Dan Margulis Applied Color Theory

 Ink Out of Nowhere?


Ink out of nowhere?
Posted by: "Marco Olivotto - LoL Productions snc"
Tue Sep 30, 2008 5:01 pm (PDT)

Hi all,

I would be very glad if someone could clarify something that happened to me today and which I cannot understand at all. Here's the scenario: a CMYK pdf is sent off to print. Proofs are ok, but the printed job is far too dark. Much darker than the original proof.

This may happen for a number of reasons, but one thing is quite amazing: one of the b/w pictures in the original pdf has a deliberately blown out white background. It measures 0 in all the CMYK channels, so the background should be the colour of the paper. When printed, the photograph has a greyish background instead. The process was this: RGB picture aptly converted to grayscale, convert to profile (ISO Coated v2, 300% ink-limit, as suggested by the printing company), import this into InDesign. The pdf correctly measures (0, 0, 0, 0) in the blown out area – but then some ink hits the paper nevertheless.

One interesting point: the original file of course measures (0, 0, 0, 0) *outside* the picture, where the actual white paper is. And this area is indeed white. The light grey background appeared only in the picture. This is scary!

This leads to two questions.

1. Where does the damned ink come from?
2. What happens exactly when the pdf leaves the computer in the printing factory and hits the RIP?

Question #2 actually means: is it ever possibile that some kind of wrong CMYK-to-CMYK conversion happened, and turned my whitest white into something that might as well be 5C 3M 3Y 0K, more or less? And if so, why only in the picture and not the area around it? They should be indistinguishable, in a pdf. I am always worried about black re-generation and so on, because nobody at the company is actually able to tell me how my CMYK values translate into the actual densities of the plates (unbelievable, I know, but I keep on getting confused replies which have nothing to do with the original question). So, whatever profile I use, I have no guarantee that it is preserved – or at least that the CMYK *values* I write into the file are preserved. That's also why I am reluctant to use High GCR, which would be the way to go in a case like this – b/w pictures. If the black is regenerated, then my work is useless. So, most of the time, I stick to the profile they suggest.

The bottom line is that everything is too dark. Shadows which obviously were meant to have detail (and had some, in the originale file) are plugged, and the client rejected the job. I may prepare a lighter version of the files, but in doing so I indirectly would admit some fault on my side, and I see none... If white turns to grey, then light grey is likely to turn into medium, medium into dark, dark into black and black into hell. Or not?

Thanks for any reply you may be able to come up with!

M.
---
Marco Olivotto

LoL Productions snc
1/A Via per Sasso
38060 - NOGAREDO (TN)
ITALY
Ph/Fax: +39-0464-490614

http://www.lolproductions.it
___________________________________________________________________________

Re: Ink out of nowhere?
Posted by: spoocobra
Tue Sep 30, 2008 5:44 pm (PDT)

Are they using some sort of ink drop technology? I know when we were setting ours up (Rampage Rip) we found that it added color (cmyk) to certain objects. Did you embed a icc profile? When we get pdfs with embedded icc profiles it changes the color of the pdf. I'd would love to see that pdf and what rip system is your printer using and guessing you didn't get color proofs for press match?

Kelly
www.universalprintingco.com
___________________________________________________________________________

Re: Ink out of nowhere?
Posted by: J Walton
Tue Sep 30, 2008 5:46 pm (PDT)

On Tue, Sep 30, 2008 at 11:45 AM, Marco Olivotto - LoL Productions snc wrote:

Here's the scenario: a CMYK pdf is sent off to print. Proofs are ok,
but the printed job is far too dark. Much darker than the original
proof.

You can pretty much stop here if you are looking to assign blame. If the printing company made the proofs and cannot match them, that is their problem.

This may happen for a number of reasons, but one thing is quite
amazing: one of the b/w pictures in the original pdf has a
deliberately blown out white background. It measures 0 in all the
CMYK channels, so the background should be the colour of the paper.
When printed, the photograph has a greyish background instead.

Someone at the company would need to explain this. It's difficult for us to figure out exactly what went wrong.

The process was this: RGB picture aptly converted to grayscale, convert
to profile (ISO Coated v2, 300% ink-limit, as suggested by the
printing company), import this into InDesign.

That's a recipe for bad black-and-white printing, and that's your fault. Why would you print a 4C grayscale picture with a skeleton black?

The pdf correctly
measures (0, 0, 0, 0) in the blown out area – but then some ink hits
the paper nevertheless.
One interesting point: the original file of course measures (0, 0,
0, 0) *outside* the picture, where the actual white paper is. And
this area is indeed white. The light grey background appeared only in
the picture. This is scary!

Sounds like there's some re-conversion going on if only images are being affected but nothing else. Again, though, it's hard to say without being there.

is it ever possibile that some kind of
wrong CMYK-to-CMYK conversion happened, and turned my whitest white
into something that might as well be 5C 3M 3Y 0K, more or less?

It's possible, yes, but that's a bad conversion. Profile-to-profile conversions are one thing, but adding tone to a highlight is just bizarre.

I am always worried about
black re-generation and so on, because nobody at the company is
actually able to tell me how my CMYK values translate into the actual
densities of the plates (unbelievable, I know, but I keep on getting
confused replies which have nothing to do with the original
question).

To be honest, I have a hard time believing that the printing company can't accurately answer a question about dot gain. I am more apt to believe that they have answered the question, but the answer was not understood. What did they say?

So, whatever profile I use, I have no guarantee that it is
preserved – or at least that the CMYK *values* I write into the file
are preserved. That's also why I am reluctant to use High GCR, which
would be the way to go in a case like this – b/w pictures. If the
black is regenerated, then my work is useless. So, most of the time,
I stick to the profile they suggest.

What possible benefit do you have by doing this? If you convert properly (using Heavy black) and they leave it alone, that's great. If they convert it to your profile, then you're no worse off than if you'd converted it improperly yourself. If you convert it using a Heavy black and then assign their suggested profile to the image you stand a good chance of having it go through unchanged.

The bottom line is that everything is too dark. Shadows which
obviously were meant to have detail (and had some, in the originale
file) are plugged, and the client rejected the job. I may prepare a
lighter version of the files, but in doing so I indirectly would
admit some fault on my side, and I see none...

When you prep the files you are taking on some of that liability. So it's impossible that there's no fault on your end, IMO. But if they provided proofs and said they could match them and they didn't - that's their problem and that's the only problem you really need to solve.

J Walton
___________________________________________________________________________

Re: Ink out of nowhere?
Posted by: Michael Jahn
Tue Sep 30, 2008 5:49 pm (PDT)

Hi Marco,

As this is not a color theory type question, but more of a "I don't really get how color management works" technical support type question, well, I 4 one will not be rude and say RTFM, but really, I think this is the sort of thing Dan frowns upon here...

but - i am feeling generous today....

On Tue, Sep 30, 2008 at 11:45 AM, Marco Olivotto wrote:

1. Where does the damned ink come from?
2. What happens exactly when the pdf leaves the computer in the
printing factory and hits the RIP?

Jahn comments - if there was an embedded profile (that is, someone assigned a profile to that image, then has InDesign set up (Edit > Color Setting) and has the "Color Management Policy" set to "Preserve Embedded Profiles" - AND they then exported to PDF/X thur some "Adobe PDF Preset" which has the settings ( In the Output section, in "Profile Inclusion Policy" - set that to "Include Tagged Source Profiles" - and you then (ignorantly or otherwise) APPLY or PROCESS these embeded profiles, this certainly could cause a situation where the image changes dramatically - but objects - or "non" object (like the paper white you are describing - does not.

I have a special set of profiles that is designed to turn flesh tones a bright blue. This can be set to an image, or a PDF file (Output Intent). I use this to test customer set ups, workflows and RIPS .

What is nice, is that the page looks perfect in Acrobat and prints with this really obvious blue cast on the flesh tones - to 'see' this in Acrobat, you need Acrobat Pro - (in Adobe Acrobat Professional - Print Tools > Output Preview).

I mention this here so I can justify why I am taking up so many eye balls here on the forum / group to answer your question - in a color managed workflow - what you might see in a PDF may not be what you get (especially if there are embedded profiles or output intenet inside that PDF!

- you probably had no idea that you could have 'seen' what you get using the Adobe Acrobat "Output Preview" tool...but, I suggest you try.

NOW, having said that, without actually seeing this PDF, I could be totally wrong, and your rip is doing something crazy or you have some setting or CRD that has gone wacko. Send me your PDF, and I will report back.

Question #2 actually means: is it ever possibile that some kind of
wrong CMYK-to-CMYK conversion happened, and turned my whitest white
into something that might as well be 5C 3M 3Y 0K, more or less?
And if so, why only in the picture and not the area around it?

Jahn comments - yes, and yes see above answer.

They should be indistinguishable, in a pdf.

Jahn comments - no, that is obviously not the case - as you have already clearly demonstrated to yourself.

I am always worried about...
..... a bunch of worries)

Jahn comments - as Dan has said many times, just because someone claims to know what you should do, and you believe him, well, that certainly does make you or anyone else "right" - all your 'worries' are caused by your not understanding how color management 'works' - you seem to be in the camp of "please, I HATE color management, make it GO AWAY and give me CMYK PDF!" - and while that is quite popular here, as you are in Europe - where PDF/X-3 is prevelant - perhaps you need to bone up on that...

The bottom line is that everything is too dark. Shadows which
obviously were meant to have detail (and had some, in the originale
file) are plugged, and the client rejected the job.

Jahn Comments - well, that is a shame. Ink does not magically appear in images unless something converted that image. my guess is that that "someting" was an embedded profile in the image stream, or some crazy output intent, or maybe some really strange CRD - or some combination of them all - again, just a guess.

Hope this helps in some way. This is indeed a good example where someone with a device link profile could take your PDF - without you doing anything - and "re-separatated" it - or, like I might do, use an Enfocus Action List to extract and remove the embedded profiles and send you back a nice clean PDF that might behave a little better in your "we are not all that sure what happened" environment.

Hope this helps in some small way, and good luck recovering your time or selling that invoice.

--
Michael Jahn
___________________________________________________________________________

Re: Ink out of nowhere?
Posted by: Michael Jahn
Tue Sep 30, 2008 6:28 pm (PDT)

WHOA - hey, that may be EXACTLY right, but not if "the original proof" was made by someone other than the printing company !

- like (for example) - if Marco walked in with the proof, and the files, and said, "hey, here you go, run it" - well, I can certainly see how I might have some proof (where the color management policy was set to "IGNORE" - or (gasp) there was no ability to actually observe and HONOR or APPLY the profiles -

having said that - J Watson is perfectly correct - if the PRINTER made the proofs (and they were fine and signed off on ) - i can't imagine them even bothering to call you up and ask you to look at the press sheets if they were significantly different than the proofs that that same PRINTER created...(which is why i suspect that they DIDN"T make the "orginal proofs"...

oh, and just because I want to be silly;

"Why would you print a 4C grayscale picture with a skeleton black?"

I can name many reasons - in fact, in Japan, they do not even RUN a black on any images (only CMY) - black is only used for the plate that contains the type, and this black is POUNDED into the paper. Black ink is the most offensive color dulling part of most printing processes - there are color separations methouds that are routinely used in gravure and flexo where I might print a greyscale image in CMY with no black whatsoever.

But I agree wholeheartedly, that in most normal commecial prnting, certainly here in the US, we would not like a greyscale CMY image. Of cousre, if you really believe what all these G7 people are saying "hey, no problem - we LIVE to keep neutrals !"

LOL

On Tue, Sep 30, 2008 at 5:33 PM, J Waltonwrote:

You can pretty much stop here if you are looking to assign blame. If
the printing company made the proofs and cannot match them, that is
their problem.
___________________________________________________________________________

Re: Ink out of nowhere?
Posted by: Stephen Marsh
Tue Sep 30, 2008 6:28 pm (PDT)

Marco Olivotto wrote:

This leads to two questions.
1. Where does the damned ink come from?
2. What happens exactly when the pdf leaves the computer in the
printing factory and hits the RIP?

Too hard to answer the second point, only the shop in question can.

How do they handle/process/output your PDF files?

Do they print from Acrobat Reader, Acrobat Pro or Apple Preview to the RIP? What version of the software and what colour and print settings? Do they impose the PDF directly in say PREPS or other imposition software before sending to the RIP? Do they print your PDF from some program to PostScript and then impose the PostScript? Then what RIP and or other software settings may be at play here?

It is obvious that a "scum dot/s" or "higher than scum dot" value is being generated for CMYK raster elements that contain 0% CMYK values - but where is this value being generated if it is not in the PDF?

Are there ICC profiles in the image and PDF? Is the image untagged but the entire PDF has an ICC tag describing the output intent? Is the output at the RIP colour managed in some way and or using Absolute Colorimetric intent? There are many things for you (the printer) to look into here!

___

Question #2 actually means: is it ever possibile that some kind of
wrong CMYK-to-CMYK conversion happened, and turned my whitest white
into something that might as well be 5C 3M 3Y 0K, more or less?

It sounds like that is what happened...

___

And if so, why only in the picture and not the area around it? They
should be indistinguishable, in a pdf.

First guess: RIP software.

You are correct that in a colorimetric sense, that a CMYK image with values of 0CMYK and that a "blank" area of a layout PDF both read as 0% and you are also correct that 0% should output as 0% whether there is actually "data" there or not.

Damn CM settings and or RIP software!

The problem is that even though the empty page has a value of 0% CMYK and the raster image has values of 0% CMYK, the blank page truly contains no data - while to the RIP the image has "data" (even if blank). Some RIPs seem to freak out when there is an ICC profile in the PDF (and despite some vocal colour management consultants claims, these are not Dinosaur PS level 1 RIPs that are over 15 years old).

There can be similar issues with feeding various RIPs values of say grayscale mode 50% vs CMYK mode 0cmy50K and differences between raster and vector of these two different colour modes. As far as the end user is concerned - it is all 50% K whether the source is raster or vector or grayscale or CMYK mode (device values take prominence)...but some software puts prominence on colour mode and profiles and not the files numbers).

___

...because nobody at the company is actually able to tell me how my
CMYK values translate into the actual densities of the plates
(unbelievable, I know, but I keep on getting confused replies which
have nothing to do with the original question).

They do not wish to talk about a suspected undisclosed conversion? I can't imagine why not?! Do they now hire monkeys instead of trained prepress operators becuase clients and printers care more about the quote than the final print quality (compared to say 20 years ago)?

___

That's also why I am reluctant to use High GCR, which
would be the way to go in a case like this – b/w pictures. If the
black is regenerated, then my work is useless. So, most of the time,
I stick to the profile they suggest.

Two separate issues here, the values in the file are one issue, the profile you (may) tag is another - if you do choose to tag CMYK these files.

One can still convert using say a Heavy GCR CMYK or a Max GCR CMYK with some UCA, then tag it with the profile suggested by the print shop.

Sincerely,
Stephen Marsh
___________________________________________________________________________

More about ink from nowhere
Posted by: "Marco Olivotto - LoL Productions snc"
Tue Sep 30, 2008 9:43 pm (PDT)

Many thanks to all who replied to my request so fast... this is a very interesting discussion for me, and I will answer a few of the questions here...

Jahn comments - if there was an embedded profile (that is, someone assigned
a profile to that image, then has InDesign set up (Edit > Color Setting) and
has the "Color Management Policy" set to "Preserve Embedded Profiles" - AND
they then exported to PDF/X thur some "Adobe PDF Preset" which has the
settings ( In the Output section, in "Profile Inclusion Policy" - set that
to "Include Tagged Source Profiles" - and you then (ignorantly or otherwise)
APPLY or PROCESS these embeded profiles, this certainly could cause a
situation where the image changes dramatically - but objects - or "non"
object (like the paper white you are describing - does not.

I must explain one thing first which may explain why we get so stuck occasionally. We are not a graphics company proper, but a record company. What we do, 99% of the time, is booklets for CDs and DVDs, and we print small batches. Usually 1000-2000, sometimes – as in this case – only 500. We have a graphic designer in-house because it makes things a lot easier, but none of us has proper pre-press experience or formation. Yet we're very serious about the 50 or so works we do per year, and we have learned through 1) trial and error, and 2) self-study. But, believe me, I think both I and my graphic girl know a little bit more about CM than the average press operator we have to deal with.

I say this because, you understand, in order to have such small batches printed we can't simply speak to big pressing companies. We're just too small for them, and we should pay outrageous prices for such small runs – admitted they're interested in getting the job. So we're off with smaller companies, and in Italy *anything* can happen in such places. And anything goes, really. The only pressing company we can use in this area for similarly sized batches is the one we've used this time. And we've tried different ones... with varying results, but this one is the best, on the average. Yet the guy in charge of pre-press there has no formation at all in the graphics field, but he's learned what he does through ten years of practical work. I am not at all sure that he has a completely clear idea of what Assign vs Convert to profile means. Not joking – that's the situation. I hear you say it's flawed – but this is what we have available. So, you see, it's us who have to make moves to avoid disasters most of the time.

Another example. We printed a more bulky work with a reputed "big" company – 20 workers, supposedly high-quality output. We approached the pre-press guy and asked about profiles. "Give me RGB, I'll handle it." Oh well. "No, we want to give you CMYK." "Any conversion profile you prefer...?" "Uhm. We're getting close to FOGRA27 at the moment." Getting close? I wanted to ask: "at which speed and angle of impact?" I objected we needed a precise match with what we had in mind, and a proof was needed. Which profile for the proof, then? "We don't proof", the reply was. "We'll match your monitor, don't worry." In the end I used FOGRA27 and attended the press run with a gun pointed at the head of the guy in charge. "More yellow, please", and that was it. I am not joking at all. And this company invoices several millions of Euros per year. Welcome to Italy... ;-)

Going back to the specific case of my problem, in a few minutes I'll be uploading two files in the group section. One is called Marco_original_page.pdf and is the specific page from the original .pdf, re-exported with the same settings. The area of interest is of course that just above the head of the man. The other file is called Marco_scanned_page.jpg and is a scan of the printed page (WARNING: deliberately ridiculous contrast here to show the very light grey area – this scan does not at all reflect the actual printed result!). In this file, the patterns you see is print-through from the back of the page, but this is of course not the source of the problem. The grey is very light and difficult to scan properly, but it's so obviously visible by eye. The interesting bit is the pdf, I think; if any of you wants to have a look and report I'll just be glad. Anyway: the original picture had an embedded profile; but the settings in InDesign are correct as far as I can see. Certainly the profile was not exported into the pdf, or at least that's what my dialogue box says. ;-)

By the way, J Walton asked why I didn't use high GCR for b/w images, and suggested that if some kind of black re-generation happens, it might make things worse, but not worse than using a profile like the one I used. Good point, I agree. But that's not the issue here. I am not complaining about tinted greys or anything – I am complaining that something that any reasonably conceivable profile apt for printing would convert from RGB into (0, 0, 0, 0) was evidently not so when the ink hit the paper.

Always J Walton wrote:

To be honest, I have a hard time believing that the printing company
can't accurately answer a question about dot gain. I am more apt to
believe that they have answered the question, but the answer was not
understood. What did they say?

I am just glad that you live in a place where a printing company knows exactly about dot gain. If you read above, you'll realise this is not necessarily the case – not here at least. And the reply, I argue, was very well understood by me: it was "I don't know what happens to the CMYK values when we go to plate, but if you use ISO Coated V2 300% I guess they should be preserved." So much for CM...

Back to Michael Jahn:

- you probably had no idea that you could have 'seen' what you get
using the Adobe Acrobat "Output Preview" tool...but, I suggest you try.

Of course I did, and that's how I was able to tell that the pdf I sent in had (0, 0, 0, 0) in the supposedly white area. I always have a look at the output preview, all channels, to check if everything is ok.

NOW, having said that, without actually seeing this PDF, I could be totally
wrong, and your rip is doing something crazy or you have some setting or CRD
that has gone wacko. Send me your PDF, and I will report back.

See above. And, since you asked in a subsequent e-mail, yes they made the proofs. All in-house.

Jahn comments - as Dan has said many times, just because someone claims to
know what you should do, and you believe him, well, that certainly does make
you or anyone else "right" - all your 'worries' are caused by your not
understanding how color management 'works' - you seem to be in the camp of
"please, I HATE color management, make it GO AWAY and give me CMYK PDF!" -
and while that is quite popular here, as you are in Europe - where PDF/X-3
is prevelant - perhaps you need to bone up on that...

Point taken, and in principle I agree with you. I may be very wrong on this, but I never use any of the "X" specs just for this reason – and the kind of people I need to work with when it comes to printing. I mean, if I do something wrong with my files – fine. But then I am *responsible* for doing it, and I don't have to bang my head against some sales manager who won't take my protest about a badly printed job. So, if I screw up, it's my fault – which implies that if I don't, then the fault must lie somewhere else. And the best way I see to do this is to ask that whatever silly value I put into my files, it should be what goes to the plates. To be honest, while we were learning how to handle CMYK in a supposedly proper way we've experimented a lot, and I can't really see a terrifying difference between images printed with, say FOGRA39 or S.W.O.P. (profiles not embedded, so in principle the numbers should not be translated); yet, in principle, the two profiles are quite different, also due to different ink specifications. This particular case is wrong *way* beyond the difference between two average commercial profiles, and this means to me that something was done to my file. I don't know why, of course, and what. But I would dearly love to see the plates, believe me.

Many thanks again for all the input. I really appreciate it!

M.
---
Marco Olivotto

___________________________________________________________________________

Re: More about ink from nowhere
Posted by: Stephen Marsh
Tue Sep 30, 2008 10:02 pm (PDT)

Marco Olivotto wrote:

Going back to the specific case of my problem, in a few minutes I'll
be uploading two files in the group section.

When I look at this image in Photoshop CS3, by extracting the image data (not rasterizing the whole page), I find 1% CMY dots in the area which you say contains "no data".

So there is "some" data. Closer to the horizon there is a darker dot similar to 2c 2m 1y.

With neg plates and film, a 1% scum dot may not make it through to press - but with I presume CTP or positive film, it sounds like it has made it through to the final print.

Honestly, if no intent was given - I would have made the incorrect guess that you *did* require a minimum dot in the background sky so that the image would not be lost white on white. That you actually *do* wish 0% in this area and not a minimum 3-5% dot surprised me (I personally hate to see the sky disappear into the white page, so for me the sky should have a minimum dot so that you know where the image ends and the page begins).

Acrobat Pro CS3 shows areas of white, when Photoshop has 1%. I have never fully trusted the Output preview window in Acrobat Pro, I much prefer Photoshop.

Stephen Marsh
___________________________________________________________________________

Re: More about ink from nowhere
Posted by: Michael Jahn
Tue Sep 30, 2008 11:27 pm (PDT)

okay, now - related to this...

Of course I did, and that's how I was able to tell that the pdf I
sent in had (0, 0, 0, 0) in the supposedly white area. I always have
a look at the output preview, all channels, to check if everything is ok.

what were the setting you used in the Output Preview ?

- was there an output intent in the "Simulation Preview" - that is, did it say "Output Intent:FOGRA27" or was there some default profile selected....

Looking at the Output Preview is useless unless all parties agree to use one. I mean, you know you can change the output simulation - even if the file is CMYK, you can change the Simulation profile to an RGB one - so, this really is pretty useless without all aprties agreeing to a satndards vbased blind excahnge, which is the who point of the PDF/X - reliable blind exchnage between parties of color PDF files for printing.

--
Michael Jahn
___________________________________________________________________________
.
Re: More about ink from nowhere
Posted by: Kelly Ralston
Tue Sep 30, 2008 11:27 pm (PDT)

\Do you have pitstop for acrobat? I will look at the pdf an the image placed if it has and embedded icc profile I know for a fact (well at least with our Rampage workflow) that it will change color. I understand you are small but someone back me up you shouldn't ever do RGB if you are going to print! I can't even count the amount of PDF jobs that we get in an go straight to press without color proofs that still match or exceed what the customer want did you get an answer from your printer about why it happened?

Kelly Ralston
www.universalprintingco.com
___________________________________________________________________________

Re: More about ink from nowhere
Posted by: "Marco Olivotto"
Tue Sep 30, 2008 11:27 pm (PDT)

When I look at this image in Photoshop CS3, by extracting the image
data (not rasterizing the whole page), I find 1% CMY dots in the area
which you say contains "no data".

This is true, I've checked. I hadn't opened it in Photoshop; surely it reads all 0's in InDesign (before export) and Acrobat Reader though
.
Curiously, though, if in the same file I draw a rectangle of colour 1C 1M 1Y this is clearly distinguishable from the white sky. The sky is whiter. Also, I can't see any difference between such rectangle and one of colour 0C 0M 0Y, which agains looks a skosh darker than the white background.

But this sampling, I suspect, changes slightly from program to program. Try doing the same in Illustrator: the area of interest is practically a dead 0C 0M 0Y.

With neg plates and film, a 1% scum dot may not make it through to
press - but with I presume CTP or positive film, it sounds like it has
made it through to the final print.

This is CTP, by the way.

that the image would not be lost white on white. That you actually
*do* wish 0% in this area and not a minimum 3-5% dot surprised me (I
personally hate to see the sky disappear into the white page, so for
me the sky should have a minimum dot so that you know where the image
ends and the page begins).

Ah, agreed completely. In this sense, the mistake on paper is much better than the original – the sky is visible. The truth is that the picture was given by a photographer (the one actually portrayed) and the sky was totally white. We asked him, and he said he'd like it like that, so we just left it.

Acrobat Pro CS3 shows areas of white, when Photoshop has 1%. I have
never fully trusted the Output preview window in Acrobat Pro, I much
prefer Photoshop.

But the original picture, converted to CMYK before import in InDesign shows a correct 0C 0M 0Y. Then something must have happened in all these passages. Never mind, also because 1% per ink is no big deal per se. Yet the grey I see is honestly more intense than this. My bet would be around 3-4%, give or take another skosh.

Many thanks for your reply!

M.
___________________________________________________________________________

Re: More about ink from nowhere
Posted by: Stephen Marsh
Wed Oct 1, 2008 12:51 am (PDT)

Marco, If you are *sure* that in RGB it is perfectly 255r255g255b, then I presume that the scum dots are appearing when you convert to CMYK, most likely your choice of CMM (is it Adobe, Apple or MS?) and or rendering intent in combination with this particular profile are causing this issue (or so I am thinking at this point in time).

Stephen Marsh
___________________________________________________________________________

Re: More about ink from nowhere
Posted by: Stephen Marsh
Wed Oct 1, 2008 1:20 am (PDT)

Michael Jahn wrote:

which is the who point of the PDF/X - reliable blind
exchnage between parties of color PDF files for printing.

And the problem for some of us is that if there is a global CMYK profile describing the output intent - some service providers RIP's will incorrectly image the PDF data.

These service providers do not advertise this fact in their file supply guidelines, it is only when the proverbial hits the fan and fingers are being pointed that this comes out. To try to save lost time, one then asks the SP to strip the profile if they have the software (rather than one remaking the PDF and sending it off again) and they reply that "things sometimes go wrong with the PDF when we do that, it is best for you to resupply with no ICC profiles in the PDF".

Which leaves end users never knowing whether it is safe to tag their PDF or not when they use a new SP, so they don't tag (once bitten, twice shy and all that)!

PS: A side note of intereset, the default Illus CS3 setting for creating an Illustrator PDF (not PDF/X) does not include ICC profiles (either object or document level).

Stephen Marsh
___________________________________________________________________________

Re: More about ink from nowhere
Posted by: "Bevi Chagnon
Wed Oct 1, 2008 2:04 pm (PDT)

Marco wrote:

yes they made the proofs. All in-house.

Another thing to consider...
At what stage of the workflow were the proofs made? Before RIP/processing or after?

Given the knowledge of this printer, who knows how or when he made the proofs.

I've had some low-caliber shops make proofs before RIP/processing ... directly from the clients' InDesign or PDF files. Geez Louise, the client could have made that kind of "proof" right on their desktop inkjet printer!

Marco, if you need any consulting help there in Nogaredo, Italy (in the Italian Alps, close to Venice), just let me know. I love taking business trips to the most beautiful parts of the world. ;-)

Ciao!

--Bevi
(and yes, my Italian relatives chuckle over my name)

........................................................................
Bevi Chagnon | Adobe ACE: InDesign CS2 | www.PubCom.com
PubCom | Trainers, Consultants, Designers for Web, Print & Acrobat
Bevi's online tutorials | http: //www.CommunityMX.com/author.cfm?cid=5931
........................................................................
___________________________________________________________________________

Re: More about ink from nowhere
Posted by: "RJay Hansen"
Wed Oct 1, 2008 2:04 pm (PDT)

On Wed, Oct 1, 2008 at 12:58 AM, Marco Olivotto wrote:

This is true, I've checked. I hadn't opened it in Photoshop; surely it
reads all 0's in InDesign (before export) and Acrobat Reader though.
Curiously, though, if in the same file I draw a rectangle of colour 1C 1M
1Y this is clearly distinguishable from the white sky. The sky is whiter.
Also, I can't see any difference between such rectangle and one
of colour 0C 0M 0Y, which agains looks a skosh darker
than the white background.

Marco,

If you looked at the screenshot I uploaded, you notice that PitStop sees a bit less than a half percent of cyan magenta and yellow where I sampled it. This would explain why making a rectangle and filling with 1% of C, M and Y can be distinguished from the rest of the sky. Because the values in the sky are all less than 1%.

Since Photoshop doesn't show fractions in its percentages, it's rounding those up to 1 which is what Stephen is seeing.

RJay
___________________________________________________________________________

Re: More about ink from nowhere
Posted by: "RJay Hansen"
Wed Oct 1, 2008 2:04 pm (PDT)

I posted a jpeg in the group's files section just now (named "MarcoPDF_Pitstop.jpg"). It's a screenshot from Acrobat using Pitstop's eyedropper to read the background of the image in the PDF. You can see it's seeing .4% of C, M and Y in the spot the eyedropper is over. If you checked the image in Photoshop after converting to cmyk and still saw 0% of every color then you must have an issue when generating the PDF that's introducing the unwanted background "color".

RJay
___________________________________________________________________________

Re: More about ink from nowhere
Posted by: "Marco Olivotto - LoL Productions snc"
Fri Oct 3, 2008 1:07 am (PDT)

First of all, many thanks to all who replied to my original post. Just to close it, I think what happened is this: the "totally white area" had indeed a very small percentage of ink, less than 1% in all channels. I was more or less able to understand from the printing company that somehow their software promotes this to 1% (but it was a muddy information, I can't be sure of this), and some ink indeed hit the paper. Plus, there was a dot gain problem for sure, because what I see on paper is much more than 1%.

We're re-doing the job anyway. I have re-separated the original b&w images using a high GCR setting – just to protect from color shifts that might happen although this was really not the original problem. Let's now see what happens...

Many thanks again, and have a nice week-end, all of you!

M.
---
Marco Olivotto