Skip navigation
Currently Being Moderated

Here is a puzzle

Jul 4, 2013 9:37 AM

Tags: #effects #transparency #dpi #low #rez

Here is another puzzle

 

A client sent us a 4 Sheet poster (60”x40:0”) all images in the poster are set to 300dpi and placed in the InDesign file at 100%.

 

When I made a pfdx4a file the pitstop preflight failed with ‘resolution of colour or geryscale image is less than 200dpi)

After placing the file in Neo, we could see that the offending images were from the Transparency effect on the type boxes.

 

After some investigation if you apply a transparency effect, Outer Glow, mode to multiply with black and if the technique size is set more than 25mm it generates the transparency effect as a 150dpi image.

 

does anyone have any ideas

 

paul

 
Replies
  • Currently Being Moderated
    Jul 4, 2013 9:41 AM   in reply to Dewynters

    Did you mean 1a for the PDF? If so, why is it being flattened?

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 4, 2013 1:10 PM   in reply to Bob Levine

    @Bob – there might be some transparency effects that always will be flattened, even if you export to PDF/X-4.

     

    At least one as I  discovered: "Inner Glow".

    So I think, we can also add "Outer Glow". ;-)

     

    Uwe

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 4, 2013 1:39 PM   in reply to Laubender

    Absolutely no transparency effects are flattened when PDF/X-4 is exported from InDesign.

     

    However, if you placed content into an InDesign for which transparency has already been flattened (such as EPS or PDF/X-1a, for example), the damage of transparency flattening has already occured for objects in that particular content.

     

              - Dov

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 4, 2013 2:13 PM   in reply to Dov Isaacs

    @Dov – I'll send you an example where "Inner Glow" is flattened when PDF/X-4 is exported.

    It's been done with InDesign CS5.5. Don't know, if CS6 or CC will work differently…

    Can I contact you over PM about that?

     

    Uwe

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 4, 2013 2:20 PM   in reply to Laubender

    Sure! But probably can't get to it until Monday! 

     

              - Dov

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 4, 2013 2:40 PM   in reply to Dov Isaacs

    Done :-)

     

    Uwe

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 8, 2013 5:18 AM   in reply to Dewynters

    Using a 3rd Party PDF viewer/editor is not a good idea at all - although you may need this for your workflow I have never heard of Neo.

     

    How does it look in Acrobat? Perhaps it's just the way it displays in Neo?

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 9, 2013 4:13 AM   in reply to Dewynters

    You shouldn't be opening PDFs in Illustrator - Illustrator is not a multi-purpose PDF editor.

     

    If the people at Esko are telling you this is a fix - then tell them that it's actually not - it's a fuzzy workaround.

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 9, 2013 4:58 AM   in reply to Dewynters

    The question in my mind is whether this is really an actual problem. Glows and shadows have no detail that needs to be preserved, and 200 ppi is still within the range of printable for photos at 150 lpi in my expereince.

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 9, 2013 5:35 AM   in reply to Dewynters

    I can't answer that question, though I do agree it would be good to know if it's as designed.

     

    If the preflight can't tell the difference between an effect and a photo, it makes some sense to me to set the trigger value at 300, but that also means you need an intelligent operator who can look at the warnings and say "ignore" for cases where it doesn't matter.

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 9, 2013 8:39 AM   in reply to Dewynters

    Paul,

     

    This is a user forum. You won't often find Adobe employees here to answer your question.

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 9, 2013 8:44 AM   in reply to Steve Werner

    Although Dov did reply in this thread and was provided a file that evidently demonstrates the issue. So hopefully he at least communicates to the op.

     

    Mike

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 9, 2013 9:24 AM   in reply to Dewynters

    There's an official bug report form at Adobe - Feature Request/Bug Report Form, but you won't be starting a dialog, and I don't beleive there's a way to attach your file, so unless someone decides they need it you won't hear a thing.

     

    I'd suggest you contact Dov directly by email (he doesn't check his private messages all the time, but his email should be in his profile) and ask if he'd like to see your test file, too. I suspect he will, and he's in a much stronger position to get this pushed in the right direction.

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 11, 2013 6:06 AM   in reply to MW Design

    Although Dov did reply in this thread and was provided a file that evidently demonstrates the issue. So hopefully he at least communicates to the op.


    @Mike – I don't know if we could call "flattening a distinct transparency effect while exporting to PDF/X-4" a bug. As Dov stated, the export mechanism should not flatten it, but hey, maybe it's absolutely necessary to do that (because the effect cannot be replicated sufficiently as a PDF construct).

     

    Let's see what Dov can find out…
    We'll have to be patient.

     

    Uwe

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 11, 2013 6:11 AM   in reply to Laubender

    I've seen this flattening effect on PDFX4 too - but for the life of me I don't remember what file or what I did to "get around it".

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 18, 2013 5:14 PM   in reply to Laubender

    Sorry for the delay (feels like I am working a dozen jobs here @ Adobe!), but I finally got to look at the PDF file provided privately to me by Uwe.

     

    Using the tools at my disposal, I don't find any explicit generation of an image in the sample PDF file such as you would have generated by flattening the PDF file, but rather, there is a mask image associated with the “inner glow” effect definition. It isn't flattening, but rather a part of the live transparency definition. This is a case of an image that really isn't an image in the true sense of how we normally think of images.

     

    The issue of whether there is “pre-flattened transparency” or not having been determined (there is not), the next question is whether having a mask used in transparency rendering below a particular resolution is or is not a problem in terms of actual output production.

     

    With the sample provided by Uwe, no, there was no visible problem caused by the resolution of the mask. I don't know whether such a mask is actually causing visual problems with the output from the file discussed by the OP, Paul.

     

    Let's assume that there are no visual anomalies in the output and that the mask used in the transparency definition is effective for what it is supposed to be doing. Then why does it make any difference whatsoever what somebody's preflight profile says. The PDF/X-4 standard does not, repeat does not, have any minimal or target effective image resolution value. Thus, the preflight profile may be looking for potential problems, but it is neither a PDF/X-4 compliance profile nor a profile that can definitively advise as to whether a PDF file will produce optimal, high quality output, in which case, who cares! Ignore the results of the preflight when prudent.

     

    Let me give another example, in this case with nothing whatsoever to do with transparency effects …

     

    Suppose you take a computer screen shot and place it 1:1 (i.e., no scaling) into an InDesign document. The effective resolution of that image will be exactly what the screen resolution is, somewhere between 72dpi and 200+dpi. It certainly won't be 300dpi or higher! Let's assume for the time being that it is 96dpi, old style Windows-defined (although not necessarily actual) screen resolution. You export PDF/X-4 with settings that downsample any image over 450dpi to 300dpi. The result is a PDF/X-4 file that validates as PDF/X-4-compliant that has a 96dpi image. Should you don ashes and sackcloth over this because it will fail somebody's preflight profile? Absolutely not! You will absolutely not get any improvement in output by editing the screen shot in Photoshop to “increase resolution” prior to placing the image in InDesign. Why? Because the typical PostScript or direct PDF RIP will do the exact same imager interpolation when printing that you would have done in Photoshop. However, if you do “increase resolution” in Photoshop first, you may actually get some degradation at the RIP. Contrary to popular belief, 300dpi is not a magic number. It may be a good guideline for high quality print, especially for photographic images, but it is probably not the exact target resolution when raster images are resampled, screened, and separated at the RIP. The best advice we can offer is to do the least resampling of raster images in your content within your workflow and never upsample prior the RIP / rendering process.

     

    Clear as mud?

     

              - Dov

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 19, 2013 1:04 AM   in reply to Dov Isaacs

    Hi, Dov!
    Thank you for checking…

     

    That's crystal clear!

     

    (For German readers here: "Klar wie Kloßbrühe!") ;-)

     

    The issue of whether there is “pre-flattened transparency” or not having been determined (there is not), the next question is whether having a mask used in transparency rendering below a particular resolution is or is not a problem in terms of actual output production.

     

    It's no problem using an effect like that I provided you with for printing production.

    Other examples (Inner Shadow) would be:

     

    http://forums.adobe.com/thread/1256331?tstart=0

     

    http://forums.adobe.com/message/4405072#4405072#4405072

     

     

    But…


    There is one thing we have to consider if we take the examples of the links above:

     

    in case you want a searchable PDF, we have a problem here since text would not be text anymore after Export to PDF.

     

    But there is a workaround:

    Make a duplicate text frame on top of it with the exact text contents, set fill and stroke of the text to "None". Then the text information will travel over to the PDF and is still searchable…

     

    My 2 €-cents…

     

    Uwe

     
    |
    Mark as:
  • Currently Being Moderated
    Jul 19, 2013 7:06 AM   in reply to Laubender

    For better or worse, some transparency effects cannot maintain text using fonts for output (although within the InDesign document itself, the text remains editable as text). Inner glow is one such effect. Other effects, such as simple transparency and drop shadows do maintain text using fonts for output.

     

    For what it is worth (pun intended) your 2 €-cents are worth more that US$0.02. 

     

              - Dov

     
    |
    Mark as:
  • Currently Being Moderated
    Sep 17, 2013 10:18 PM   in reply to Dewynters

    I found this discussion searching for some clues to my own problem with inner glow, so I thought I'd post the only solution I could find.

     

    My strange situation involved placing Illustrator files into InDesign and applying an Object Style to the frame that was a simple 10% of cyan fill and a 2-pica inner glow of a darker blue swatch color. Most of the Illustrator files were fine, but if the AI file contained any gradients, my ID preflight profile would report an error: graphic frame is 150 dpi.

     

    I tried many fixes, including adjusting the AI file's raster effects to 300 dpi (even though there were no raster effects), and I changed the gradients from "color" to white to "color" to 0% "color," but nothing worked.

     

    The only thing that cleared the preflight error was to cut the graphic from my Object Style frame and paste it in front of that frame in its own frame with None Object Style.

     

    In my case, I couldn't simply ignore the preflight warning, since my client requires that vendors like me certify our PDFs with a PitStop profile they provide. These inner glow frames that  ID's preflight reported as low-res also prevented the PitStop certification from passing.

     

    Anyway, it's clear that something weird is going on with inner glow in ID. Maybe someday it will be fixed.

     
    |
    Mark as:
  • Currently Being Moderated
    Sep 21, 2013 9:14 PM   in reply to MrScreens

    Postscript to my previous post: frames with my Object Style with the inner glow that contained an AI file with a placed bitmap TIFF didn't report a preflight error, but when I attempted to print any page that contained one, I got the dreaded message "Printing Error: The Adobe Print Engine has failed to output your data due to an unknown problem."

     

    The fix was the same: cut the AI file from the frame, then paste it in front in its own no-style frame.

     

    Fun stuff!

     
    |
    Mark as:
  • Rob Day
    3,142 posts
    Oct 16, 2007
    Currently Being Moderated
    Sep 22, 2013 10:12 AM   in reply to MrScreens

    In my case, I couldn't simply ignore the preflight warning, since my client requires that vendors like me certify our PDFs with a PitStop profile they provide.

     

    It's too bad 300ppi has become so standardized. As Dov points out with his screen capture example nothing magic happens at 300. There could also be a reverse example where 300ppi isn't enough for certain kinds of images—like a photo with high contrast, small type.

     

    And what exactly happens when preflight finds a 200ppi image? Does the client sample it up?

     
    |
    Mark as:
  • Currently Being Moderated
    Sep 22, 2013 11:12 AM   in reply to Rob Day

    Actually, I think my client's PitStop profile fails a job if it finds something under 250 dpi, but ID's own preflight is reporting these graphic frames as 150 dpi, so they earn the dreaded BIG RED X.

     

    Pitstop offers variious tools to correct problems it finds (e.g., if it fails a PDF because it found rules under .25 pt wide, you can opt to selectively or globally fatten them up to .25 pt. and rerun the preflight). There is an option to upsample low-res graphics found by the preflight, but it had no effect in this case.

     

    If a vendor feels the PitStop failure is in error, we can say so. But if the problem turns out to adversely affect print production, we get to pay for the printer's efforts to fix it. All in all, I prefer that all my PDFs pass their preflight.

     
    |
    Mark as:
  • Rob Day
    3,142 posts
    Oct 16, 2007
    Currently Being Moderated
    Sep 22, 2013 12:44 PM   in reply to MrScreens

    There is an option to upsample low-res graphics found by the preflight, but it had no effect in this case.

     

    But that's my point upsampling is futile, it's not going make any meaningful improvement in image quality, and in the screen capture example it would be destructive. If I'm understanding Dov's post the mask resolution is for the glow, which doesn't need much resolution because there are no edges or details that need to be defined with the extra resolution.

     

    So a photograph of clouds might need less resolution than fine hair because there are no high contrast edges to define—300 or 250 are  arbitrary thresholds.

     
    |
    Mark as:
  • Rob Day
    3,142 posts
    Oct 16, 2007
    Currently Being Moderated
    Sep 22, 2013 3:00 PM   in reply to MrScreens

    Like in this case where the image on the left is at 100ppi and on the right 300ppi. By the time you output these to a halftone screen there will be no discernible difference in quality because there are no edges to define.

     

     

    Screen Shot 2013-09-22 at 3.36.49 PM.png

     
    |
    Mark as:

More Like This

  • Retrieving data ...

Bookmarked By (0)

Answers + Points = Status

  • 10 points awarded for Correct Answers
  • 5 points awarded for Helpful Answers
  • 10,000+ points
  • 1,001-10,000 points
  • 501-1,000 points
  • 5-500 points