Monitor calibration-->CMYK printing

[monokakata]

I create book covers for a very small press. I could link to the press website but it might be considered self-promotion, so I won't.

I create them in Photoshop, and drop the jpg outputs into an InDesign template that comes from the on-demand printer (Lightning Source). I always make the jpgs CMYK (in Photoshop) before I start working on them. InDesign then puts out a PDF for the printer.

So far, no one has cared how closely the printed covers match the originals (almost always photographs). All the authors have loved their covers. There's never been and won't be a client who requires matching a logo or corporate color or anything like that.

Right now I'm working on a cover and I have gotten to a shade of red that works well -- by trial and error. I'm worried because if the printed version drifts very far from what I see on my screen I don't think it's going to work. It could tip into something muddy.

I have no control over the printing process. All I can do is wait for a physical proof of the book (and an e-proof (PDF)) and if it looks OK, the publisher approves it and that's that.

Resubmissions are expensive for a small-volume press.

I'm sure that one of these days (we do 3-4 books a year) I'm going to create something I like, the author likes, the publisher likes, but that we're not going to like what we get back from Lightning Source (and it won't be LS's fault -- I'm not saying that at all).

My question is: is it worth getting a calibration setup for my Dell U2711 (and/ or rMBP) to calibrate for RGB even though I'm putting out CMYK?

The publisher is completely willing to spend the money, and I like the idea, but I'm not sure how (or even if) I can calibrate my monitor to show me what I'm going to see after Lightning Source takes my CMYK PDF and prints it on whatever commercial printers they use.

As I said, everybody's been happy so far. But the world being what it is, sooner or later I'm going to have a color fail. And I'd like to do whatever I can to make that unlikely.

I was thinking of the Datacolor Spyder4Elite S4EL100, but I'd be grateful for suggestions.

 

[Apple fanboy]

I create book covers for a very small press. I could link to the press website but it might be considered self-promotion, so I won't.

I create them in Photoshop, and drop the jpg outputs into an InDesign template that comes from the on-demand printer (Lightning Source). I always make the jpgs CMYK (in Photoshop) before I start working on them. InDesign then puts out a PDF for the printer.

So far, no one has cared how closely the printed covers match the originals (almost always photographs). All the authors have loved their covers. There's never been and won't be a client who requires matching a logo or corporate color or anything like that.

Right now I'm working on a cover and I have gotten to a shade of red that works well -- by trial and error. I'm worried because if the printed version drifts very far from what I see on my screen I don't think it's going to work. It could tip into something muddy.

I have no control over the printing process. All I can do is wait for a physical proof of the book (and an e-proof (PDF)) and if it looks OK, the publisher approves it and that's that.

Resubmissions are expensive for a small-volume press.

I'm sure that one of these days (we do 3-4 books a year) I'm going to create something I like, the author likes, the publisher likes, but that we're not going to like what we get back from Lightning Source (and it won't be LS's fault -- I'm not saying that at all).

My question is: is it worth getting a calibration setup for my Dell U2711 (and/ or rMBP) to calibrate for RGB even though I'm putting out CMYK?

The publisher is completely willing to spend the money, and I like the idea, but I'm not sure how (or even if) I can calibrate my monitor to show me what I'm going to see after Lightning Source takes my CMYK PDF and prints it on whatever commercial printers they use.

As I said, everybody's been happy so far. But the world being what it is, sooner or later I'm going to have a color fail. And I'd like to do whatever I can to make that unlikely.

I was thinking of the Datacolor Spyder4Elite S4EL100, but I'd be grateful for suggestions.

The Spyder4Elite is a capable monitor calibrator. Another one to look at is the Xrite i1 Display Pro. It measures more colour patches than the Spyder so is slightly more accurate. But either of these would give you a good result if used properly.

 

[citizenzen]

My question is: is it worth getting a calibration setup for my Dell U2711 (and/ or rMBP) to calibrate for RGB even though I'm putting out CMYK?

As a print designer for over 20 years I highly recommend you calibrate your monitor. I could not imagine trying to color correct without it. Even though your monitor displays in RGB, it's vital that it be as accurate as possible.

One note about the red that you mentioned ...

An offset printer once passed on some wisdom to me. He showed me a swatch of Magenta and told me that no matter what, this was the most "red" he could put on press. He could add yellow to it ... but that wouldn't make it more "red". He could add black to it, which would make it darker, but it wouldn't make it more "red".

Once you realize the limitations of the inks involved you become a little more forgiving about what they can't do, and appreciative of the amazing things they can do despite those limitations.

And a final note about workflow ...

While I always check with my printer to see what they require, I've had great success at not converting my photos to CMYK. I leave them 16-bit, RGB and let the CMYK conversion happen during the color separation for making the plates. Occasionally you'll run across a photo that you want to see in CMYK, and for that I use Edit > Convert to Profile to test it for color shift. I adjust it in RGB until there's virtually no change when converted to CMYK.

----------

Another one to look at is the Xrite i1 Display Pro.

This is the one I use.

 

[Apple fanboy]

----------

[/COLOR]

This is the one I use.

Me too mostly. Sometimes I borrow the ColorMunki Photo instead. (I borrow both from work)

 

[thekev]

While I always check with my printer to see what they require, I've had great success at not converting my photos to CMYK. I leave them 16-bit, RGB and let the CMYK conversion happen during the color separation for making the plates. Occasionally you'll run across a photo that you want to see before the color shifts that will occur, and for that I'll use Edit > Convert to Profile to test it and adjust it in RGB until there's virtually no change when converted to CMYK.

You have to be careful about things that might cause posterization or fall out of gamut on conversion. Gamut warning can be helpful for that. If they're digital raws, you have to ensure that nothing being processed in a raw processor is clipping on the initial RGB as well. Otherwise by the time it gets to CMYK it's a complete mess. It's also true that it's a good idea to calibrate the display, because it gives the OS some idea of how the monitor actually responds and the limits of its primary colors. It's still not a perfect match to print though. At the very least an inkjet from a decent RIP (GMG, EFI, etc.) viewed under D65 lighting is a better approximation. If the final image will have a UV coating applied, figure that is typically a 2% offset or so in terms of yellow. It's approximate but it usually works. I think a good display is really essential, but I wouldn't completely disregard proofs and reference prints in favor of making judgements based on the display.

Edit: That's not really an argument from me. I just wanted to add more detail.

 

[monokakata]

And a final note about workflow ...

While I always check with my printer to see what they require, I've had great success at not converting my photos to CMYK. I leave them 16-bit, RGB and let the CMYK conversion happen during the color separation for making the plates. Occasionally you'll run across a photo that you want to see in CMYK, and for that I use Edit > Convert to Profile to test it for color shift. I adjust it in RGB until there's virtually no change when converted to CMYK.

Thanks very much for the workflow advice. I thought that by immediately working in CMYK I was doing the right thing. I totally get it (now) about going back and forth until there's no change. Makes perfect sense as a technique. Thank you.

I hope I didn't seem like an idiot. I fully understood that when headed towards magazines, brochures, color plates etc., that calibration and standards all along the line were essential. But I figured that in the work I was doing there was no standard of comparison -- that no printer or the publisher or the author would be saying, "this color isn't what it's supposed to be," because the book cover effectively defined itself.

 

[monokakata]

You have to be careful about things that might cause posterization or fall out of gamut on conversion. Gamut warning can be helpful for that. If they're digital raws, you have to ensure that nothing being processed in a raw processor is clipping on the initial RGB as well. Otherwise by the time it gets to CMYK it's a complete mess.

Thanks. If it matters, the covers are 6x9 or 5.5x8.5, which is to say pretty small.

Usually we use a raw image I've made myself. I'm more comfortable in Lightroom than in Photoshop, so I manipulate the image in LR to get it looking the way I want it to.

Then I export it and bring the JPG (full strength) into PS for cropping and actual cover creation. Finally I export a JPG and that goes via InDesign into a PDF.

Then either InDesign or Adobe Acrobat X issues warnings if there are any. The first time I saw "more than 100% black" I was mystified.

Do you suggest that when I'm finished in LR I give the raw or a DNG over to PS, rather than a JPG and not make a JPG until the last moment (apart from tests to show people)? I never re-edit a JPG; I only save as PSD.

 

[thekev]

Do you suggest that when I'm finished in LR I give the raw or a DNG over to PS, rather than a JPG and not make a JPG until the last moment (apart from tests to show people)? I never re-edit a JPG; I only save as PSD.

I would just export from LR as a tiff or psd rather than a jpeg, even if the difference is minor at lower levels of compression. By more than 100% black do you mean your K channel was clipped or you were beyond the total ink limit of the provided specification? Technically a CMYK profile can go to 400%, but you can't print that. Depending on the type of printing, 300-350% is more typical.

The things I was trying to point out were a bit different. One was that the monitor is never a complete reference. A proof is a better reference, but for an exact match it should ideally be viewed under a color corrected viewing lamp. You can't merely assume that what is on your monitor will show up. Also the Dells are okay. I like Eizo, but I've had to do a lot of heavy composite work where any minor issue in screen uniformity is terribly annoying. A contract proof or inkjet proof from a decent RIP is much more valuable in the end. The other point I was trying to make was to check your clipped values at various points. Many people make the mistake of processing things from very large gamuts into much smaller ones without appropriate compensation. This is especially prevalent with digital medium format brands. The gamuts of some of these phase one backs are pretty wide. Someone then processes a photo with really red lipstick or a red sweater or a saturated blue sky into Adobe RGB with relative colorimetric conversion without checking on what actually fits, and they end up with a mess. Converting to cmyk later makes it even worse, where good raw processing allows you to output something that will actually fit in that smaller gamut. There are other quirks to RGB and CMYK conversion, but the biggest ones really are saturated blues and reds followed by greens. Some people have difficulty with metallics as well. I'm puzzled by that one.

 

[torana355]

While I always check with my printer to see what they require, I've had great success at not converting my photos to CMYK. I leave them 16-bit, RGB and let the CMYK conversion happen during the color separation for making the plates. Occasionally you'll run across a photo that you want to see in CMYK, and for that I use Edit > Convert to Profile to test it for color shift. I adjust it in RGB until there's virtually no change when converted to CMYK.

----------



This is the one I use.

Supplying files to a printer in RGB is terrible advice. Ive worked in prepress for over 15 years and can speak from experience that Converting to RGB at the prepress stage can lead to inconsistent results. If you have a job that is Colour critical you need to get either a machine proof or an Epson/HP proof from a proofer that is calibrated to the press using an ICC profile.

 

[monokakata]

I would just export from LR as a tiff or psd rather than a jpeg, even if the difference is minor at lower levels of compression. By more than 100% black do you mean your K channel was clipped or you were beyond the total ink limit of the provided specification? Technically a CMYK profile can go to 400%, but you can't print that. Depending on the type of printing, 300-350% is more typical.

The things I was trying to point out were a bit different. One was that the monitor is never a complete reference. A proof is a better reference, but for an exact match it should ideally be viewed under a color corrected viewing lamp. You can't merely assume that what is on your monitor will show up. Also the Dells are okay. I like Eizo, but I've had to do a lot of heavy composite work where any minor issue in screen uniformity is terribly annoying. A contract proof or inkjet proof from a decent RIP is much more valuable in the end. The other point I was trying to make was to check your clipped values at various points. Many people make the mistake of processing things from very large gamuts into much smaller ones without appropriate compensation. This is especially prevalent with digital medium format brands. The gamuts of some of these phase one backs are pretty wide. Someone then processes a photo with really red lipstick or a red sweater or a saturated blue sky into Adobe RGB with relative colorimetric conversion without checking on what actually fits, and they end up with a mess. Converting to cmyk later makes it even worse, where good raw processing allows you to output something that will actually fit in that smaller gamut. There are other quirks to RGB and CMYK conversion, but the biggest ones really are saturated blues and reds followed by greens. Some people have difficulty with metallics as well. I'm puzzled by that one.

This is all so helpful. The % black thing was, I'm pretty sure, that I was beyond the ink limit set by Lightning Source. When the cover came back, the black wasn't nearly as black as I'd thought it would be. It was towards the brown a bit.

I'll be careful about gamut.

You're in a different world than I am, I think. We're talking about books that would do well to sell 200-300 copies. Each revised PDF that goes to the printer costs us $40 and now that we're operating in Hawai'i, the delivery of the proof book has gone from $30 overnight to something awful like $100. So you can see that one revision and one new proof copy would mean about $1 more cost per $15 or $17 book that wholesales at $5-$7.

A micro-press can't afford that kind of back-and-forth, so it's basically a one-shot. I do the best I can and then LS does the best it can. It's been all good so far, as I said, but I'm a cautious guy and thus my interest in calibration.

One more question (for this go-around). Do you think I'd do better looking at my colors on the rMBP or the Dell, assuming calibration?

I'm still reluctant to give a link to the press website. I don't have a financial interest in it, and I don't get paid for my work, but the publisher is my life partner and so obviously I'm all tangled up in the affairs of the press. But I don't get any money from it (at least partly because it's lost money for the first 3 years it's been in existence).

If any of you professionals are curious, I have the covers in a Dropbox and if you send me a PM I'll give you the link. Most of them are in print and won't be modified. But three are in progress. I'd be happy to know what anybody thinks about them.

 

[thekev]

Supplying files to a printer in RGB is terrible advice. Ive worked in prepress for over 15 years and can speak from experience that Converting to RGB at the prepress stage can lead to inconsistent results. If you have a job that is Colour critical you need to get either a machine proof or an Epson/HP proof from a proofer that is calibrated to the press using an ICC profile.

I know what you're saying here, but consider the audience. Presses do shift. They aren't entirely consistent, and it's not like you can precisely calibrate to that. With a decent proofing system consisting of a printer, printing RIP, and preferably a spectrophotometer to measure output values, you can get a lot closer. I don't think it's like the OP could budget to set up such a thing and learn it overnight. I completely agree about not just sending out whatever in RGB, as they are not going to handle out of gamut problems well, which is one of the most common reasons for color shifts and loss of detail.

I should probably tack on that it's not like a shift waiting to happen is always apparent on screen. I know you implied that, but I wanted to mention it explicitly.

This is all so helpful. The % black thing was, I'm pretty sure, that I was beyond the ink limit set by Lightning Source. When the cover came back, the black wasn't nearly as black as I'd thought it would be. It was towards the brown a bit.

I'll be careful about gamut.

You're in a different world than I am, I think. We're talking about books that would do well to sell 200-300 copies. Each revised PDF that goes to the printer costs us $40 and now that we're operating in Hawai'i, the delivery of the proof book has gone from $30 overnight to something awful like $100. So you can see that one revision and one new proof copy would mean about $1 more cost per $15 or $17 book that wholesales at $5-$7.

I understand your frustration. There really is more to this than I can explain in a post or two. I'm not sure how it would go beyond a 100% K value under any kind of normal circumstances. I just wanted to point out that there are hard ink limits and that gamut must be watched at each step in the process. Regarding brownish blacks, that isn't so much an issue of density. It's an issue of too little black ink and too much cyan, magenta and yellow in the deeper tones. I'm not sure if you were given any kind of spec information from the printer. It would have notes regarding separation type, total ink limit, etc.

A micro-press can't afford that kind of back-and-forth, so it's basically a one-shot. I do the best I can and then LS does the best it can. It's been all good so far, as I said, but I'm a cautious guy and thus my interest in calibration.

One more question (for this go-around). Do you think I'd do better looking at my colors on the rMBP or the Dell, assuming calibration?

If it was me, I would go with the Dell. It's one of their better displays. Apple is still working within the constraints of notebook screens in the rMBP. For calibration, just stick to something somewhat standard like D65 gamma 2.2. Give it 30-45 minutes to warm up prior to calibration (can't stress that enough). They all require some kind of warmup time. I disable graphics switching just to be safe. Most calibration software requires that you disable any kind of mirroring and set it as the dominant display just in case you're attaching this to your rmbp. Note that profiles don't transfer between machines. You'll want to do this under whatever setup you will be using.

You'll also notice that typical soft proofing setups often recommend really high brightness. It's because in a pre-press setting the viewing booths used are very bright, and its an attempt to match the amount of light reflected back from the print in the direction of the viewer with the amount emitted from the screen within a certain tolerance. For your purposes it would probably result in bad judgements, as it will tend to lie to your eyes when it comes to contrast. I don't have that much experience with Dells (mostly NEC and Eizo, Sony Artisan years ago).

If you really must go without any kind of printed reference, even an inkjet proof, set it around 80-100 cd/m^2 via whatever method of backlight adjustment. Some of them look terrible when dialed down that far. At least get it below 120, as like I mentioned, you're not comparing against a print under a big normlicht or something like that. I would also suggest you make your corrections in a very dark room, so your eyes are not influenced by other things and to lessen the effect of reflections on your display. This stuff causes confusion for a lot of people. Just be aware that calibrating your monitor is not a catch all solution. Reference prints or actual proofs are much more reliable, which is why I suggest that. It can still be kind of a shot in the dark without appropriate proofs, and I am concerned about the brownish blacks issue which I explained above.

 

[monokakata]

I went back to the original PDF I sent to the printer, and ran it through Acrobat X's preflight (which is where I remember seeing the warning). But it wouldn't do it again today.

Lightning Source (who are really great to work with) don't give exact specs. This is all I ever found:
---

ICC Profiles are color profiles that can be applied to files (and images & objects in them) intended to assist with the printing process. We do not recommend using them.

Color Book Interior/Cover:
Because color book files are required to be PDF/x-1a:2001 compliant, ICC profiles cannot be used for any color book files.

------

LS's business is print on demand -- fast, relatively cheap, and so on. A person would never go to them for an art book or a book with photos. The printing in the interior does OK with small grayscale images (such as an author photo). I generally give them interior images (when required) that are meant to print at 20% - 30% transparency (InDesign). Those print well, and because the images are always clearly manipulated, the reader never complains about them -- in fact, they think they're cool. It's clear that they're meant to be design elements rather than photographs. The author I'm working with now wanted to have "straight" grayscale photos heading each chapter, but I refused. It wouldn't have looked good at all. (I do the interiors also.)

I'm driving the Dell from my Mac Pro 5,1 with a 5770, Mavericks.

I do think I must have made some mistake somewhere along the way with the brownish black cover, probably in Photoshop. I did another mostly-black cover and it was just fine.

I ordered the Spider Elite and it should be here next week. Thanks for the tips on monitor settings.

I do have an Epson Stylus R2400 that I don't use much, but I could put it back into service. Do you think I'd get useful information printing to it?

 

[spacedcadet]

CMYK Preview

Enjoyed reading the detailed information from previous posters.

A quick shortcut, Cmd-Y in Photoshop will show you the CMYK preview when working in RGB, easy to toggle back and forth and see which colours change.

For what it's worth, I work on book covers and our workflow goes pretty much as follows.

We have Eizo monitors calibrated with an X-rite hardware calibrator.

Images are CMYK imported as TIFF or PSD into InDesign. I work mainly directly in CMYK as that's how my brain does colour for print. RGB if I need an effect or adjustment layer in PhotoShop that doesn't work in CMYK.

Knowing how to read CMYK percentages and how they print is really useful.

Pure CMYK red is 100M 100Y. Take some Magenta out and you go toward orange etc.

Blacks can change or get muddy if they have too much of the other colours in. A nice cold dark black I use a lot is 60C 40M 100K.

The dense neutral black that comes out of our colour profile conversion is 78C 68M 58Y 94K.

Internal test prints are output to a calibrated colour laser copier. Not ideal as it's not good at pale gradients, pastel shades and tends to be a little murky in dark image areas. However it's the best combination of speed, cost and colour matching for our studio.

Before final print we get an epson digital inkjet proof from our repro house.

One of the most important things for us is using the correct colour profiles in the CS Suite. Ours is based on ISOcoated_v2_eci.icc for European offset litho. When sending stuff to print internally, Photoshop or InDesign handle colours and use that profile. CS Suite is set to Synchronize across the apps via Bridge.

Creating PDFs is done via printing a postscript .ps file and running through Distiller/Enfocus Pitstop with our settings to PDF 1.3 compliance.

We find the colour copier output a little darker than on screen, but a pretty good match and the digital proofs are closer to what we see on screen. All in all the process works pretty well. It can never be exact, but we aim to be close and consistent, so even if it's wrong it's predictable.

Appreciate you are doing low volume work, but I think if you callibrate monitor, finalise work as CMYK and know your final output devices capabilities, then you should get pretty close to where you need to be.

 

[torana355]

I know what you're saying here, but consider the audience. Presses do shift. They aren't entirely consistent, and it's not like you can precisely calibrate to that. With a decent proofing system consisting of a printer, printing RIP, and preferably a spectrophotometer to measure output values, you can get a lot closer. I don't think it's like the OP could budget to set up such a thing and learn it overnight. I completely agree about not just sending out whatever in RGB, as they are not going to handle out of gamut problems well, which is one of the most common reasons for color shifts and loss of detail.

I should probably tack on that it's not like a shift waiting to happen is always apparent on screen. I know you implied that, but I wanted to mention it explicitly.

Fair enough i was just making the point that the conversion of RGB to CMYK is not standardized. One printer may convert an RGB image differently to the next. You are just adding another variable where things can go wrong. Just to let you know, Yes printing presses can and do shift but any decent printer keeps their proofing system and Press calibrated to a delta-e under 5.

If the printer the OP is using does not have the facilities to run a proof that is calibrated to the press standard or is not running to a common standard like ISO 12647-2 then playing around with calibrating a monitor for colour is actually pointless. If Colour is important i would advise doing actual machine proofs on the press.

 

[monokakata]

Thanks, spacedcadet and torana. More useful advice, especially having those CMYK values to start with.

I don't think I've made one important thing clear, and that's the nature of the printing company I use (and others like it). This is POD (Print On Demand), and it works differently from traditional printing in a way that probably explains why I don't know anything about their requirements except that the PDF must use CMYK and must meet certain Adobe standards.

Lightning Source (LS) gets the PDF and does whatever setup they do and then ships one physical proof of the entire book to the publisher, and makes an e-proof (PDF) available on line.

When the publisher signs off on the proof, the book is then live and -- here's the biggest difference -- LS then prints orders that come to it from Amazon or Barnes & Noble, one or two books at a time. There's nothing like a "print run" unless somebody orders a bunch of books. I think the most the publisher ever ordered at a time was 50-odd.

Thus whatever setup LS makes initially is the setup that's called on for every print copy, whether it's one or a dozen, tomorrow or next year. I'm sure they work hard to keep all their output devices synched but I don't think the operators or system choose which device to print on unless it's by size. Nobody's pulling a copy and looking at it, so far as I know.

In case you're wondering why POD -- it's perfect for very small publishers. LS is responsible for putting the book into the big wholesale book distribution systems, from their own shop in TN. When a wholesaler orders books, the publisher doesn't even necessarily know about it. No warehousing, no accounting department, no shipping department, and so on . . . perfect for a very small outfit and especially useful for one out in the middle of the Pacific, where shipping costs are astronomical.

A big outfit doesn't need that, and of course a big outfit has people to talk to the printer in order to get everything just right for a long run.

Thanks again to everybody. I feel as though I have better direction now than I did yesterday, and I appreciate it.

 

[citizenzen]

Supplying files to a printer in RGB is terrible advice. Ive worked in prepress for over 15 years and can speak from experience that Converting to RGB at the prepress stage can lead to inconsistent results. If you have a job that is Colour critical you need to get either a machine proof or an Epson/HP proof from a proofer that is calibrated to the press using an ICC profile.

I will respectfully disagree with you.

The promise of RGB (and PDF) for years has been the capability for one file to be correctly color separated using whatever profile it is processed with when rasterized. The workflow is sound, provided your vendor is likewise profiling their color separators, proofers and presses.

You should always check with your vendor before assuming this is the case.

 

[thekev]

Fair enough i was just making the point that the conversion of RGB to CMYK is not standardized. One printer may convert an RGB image differently to the next. You are just adding another variable where things can go wrong. Just to let you know, Yes printing presses can and do shift but any decent printer keeps their proofing system and Press calibrated to a delta-e under 5.

If the printer the OP is using does not have the facilities to run a proof that is calibrated to the press standard or is not running to a common standard like ISO 12647-2 then playing around with calibrating a monitor for colour is actually pointless. If Colour is important i would advise doing actual machine proofs on the press.

I'm not so much in disagreement with you here. There are only a couple things I would point out. A contract proof with the shipping costs mentioned and everything seems to be cost prohibitive to the OP. Even delta-e approaching 5 is still noticeable in side by side comparison anyway. I think he's sort of limited on options. I kind of agree with you about the monitor, yet I don't think it's completely unhelpful. It can clear up some of the most obvious color casts due to a poor description of the monitor's own gamut/behavior in whatever canned profile is loaded by default. I tried to ensure that I didn't over-promise anything with that claim. My evidence is purely anecdotal having dealt with several others in similar situations. I mentioned the lowered brightness because the most frequent mistakes by those without the full setup of necessary equipment is to adjust things to be a bit too dark.

As I mentioned brownish blacks also suggest that in the final values he has too much CMY relative to K. If it's way far off, it may have been converted incorrectly somewhere along the line.

 

[Consultant]

1. Get display calibrator.
2. Get printer calibrator and / or printer profile.

Depending on the process, in some large volume precision press work, you have to actually check the press run ever few hours (no sleep!) while the press is running.

 

[torana355]

I will respectfully disagree with you.

The promise of RGB (and PDF) for years has been the capability for one file to be correctly color separated using whatever profile it is processed with when rasterized. The workflow is sound, provided your vendor is likewise profiling their color separators, proofers and presses.

You should always check with your vendor before assuming this is the case.

Well im currently in charge of colour profiling 2 Komori 10 Col presses, 1 Heidelberg 10/col, a Large format KBA press and 3 Epson Proofers and i have to disagree, for starters you don't rasterize a pdf in any normal workflow, you want to keep type and ai files as Vector where possible. Now you can setup a full RGB workflow, IF the printhouse is setup for it and each stakeholder is using the exact same RGB profile from the monitor calibration all the way to the Rips RGB to CMYK conversion profile. I have extensive experience with this and the results were varied and nowhere near as consistant as our current ISO CMYK workflow. Just being able to get all our clients to operate under the same RGB standard is near impossible so we had to have different workflows for each client. In the end it is much easier and much more consistant for the client to just convert their files to CMYK.

 

[monokakata]

1. Get display calibrator.
2. Get printer calibrator and / or printer profile.

Depending on the process, in some large volume precision press work, you have to actually check the press run ever few hours (no sleep!) while the press is running.

I'd love to see a large high-quality printing operation at work.

Unfortunately, my press is a print-on-demand operation in Tennessee and I'm out in the Pacific. So I won't be standing near their printers looking at output anytime soon.

I did get an e-proof today for the cover I was talking about. I bailed on the red as being too risky and went with a blue-black, and the e-proof seems to show that they nailed it. The physical proof comes later in the week.

The Spider calibrator hasn't arrived yet. Yeah, the wide Pacific, plus I'm on an outer island.

Thanks to all of you. I learned a lot and have used the 78-68-58-94 black.

I'm going to guess that it's OK to link to the press website. I don't get paid money for my work there but there are other benefits.

www.saddleroadpress.com

The cover in question can be found under Catalog/Forthcoming/The Animals. The author wanted to use the pig etching, but it was a dull orange. The book's a kind of wild and crazy memoir so I convinced her to let me make the pig wilder and crazier than it was in the etching.

I'm finishing up the cover for Body on the Wall (also in Forthcoming) this week. It's a much easier one with no large single-color areas.

 

[blanka]

Oh, and try to stop working in JPEG. I don't know if the covers are "photo-like", but if they use any objects in solid colours, like type, then DON'T use JPEG. It will introduce significant edge-artifacts that are very visible on rasterized CMYK print. Especially edges with magenta/red are becoming horrible.

 

[monokakata]

Oh, and try to stop working in JPEG. I don't know if the covers are "photo-like", but if they use any objects in solid colours, like type, then DON'T use JPEG. It will introduce significant edge-artifacts that are very visible on rasterized CMYK print. Especially edges with magenta/red are becoming horrible.

I'm not sure what you mean, but I'd like to know.

I begin a cover either with a scan (tiff) or maybe a raw image sent over from Lightroom, or a full-strength JPG sent over from Lightroom. As soon as I get it into PS, I convert to CMYK (sometimes to grayscale). I work on it in PS as appropriate, including creating some but not all of the text.

When I'm satisfied, I export a JPG highest quality and place it in the InDesign template the printer prepares and emails to me for the book (given size, given number of pages (so the spine can be correct), given UPC, etc.) and I'll generally drop the blurb text onto the back cover in InDesign.

Then I export a PDF, run it through Adobe Acrobat X to check compliance with PDF/X-1a:2001, which is what my printer requires me to send, and upload it.

If I'm not pleased with the cover after I've placed one or more JPGs from PS, I flush the JPGs, continue working in PS, saving the project as a PSD, and then export another JPG.

So unless I'm not understanding what "working in JPEG" means, I don't think I'm working in JPEG. But maybe I am. I'd like to know, if so. I'm learning a lot here.

 

[Consultant]

They should have a printer profile you can apply. Once you calibrated your display of course.

I'd love to see a large high-quality printing operation at work.

Unfortunately, my press is a print-on-demand operation in Tennessee and I'm out in the Pacific. So I won't be standing near their printers looking at output anytime soon.

I did get an e-proof today for the cover I was talking about. I bailed on the red as being too risky and went with a blue-black, and the e-proof seems to show that they nailed it. The physical proof comes later in the week.

The Spider calibrator hasn't arrived yet. Yeah, the wide Pacific, plus I'm on an outer island.

Thanks to all of you. I learned a lot and have used the 78-68-58-94 black.

I'm going to guess that it's OK to link to the press website. I don't get paid money for my work there but there are other benefits.

www.saddleroadpress.com

The cover in question can be found under Catalog/Forthcoming/The Animals. The author wanted to use the pig etching, but it was a dull orange. The book's a kind of wild and crazy memoir so I convinced her to let me make the pig wilder and crazier than it was in the etching.

I'm finishing up the cover for Body on the Wall (also in Forthcoming) this week. It's a much easier one with no large single-color areas.

 

[monokakata]

They should have a printer profile you can apply. Once you calibrated your display of course.

I can query them. This is all they say in their instructions, although they do mention ".eps files":

COLOR:
All images in a cover file must be CMYK. Images that are less than 200 dpi may be rejected for higher-resolution submissions. (resolutions this low are considered too low in quality for printing)
• BODY TEXT: For best results for text that is 24 pt. or below, please use 100% black only.
• BLACK/RICH BLACK/DENSITY: We recommend a rich black with CMYK values = 60% Cyan / 40% Magenta / 40% Yellow / and 100% Black. CMYK total value should not exceed 240%. Elements should not be built in ‘Registra- tion’ (100% of all colors). *Files sent with densities higher than 240% may be rejected for correction.
• SPOT COLORS/RGB: Please convert all spot colors (PMS / PANTONE) with/without transparencies to CMYK, (even in supporting .eps files). *Note: RGB files received will be converted to CMYK before printing. RGB / spot color with/without transparency may produce unexpected color results when printing.
Any dissatisfaction with color shift will be the publisher’s responsibility to correct.

The calibrator arrived and I'll get to work with it this afternoon. Unfortunately because of deadlines both covers have gone off to Lightning Source already. But for the next job, which won't be for a few months, I'll have a calibrated monitor.

 

[Consultant]

I can query them. This is all they say in their instructions, although they do mention ".eps files":

COLOR:
All images in a cover file must be CMYK. Images that are less than 200 dpi may be rejected for higher-resolution submissions. (resolutions this low are considered too low in quality for printing)
• BODY TEXT: For best results for text that is 24 pt. or below, please use 100% black only.
• BLACK/RICH BLACK/DENSITY: We recommend a rich black with CMYK values = 60% Cyan / 40% Magenta / 40% Yellow / and 100% Black. CMYK total value should not exceed 240%. Elements should not be built in ‘Registra- tion’ (100% of all colors). *Files sent with densities higher than 240% may be rejected for correction.
• SPOT COLORS/RGB: Please convert all spot colors (PMS / PANTONE) with/without transparencies to CMYK, (even in supporting .eps files). *Note: RGB files received will be converted to CMYK before printing. RGB / spot color with/without transparency may produce unexpected color results when printing.
Any dissatisfaction with color shift will be the publisher’s responsibility to correct.

The calibrator arrived and I'll get to work with it this afternoon. Unfortunately because of deadlines both covers have gone off to Lightning Source already. But for the next job, which won't be for a few months, I'll have a calibrated monitor.

Maybe this will help. Illustrator:
http://help.adobe.com/en_US/illustrator/cs/using/WSfd1234e1c4b69f30d2a5051004d659b1c-7ff6a.html