ID Export to PDFX1a different from PS to PDFX1a

[Colorblind]

Have a very simple file that has a white box on top of a colored background tiff image. Both boxes have bleed about 0.5" outside trim. Tried the direct export from ID to PDFX1a using 0.25" bleed and compared to a file generated from the ID to PS to PDFX1a (same PDF settings) and found out that the PDFX1a generated from ID has dark lines from whats under the white zones of my file (not atomic region because they print when sent to my ApogeeX APPE rip) and shows clearly in Acrobat. The PS to PDFX1a doesn't have that problem and is a better representation of the ID file IMO. Any opinion?

 

[leonardr]

Any opinion?

Not without seeing actual PDF files....

 

[Dov Isaacs]

Need Files / Try PDF/X-4

Need Files / Try PDF/X-4

I fully agree with Leonard's assessment that we would need to see the files (and I'd also include the source files and joboptions files used) in order to assess what is really going on.

Having said that, if you are using a RIP that directly interprets PDF - and your Agfa ApogeeX RIP with the Adobe PDF Print Engine certainly does that very well, you really should be exporting PDF/X-4 with live transparency and color management! PDF/X-4 was actually specified with that technology directly in mind!

- Dov

 

[Lukas Engqvist]

If i am not mistaken the thin hariline is the underliying pixel image?
Does the underlying image have a crop box? If so then what you are probably seeing is in the export to PDF Indesign sees the cropbox as being outside of the bleed, and thus makes it redundant. and so your pixel is not masked as tight as in the PD to PDF workflow. The PS distiller will not be removing redundant masks and therefore give you a cleaner bleed.
It could also be an issue of smooth graphics where the PDF is antialiasing the pixels and therefore you are seing a hairline.
Would be interesting to know, if you have the objects withing printing area do you see the same thing?
If you turn of smoothing graphics in acrobat do you still see the eclipsed bitmap image.

 

[Prepper]

PDF-x4

PDF-x4

Is it safe to use the Downsampling with Automatic JPEG? I'm assuming it would be since its used in PDF-x4?

I have resisted using this but am getting larger and larger files with image placed in Indesign and scaled down making effective resolutions around 1000dpi each and 3D rendering models and such. Our 42pg magazine that was 830meg exported without downsampling is only 120meg exported as PDF-x4, this is a huge difference but is it 100% safe? I have recently been downsampling but using Zip thinking I'm getting lossless compression, is that correct? This gives me a 530meg file.

Thanks

 

[Colorblind]

Here are the PDFs for you guys to look at.

TBYLT02_turnstiles_side.pdf is the PDF exported from InDesign CS4
TBYLT02_turnstiles_side.indd.pdf was generated using PS to Distiller

The PDF setting that was used for both is the standard PDFX1a 2001 with downsampling at 72dpi and JPEG High compression (just so I could post my files here)

You can clearly see the line outside bleed area on the PDF exported from InDesign.

Dov, the reason why I use PDF/X-1a is because most publications we deal with require us to do so. I basically use my ApogeeX workflow to preflight/fix the PDFs it receives and to proof them. I haven't found a single publication (magazines, newspapers) that would allow us to send them PDF/X-4...

 

[Colorblind]

Dov, Leonard, the attached PDF file is the kind of spec sheet we get all the time from publications. This one has been updated last month and I received it today. As you can see, not only PDF/X-1a is the required PDF flavor but they want you to take the PS to Distiller road. A lot of publications are still afraid to take PDFs that haven't been made from PS.

 

[Lukas Engqvist]

Oh you didn't tell us the white shape had a drop shadow applied in indesign
If you look at the two as line art you will see in the postscript there is a large white object with a clipping mask. In the file from indesign you have just one shape a "pathfinder" function has been preformed to remove excess object.
You will also always get maths happening as you down sample, and some how the downsampling will be to a grid. the line is less than 1/2 a pixel (if ppi is 72)

now one more thing, if you look at the two files in illustrator you will se a structural difference.
The indesign file is a bitmap has a clipping path and is then in a group.

In the post script version of this
The bitmap is
…has a clipping path and is then in a group.
…that has a clipping path and is then in a group
…that has a clipping path and is then in a group
…that has a clipping path and is then in a group
...that has a clipping path and is then in a group

Yes, five levels of groups and clipping paths. (for each atomic region)

Now a drop shadow on the edge of an object would only appear like this on a bleed, sinse else you would have a clipping path around white object, and background or the shadow would protrude.
You are asking why is edge of the bleed not rendered in export, well it is not the artwork, that is what the bleed is for. Postscript does not have bleed boxes by definition, and therefore these will be another level of clipping paths. where all within the artbox is rendered meticulously, including irrelevant information such as downsampled details in the bleed, at the expense of making irrationally complex structure. (I remember the times we had to ungroup groups to get them through the rips).

It is to me an explanation of how those coding must draw the line some where to get rational code and it is up to us to understand those boundaries in which we are to work. In this simple file it seems like a bad decision, but remember the rip needs to render things that are invisible in a file, and had there been lotts of detail outside the bleed you would find that the decision to do away with what is outside the bleed (or should we say aputate since we are talking about bleeding? ;P) is on the whole a good one.

 

[Bret Hesler]

Is it safe to use the Downsampling with Automatic JPEG? I'm assuming it would be since its used in PDF-x4? ...

Downsampling is OK, JPEG is not recommended. I have seen many image quality problems with jpeg compression in PDF creation. Also, check out this article:
<http://blogs.whattheythink.com/blognosticator/2010/08/ansel-would-not-approve-part-ii>

I find it odd that Adobe has JPEG compression on by default in all of the predefined PDF settings, and even more odd that the submission specs say ,"Do not use JPEG compression." and then say to use the default adobe setting which uses JPEG.

Bret

 

[Lukas Engqvist]

@ Bret + Prepper
Might be idea to start a new thread if you have a different topic, makes it easier for all to follow.

In many jobs JPG compression at high or maximum is the least problem to worry about and JPG can be used without perceivable loss in quality, but it depends on the image. A poor quality image will be worse afected than a good image. Noisy images can be really messed up by jpg. But generally the surface of the substrate is more likely to cause image detail loss.

 

[Prepper]

PDF-x4

PDF-x4

Yes, Lucas, you're right about where to post, sorry, but while here....

Tried some files and JPEG is introducing artifacts into the files and although it may not be enough to notice I am uncomfortable "introducing" artifacts myself just by converting something. Our designers are shooting at high res, wanting to use FM screening, and keep everything as high res as possible.

Am I right that compression with zip is lossless?

Also say yesterday that Preps manual says not to use any PDF-X standard, but doesn't state why?

My workflow is exporting to PDF out of Indesign, impo in Preps, output on Xitron (Harlequin) rip to Screen platesetter. Have had problems with lines in PDFs, if I export as PDF 1.5 instead of 1.3 no lines, also lines seem to show up on screen in Acrobat and Preps but not in final output at rip. Preps says it can take 1.5 and rip says it can take up to 1.7, so to avoid the lines in steps b4 the rip should I just go with 1.5?

Thanks

 

[Lukas Engqvist]

The lines, if may be a due to "smoothing" in acrobat, and if so do not turn out in print.
Don't know why preps would advise against industry standards, and think they ought to explain how they are thinking.
Yes ZIP is lossless, and there is also a lossles form of JPG 2000, for higher versions of Acrobat.

 

[Colorblind]

Hi Lukas, great analysis. You're right, a drop shadow effect was used under the white box. I understand that both PDFs are made 2 different ways and I also understand that my APPE rip is supposed to print the dark lines since they are in the InDesign exported PDF. My point is that both PDFs generated by the very same settings are different when opened in Acrobat or printed. That could be one reason why publications that are used to PS generated PDFs are so reluctant to accept exported PDFs.

 

[rich apollo]

Colorblind,
The difference looks, to me, like the result of some rounding. The mask around the white element looks to be a slightly different size.

You're correct, printing and distilling will yield somewhat different results than exporting. Sometimes one will work in your favor, sometimes the other. On the whole I dislike PDF/X-1a - I would much rather have live transparency.

Prepper,
Absolutely! If your RIP can handle live transparency, then leave it in! The lines you're describing sound like the result of flattening. While most of the time they won't show in print there will come the odd occasion in which they'll bite you. Flattening can cause a host of trapping issues, and can make color management impossible.

 

[mattbeals]

PDF/X-1a has a place and a purpose. Color management should have already been applied and transparency flattened. It is still a very viable format for many applications. But you are right that PDF/X-4 is much more flexible and desirable. As you said "Sometimes one will work in your favor, sometimes the other".