Do numbers lie: 93% Black for offset printing?

[lowep]

Preparing photos for offset printing, the best advice I have come across is that a value higher than 93% black for a CMYK image should not be used as this is the maximum an offset press can handle?

However, when I use image > adjustment > "selective color" > blacks > black in Photoshop to reduce the black of the darkest pixel to a value of 93% K according to the pixel info, the result is the image on my calibrated screen with proofing set to CMYK looks somewhat washed out.

Do I guess correct this is simply because no screen (unless it is specially customized) can accurately reproduce the end result of offset press printing on paper, so it is the screen that is misleading whereas the numbers do not lie?

Or am I barking up the wrong tree?

[LenR]

Prior to the days of color management the way I converted rgb files to cmyk was to call the printer and ask for these three bits of information regarding their press condition:
1) Start point for cyan
2)Maximum black
3) Total ink

I would take the provided values and plug them into ResoLut and get a file that printed perfectly every time.  Clients came to me from Hi and Low for this service.  Ahh... those were the good old days:)

Today I would open Levels and set the end points to 20 and 240 and probably be be done with it.


Don't use Selective color for what you are trying to do.

[lowep]

Today I would open Levels and set the end points to 20 and 240 and probably be be done with it.


Don't use Selective color for what you are trying to do.

Thanks Len for this advice,

So the numbers could seem to lie if one doesn´t understand the language (as I obviously don´t  )
ie 93% K is not a good indicator of where the darkest black should be?

This looks like another viable solution for dummies here

[Wayne Fox]

So just curious if that’s still standard practice, sounds a little dated to me. I admit I don’t do offset, but why wouldn’t the best route be acquiring a CMYK profile for the offset output device and soft proof, and let color management handle mapping the black point. 

[ColourPhil]

Hi,
I second what Wayne said. Have been involved in Offset and Colour Management for too long. Just contact your printing co., ask them which profile to use, and if not a common one, can they supply it. In Photoshop, under 'Edit' go 'Convert to Profile' (NOT Convert to CMYK as you'll get the default which could be anything), use either Perceptual or Relative (with Black Point Compensation) for rendering.
The correct CMYK profile (which should only be SWOP if web offset) will give you the correct Black, Grey Balance, Total Area Coverage, etc.
That's the way we do it today.
Cheers,
Phil
www.colourphil.co.uk

[Phil Indeblanc]

Yes, you should be matching (Likely web coated) SWOP standard in the US (FOGRA for EU).
A color match depends on your printers calibration and cleaning practice. The other thing you should consider is him being flexible to get max Total ink down. Many shops try and keep ink usage low at under 280 total ink. I have done prints with 320..but paper makes a difference.  Make sure your work is properly converted to CMYK. Some of those amazing sunsets will flatten out. If your blacks arn't "rich" they can have color etc.  You best bet is to run a Contract Proof, or at least a profiled proof to get a ballpark of your final output.  I make contract proofs and prep files for offset on a regular basis.

[LenR]

Your correct Wayne, today (In the US) I would convert to SWOP for general commercial printing.  
I went digital in 1991 and for that whole decade virtually no presses could (would) provide a profile.
Or I would get a response like "do you want TODAYS profile?"
I could see why it would be tough profiling a press given the variables they deal with.
Giant warehouses with seasonal temperature and humidity variations, custom paper types etc.
The "numbers" was something they understood and used as targets for their imagesetters and the rest of their workflow.
Today most seem fine accepting SWOP profiles and do well with it.
I have so much respect for these people... you wouldn't believe the talent it takes to do what they do.

[lowep]

I could see why it would be tough profiling a press given the variables they deal with.
 
I have so much respect for these people... you wouldn't believe the talent it takes to do what they do.

+2!

I also have a lot of respect for the people running the press I am working with here in Mexico. Like you Len and others who have been kind enough to reply to my question, they have many years of experience in a rapidly changing industry and as you say have a tough job juggling many different variables. Also they have to run hard to make ends meet and do not have as much time as I do to gild lilies.

I asked about a profile a few days ago but so far nothing has materialized. Maybe one will turn up but it is not certain.

As suggested my first step was to convert to CMYK using the standard Photoshop settings for offset prepress but I was not happy with the result, so opted to experiment with RGB to CMYK conversion via multichannel. That was an interesting detour that produced some very nice results and also some unexpected headaches, including all the total ink levels going awol as they would not have done if I had stuck with the straight Photoshop RGB-CMYK conversion engine.

In the end I found the best results I could get was by combining the files automatically converted by Photoshop together with the ones I converted via multichannel. I was happy with how the results looked on my calibrated screen set to CMYK proofing. But that still left unsolved the problem of how to correct the amount of black and maximum ink levels of the files that I had already converted to CMYK and worked on. Hence my post this morning.

As you kindly pointed out Len trying to do this by fiddling with the black slider in the selective color dialog box to get a K setting of 93% really was barking up the wrong tree. I appreciate very much you pointing this out as it motivated me to stop bulldozing ahead up this wrong road and to look more into other options, which lead me to the numerical solution I linked to in an earlier post that has (for me) turned out to be a good solution.

At least now I am confident the result is as good as I can manage to get to before I sit down with guys at the printers and figure out talking together with them and looking at proofs how to do the final tweaking. Unless of course there is an earthquake, a revolution, a birthday party or a funeral.
    

[pfigen]

Ninety three percent in the black is generally plenty for a maximum black as long as the other inks - cyan and magenta being more important for density than yellow - all add up to a reasonable total ink. Almost all of the old Lino profiles from the early 90's maxed out at 93 percent black.

How your monitor displays your file that has been brought down a few points in the black depends on the profile that you're viewing the file through and what the total ink in combination with the max black is, as well as your actual monitor calibration. You can also rebuild many profiles with whatever total ink, black start and black max numbers you want, if you have the software and the data to do it. I often use ProfileMaker or Profiler to build alternate profile versions using different ink totals or, especially, different black generation schemes.

If you're only pulling the black down a few points like you say you are, then Selective Color is a fine way to do that. I've used that for thousands of files over the years to make very fine last minute tweaks. Generally though, if you're knocking the black down seven or eight points, you also might want to bump the others up a couple to keep the total ink in the file where it needs to be.

[lowep]

I've used that for thousands of files over the years to make very fine last minute tweaks.

Thanks for this good suggestion. I have managed to get the total ink down to about 290 for all my files that if I understand right is the most important concern, leaving black values higher than 93% to be tweaked as necessary when I have the proofs.

[pfigen]

"As suggested my first step was to convert to CMYK using the standard Photoshop settings for offset prepress but I was not happy with the result, so opted to experiment with RGB to CMYK conversion via multichannel."

When you say you did you initial conversions using the Ps defaults, what exactly were your default settings?
Where did you come up with 290% Total Ink?
What kind of press are you printing on?
When you say you were not happy with the results, what are you basing that on - what the files look like on screen, what the printer's proof looked like, or actual press output?
If it was from a printer supplied proof, what type of proof was that and to what standard was that proof calibrated to?
Remember you can have two very different files, in terms or black density proof virtually the same but act quite differently on press.
Finally, what is the nature of the images you're going to press with?

[lowep]

Good questions that I attempt to answer:
When you say you did you initial conversions using the Ps defaults, what exactly were your default settings?
North America Prepress 2 with Intent changed to Perceptual
Where did you come up with 290% Total Ink?
Here, a good resource though by a different printer than the one I am using.
What kind of press are you printing on?
Heidelberg, looks like this, exact model not sure.
When you say you were not happy with the results, what are you basing that on - what the files look like on screen, what the printer's proof looked like, or actual press output? If it was from a printer supplied proof, what type of proof was that and to what standard was that proof calibrated to?
(i) what the default converted CMYK files looked like on my calibrated NEC P221w monitor in Photoshop CS6 (view > proof colors > working CMYK)
(ii) one printer supplied "proof" of a photo from a digital printer said to be "95% similar" to the final offset print, which I doubt.
Finally, what is the nature of the images you're going to press with?
environmental portraits

I should also add my initial concern about the images looking washed out arose when I tried using selective color to reduce the black level in the black dialogue box until the K value in the info. panel hit 93%, a dog-eared detour that I have got over thanks to the good advice received here in response to my post and other input. Now the files no longer look washed out. But I am still quite nervous about how the final result will be.

[pfigen]

I don't think I'd put much credence in either of those references. Really, in order to make use of what's presented there you'd already have to be proficient enough to not need it.

While Deke's video is interesting, and it does provide a way to keep pure colors as pure as possible, it also craps on the rest of you image, which is something you really don't need for people. I didn't look for the subsequent video about what I'm presuming to be Under Cover Removal or Gray Component Replacement (UCR and GCR), which are common schemes for sucking neutral amounts of cyan, magenta and yellow from the shadows replacing with black - to lower the total amount of ink and help neutrals from shifting on press. You're generally much better off leaving that to the correct profile for your output and not trying to jury rig on top of jury rig.

The Total Ink page is really in reference to the requirements of a specific company and right now you have no idea if that applies to your situation. In addition, if you look at his example, it's clear that he doesn't know what the hell he's doing. He's already at 100 percent black and now he's bumping that up another +3. It can't go higher than 100 percent, but that small bump can certainly affect the near blacks and block them up, sacrificing shadow detail. Not great advice overall.

I would not move forward without getting more definitive information from your printer. It would really suck if you had to make rounds and rounds of proofs to get them looking good on their proofer only to have the whole thing fall apart on press if they can't match their proofs. I would ask them for some actual samples of both proof and final printed piece to see what they mean by "95 percent". If they are unwilling or unable to get you a good profile for their press/proofer, and you're not able to generate a profile on your own, you're sort of in a no win situation. Plus, you may also have a deadline you need to meet.

If you get a proof from them that is at least close, and you are able to go to a physical press check and supervise the process, you should be able to get the press man to make corrections on press that will please you. I would try at least once more, and push to get to the person in charge of prepress at the plant - for the correct profile or at least one that will work for you.

Good luck there. It's clearly not an ideal situation.

[lowep]

Yes it is a good learning opportunity  

Fortunately, forewarned is forearmed.

The beauty of adjustment layers is that everything can be tweaked right up to the last moment.

I also noticed that bumping the black level in the black dialogue box of the selective color adjustment tool up +3 "seemed" to be only adding mayhem to madness, at least on my monitor, so reduced to +1 or +2 rather than +3 depending on what it did to the image, but was fearful of going further down than that. This is what I will most likely look at most carefully while listening to the printer´s opinion about what needs to be done when I get whatever additional proofs I can before the printing starts and see what happens.