otherwise you will screw up the rest of your outputs. This was the necessary kludge before CM, and is still argueable necessary in its abscense.
The hardware calibfration i did still needed a tweak, because i could see that it was off, blacks and gamma werent quite right, convergance was a little shot away as well. This was in RGB. Dont ask me why it wasnt right even after several stabs, but black is black and the lacie wouldnt give me it. Probably should have gone for the sony, but the 2:1 lacie price vs the sony swayed me…should have listened to mike, the v4 lacies arent as good as they v3’s.
We checked the monitors in rgb against other cailbrated monitors, so we knew we were driving down the right road. After that we searched for a profile whose reds werent wacky, whose saturation wasnt blown out of possible repro cymk etc etc. PS 4 fitted, and it works, our colours dont shift, the gamma looks right for the inks, and it gives us a good representation. The problem comes when an unprofiled house proofs our image and it doesnt match our screen because the output device is a different beast to the norm. Profiling for that one job isnt a reality because of time constraints, i cant tell the client their workhouse is shite either, so its out with the kludge (if possible), or we have to supply them a proof that matches our system and knock it out with them or adjust it on their system…all a nause.
Perhaps adobe needs an adobe CM certification program? That could go along with mikes profile layer idea, and some sort of info on what discarding, converting or assigning to an imported file will do before the damage is done and the originators intent is lost.
lacie price vs the sony swayed me…should have listened to mike, the v4 lacies aren’t as good as they v3’s. <
no kidding.
Just got one, a month ago and the damn thing is blue on the 1/3 left side.
Eventually, the Artisan will be the only choice because the CRT manufactures are being squeezed due to the LCD’s price and it’s affecting the quality of the CRT’s.
progress,
I don’t know the origination of the files you’re trying to make look correct by your methods described editing in an unknown custom PS RGB space with PS 4 CMYK preview. To many unknowns for me to offer advice or alternate solutions.
What do you mean by quady, hue bent and unnatural? And you are suppose to get a preview shift when assigning a different CMYK profile after converting to CMYK from RGB with PS set to US Prepress defaults. If you have a custom RGB space other than US Prepress defaults, there’s no telling what PS is doing to your previews-RGB or CMYK.
Going by memory color, flipping through my MacWorld mags and other publications featuring the Mac interface, the images look dead on to my display. At least I know my display is optimized to match these guys. Your client’s capture and output space must truly be a beast if you have to resort to the methods you describe.
Mike,
Are they that bad? My 1998 CRT is still holding strong at least according to images matching to Peter Figen’s Epson print off his Artisan.
I bet the Lacie III is going to be hard to find when I need to replace this dinosaur.
Tim,
We originate the files in rgb (in an unassigned and unmapped rgb space because they come from 3d renderers)…we assign or convert to apple rgb or srgb so we know they’re we’re on a level playing ground as far as PS is concerned and we can still retain the colour intent of the original render. That way we dont edit an unknown space, because we have assigned or converted to an rgb space.
Yes, you will see a colour shift, but you wont necessarily see one when assigning some profiles. Remember theres a big difference between assigning and converting profiles. But the main important thing is that your profiled on screen image matches your output…whether you like the image or not, thats the most important thing. You can only compare files that you have output, you cant just say theres a mac screen shot in a mag and thats the same as what i see on screen because you dont know how that mag sceen shot was handled. You certainly cant base a profile choice around it, theres a whole world of fettling that could have happened inbetween. You can only base a profile choice around having a proof of a profiled file in your hands with a cailibrated screen. Having a calibrated screen is your first step, your blind without it. If you have a calibrated screen you can see by working within a profile if you will at least be getting natural results with certain colours, when you are determining profiles that may somewhat match your general output. US prepress defaults dont for me. Dont confuse PS4 defaults with PS 4 cymk, one is a setup the other a profile.
If you have the time and a constant output source, then perhaps a custom profile is worth the effort, unfortunatley with us, thats a practical impossibility, so we’re left in the dark, almost ignored by adobe’s cm and by those that ignore it.
Unnatural colours mean blues going cyan/greenish hue, reds looking too saturated like pure RGB red or going magenta, greens going blueish and heavy and overall heavier. US print always looks gaudy to us this side of the pond anyway, but US prepress aint for us. If however these gaudy colours matched my output then i would go with them, but they dont, and im not sure how many people over here can get a match with US settings etc.
You’ll be looking at an artisan for replacement of the lacie v3 unless you can really test a v4 out beforehand, if you want to be certain of a good high quality monitor. The price is the problem…2 for 1 artisan.
