Consistent Saturation Different Sources

bbq_roadrunner

New member
Hi,
I work at a production graphics house printing various flyers and marketing material on offset paper with a Didde web press.
Our customers are not computer savvy and generally send in there photos anywhere from professional photography to scans of polaroids shot in the 70's.

Question Is.... Since the source of these photos are so erratic, is there anyway to get the levels, lighting, ink saturation and overall print quality (not image quality obviously) to be consistent.

On more then one occasion I have had crappy photos print great and high quality images print poorly and vice versa. Is there a way to take the Photoshop values of a picture that prints great and apply them to all?? And if so what is the best route to take. Or if I'm completely off base, let me know.

I am currently using Adobe RGB (1998) / U.S. Web Uncoated v2 on all programs. Outputting files directly from InDesign or Quark to RIP.

Any help would be greatly appreciated.
 
I am currently using Adobe RGB (1998) / U.S. Web Uncoated v2 on all programs. Outputting files directly from InDesign or Quark to RIP.

On handling of RGB files/images:
While AdobeRGB is fine if you're capturing/scanning your own images, I think for an *ASSUMED* RGB working space, sRGB is the better choice....with your color management policies set to "Preserve Embedded Profiles" of course.
The logic goes like this: For "savvy" customers that have embedded their RGB working space, preserving/honoring their embedded profile is the right thing to do...the results will be what the customer intended....just don't go changing their embedded profile unless you know what you're doing.
For those customers not savvy enough to embed an RGB profile, there's a better-than-even chance that the image should be *assumed* to be in sRGB space. You can always *assign* a different RGB working space if you feel the image looks "wrong" with sRGB but then you're making your own judgement call.

I think "US Web Uncoated (v2).icc" is a reasonable choice for offset paper, at least until we get something that better represents uncoated offset papers.
US Web Uncoated has decent gray balance and total ink limit (around 250-260%). The two main gripes I would have with this profile are 1) the paper L*a*b* of 91/1/-2 is both not bright enough (should be a higher L*) and not nearly "blue" enough for uncoated offset and 2) the assumed dot gain of this profile is a bit low in my opinion (around 18% TVI). The result of this too-low dot gain would be separations that might be too "full" for your press conditions. You'd need a profile with a higher assumed dot gain that would produce separations with the proper amount of midtone reduction for your printing conditions. Of course, if you've applied dot gain compensation curves in your RIP that allow your press to mimic the dot gain/TVI found in US Web Uncoated, you'd be all set.

In thinking about what might be the source of your inconsistencies, I'm thinking proper file handling of customer-supplied images is most likely the problem. If you're "forcing" AdobeRGB on all incoming RGB images, that would be one area of concern. Use/honor the embedded profile if the image has one; if not, assign sRGB to the image and go from there. Once you've found the "correct" RGB profile for the image, then go ahead and covert to US Web Uncoated...I would recommend the perceptual rendering intent for an uncoated press profile but relative colorimetric w/ black point compensation might work also.

With supplied CMYK, same basic thing but with a twist....use/honor the embedded CMYK profile if it has one (unlikely) and then *convert* to US Web Uncoated to bring the TAC down and set the correct dot gain. For CMYK *without* an embedded profile, I would be inclined to assume/assign US Web Coated (SWOP) v2 to the image and then convert to your US Web Uncoated profile using relative colorimetric w/ black point compensation. "US Web Coated (SWOP) v2" is sort of the "sRGB" of CMYK.....with no embedded profile to go by, US Web Coated (SWOP) v2 is about the safest assumption you can make as to what flavor of CMYK you might be getting.

Hope this helps,
Terry
 
Consistent Saturation Different Sources

Let me try to condense Terry's very thorough answer a bit and add a thing or two. Every image on every page must be correctly tagged with a profile coming in. This is the only way the color will be interpreted and converted correctly for output on your press, presumably using an uncoated profile.

When they aren't tagged you have to guess, assign some likely profiles--sRGB for camera images, sometimes Adobe RGB. Scanner? We hope it's one of these because the other possibilities are endless. Just keep assigning different profiles until one looks right, then convert and replace the image. If you're going to uncoated, perceptual rendering intent is safe.

Tip: When an image comes up too saturated it means that its originating color space was smaller than your working space, assuming that the creator didn't actually make it too saturated intentionally. So if your working space is Adobe RGB, you might try assigning Colormatch RGB, for example (not used much these days but often some years ago). Even sRGB is smaller, and common. BTW, the reverse can easily happen: Working space set to sRGB, and incoming images created in Adobe RGB look dull. As Terry says, same phenomenon with incoming CMYK.

Color management has a certain logic, and it must be followed rigorously throughout the system, including the CM settings in all your applications. You are well advised to hire a good CM consultant to vet your system and put in place iron-clad procedures that guaranty the entire "chain of custody" of images from acquisition to press.

Good luck,

Mike Strickler
MSP Graphic Services
 
Any good white papers on color management?

Any good white papers on color management?

Working space set to sRGB, and incoming images created in Adobe RGB look dull. As Terry says, same phenomenon with incoming CMYK.

OK, see.... that's whats happening now but on press. Alot of images are coming out dull while others "pop" off the sheet. I understand there is going to be a certain amount of absorption by the paper which will kill brightness but it's bothering the hell out of me when some are good and others aren't. No customer has complained yet, but I'd like to figure this out.

As far as hiring a good Color Management Consultant, is there a company that specializes or would most of these guys be freelance?? Any recommendations. I hate hiring people, have them show up and then they wind up knowing less than our guys. Happens alot with our equipment techs.

I'm in California. I haven't seen any offferings in the California Printer or HorseTrader.

meanwhile I'll be switching over to sRGB, see how we fair. Thanks for all the help
 
Last edited:
I think you're talking about photo manipulation maybe color correction and not color management or calibration on your press. You said that the good photos "'pop' off the sheet" and others are dull on the same page, this tells me that your press and platemaking are fine. You just have "sucky" photos. I would compare the histograms on the good and bad photos and you will find that your highlights and shadows are more defined in the good pictures and "flatter" in the bad ones. This can be simply fixed with Photoshop's "Auto Levels" button in the "levels" or "curves" adjustments. Then I would adjust the midtones or even apply an "s" curve to give more contrast. Warning though, this may look good to you and me, but, your client might not like it. It's all subjective.

Printnewb
 
BBQ -

Both Terry Wyse and Mike Strickler are excellent CM consultants but are both loathe to say so here in forums. They do not promote their offerings and that is a testament to their ethics and abilities.

Mike is in CA.

Good luck.
Ian
 
Color Issues

Color Issues

Yeah, printnewb has a point. This is what I've been doing for each and every photo. Going in changing curves, levels, contrast, sharpness, whatever needs to be done. They always end up looking good on screen, then off to print. But

1. Lets say I have 2 photos, side by side in PS. They both look good visually. I put them in their distinctive layouts and RIP for plates. One prints dull and the other prints fine. Is there something I'm missing here? Saturation levels maybe???

and 2. If it is just some Photoshop adjustment that I have been missing.....is it possible to have one PS Action I can set up to automatically make most of the adjustments I need to keep all images consistent.....Or...... because the lighting, shadows, settings and resolution are all different will I have to manually make the adjustments based on each picture?

Thanks for all the Help

btw..... -Mike if your in southern california, would you be willing to come in for a few Hours??? I'll Pay cash....
 
Yeah, printnewb has a point. This is what I've been doing for each and every photo. Going in changing curves, levels, contrast, sharpness, whatever needs to be done. They always end up looking good on screen, then off to print. But

1. Lets say I have 2 photos, side by side in PS. They both look good visually. I put them in their distinctive layouts and RIP for plates. One prints dull and the other prints fine. Is there something I'm missing here? Saturation levels maybe???

and 2. If it is just some Photoshop adjustment that I have been missing.....is it possible to have one PS Action I can set up to automatically make most of the adjustments I need to keep all images consistent.....Or...... because the lighting, shadows, settings and resolution are all different will I have to manually make the adjustments based on each picture?
.

In regards to #2 above, I'm gathering you're looking for something to apply auto image enhancement rather than straight color management. Fuji used to market a product called C-fit that was geared toward this. I haven't heard much of it in recent years, but it can still be found on their website. FUJIFILM North America Corporation, Graphic Systems Division
 
Even if the photos look the same on you monitor the difference maybe in the CMYK separation profiles or something more complex like the colors in one photo are outside of the color gamut of the press.
It is simple to record your "actions" in Photoshop for quick one key stroke corrrection or to propel you past a series of repetitive adjustments. You may have various scripted actions depending on the condition of the photo.

Printnewb
 
This thread has prompted some questions in my little brain.

1) If a customer sends an image in CMYK mode to my print shop, my understanding is that the halftone values in the image do not change as it passes through the workflow.

2) The halftone values in the CMYK image may change though if there are some kind of curves being applied when the plate is imaged.

3) How the customer see's the CMYK image on their screen vs how I see the same image on my screen will be affected by what profile has been embedded (and whether I honor it or strip it out) in the image - however the halftone values in the image do not change. Just how they are displayed.

4) Irrespective of the embedded profile, the display of the CMYK image will also depend on the calibration and profiles associated with my monitor vs that of my customer - but again, the values in the image remains unchanged.

Put another way - if my image is CMYK and only consists of a 50% cyan tone then a 50% tone value will be sent to the plate (and may or may not be altered by a transfer curve at the plate imaging stage). However, the hue of that 50% cyan in my CMYK image will change according to what profile has been embedded in it and the calibration state of my display device.

Is all the above correct? Or am I confused? Or...?

Thank you, J
 
Last edited:
This thread has prompted some questions in my little brain.

1) If a customer sends an image in CMYK mode to my print shop, my understanding is that the halftone values in the image do not change as it passes through the workflow.

2) The halftone values in the CMYK image may change though if there are some kind of curves being applied when the plate is imaged.

3) How the customer see's the CMYK image on their screen vs how I see the same image on my screen will be affected by what profile has been embedded (and whether I honor it or strip it out) in the image - however the halftone values in the image do not change. Just how they are displayed.

4) Irrespective of the embedded profile, the display of the CMYK image will also depend on the calibration and profiles associated with my monitor vs that of my customer - but again, the values in the image remains unchanged.

Put another way - if my image is CMYK and only consists of a 50% cyan tone then a 50% tone value will be sent to the plate (and may or may not be altered by a transfer curve at the plate imaging stage). However, the hue of that 50% cyan in my CMYK image will change according to what profile has been embedded in it and the calibration state of my display device.

Is all the above correct? Or am I confused? Or...?

Thank you, J

#1: Depends on your workflow. Most folks would design their workflow such that CMYK values are not changed as it passes through the RIPing process (when you say "workflow", I interpret that specifically to mean your "RIP" workflow system such as Harlequin, Prinergy, Rampage, Nexus, etc., not your overall workflow). But it's becoming more common to introduce file/color conversions into the workflow itself. Much of this (and the questions that follow) become more about whether you see your job as delivering to plate/press the halftone values in the original file/image vs. delivering the *color* the customer expects...these can be two different things. I would contend that it's the prepress/printer's job to deliver on customer expectations and not adhering to some rigid notion that their job is to simply image halftone values to a plate.

#2. Certainly. I find many printers that consider themselves "anti-color management" and believe their job is to deliver exactly what the customer supplied to them in terms of halftone values....but these same printers don't have any qualms about altering these dot values via a dot gain compensation curve during plating! Like it or not, this is a form of color management that is altering the customer's data. If you're willing to take that step, why not do it correctly and start implementing a sound color management workflow?

#3: Of course. But, assuming you both are working on calibrated/profiled displays, you should essentially see the same thing as the customer, assuming you've honored their profile. As far as "stripping out" the embedded profile, I would say this is a questionable practice. At the very least, this is your clue as to both what the customer was seeing on their display (and possibly proofed) and the printing condition they separated for. You could elect to take this information and use it to perform a proper conversion to your specific printing conditions or simply ignore the profile and pass the values through your workflow as-is....but you should take this embedded profile as a way of the customer telling you what this image is supposed to look like. You ignore this at your own peril in my opinion.

#4: Yes, of course, but I would contend that if both you and your customer have taken the step(s) of calibrating/profiling your displays, there's a reasonable chance that what you're seeing on your display is not substantially different than what they were seeing. Sure, different profiling packages and measurement instruments will result in visual differences, but I don't believe these differences are enough to throw out the baby with the bath water. I'll take my chances with a calibrated/profile display than not! :)

As far as your "put another way" example, all things being equal, yes, you're correct....but you also need to recognize that there's several ways you set yourself up for an unintended conversion of these values if you do not have your applications and your workflow set up correctly....and the answer isn't to simply "disable color management". There's smarter strategies than that.

Regards,
Terry
 
Thanks Terry!

I find that sometimes people use vague or incorrect terms when posting questions and answers, which in turn makes me a bit confused about what I think are the basics of the process and what they are describing they're doing.
Thanks for the clear answer and reassurance.

J
 

PressWise

A 30-day Fix for Managed Chaos

As any print professional knows, printing can be managed chaos. Software that solves multiple problems and provides measurable and monetizable value has a direct impact on the bottom-line.

“We reduced order entry costs by about 40%.” Significant savings in a shop that turns about 500 jobs a month.


Learn how…….

   
Back
Top