Ideal PDF settings? Does anyone know?

[Phish]

Hi all,

I was wondering if anyone out there has any help in setting up PDF settings? Is there any websites of pages that can go through what you need to check in order to get what you want or what these options actually represent?

I have looked through web pages saying set your distiller to this or that however they do not actually go through why?

With so many different settings or preset up settings, how do we know what we are picking is what we are wanting?

Many thanks and i hope you can give us some feedback

Phil.

 

[Stephen Marsh]

In Adobe Software, a good place to start for print output is to use one of the PDF/X standards. You should not need to edit these, "they are what they are".

Another good source is the Ghent (PDF) Workgroup:

Application Settings - Ghent Workgroup

Some elect to use and or modify the Adobe "Press Quality" setting. Don't modify settings unless you know what you are doing. Adobe software has help guides that explain the PDF/X and various PDF creation settings.

Links explaining the various PDF/X standards below:

http://www.adobe.com/enterprise/pdfs/acr6_pdfx_faq.pdf
PDF Standards - PDF/X
An overview of the most important differences between PDF/A and PDF/X [callas software]
What is the difference between PDF-X 1, 2, 3, 4 and 5? It's a variety pack | Output


Stephen Marsh

 

[Phish]

Many thanks Stephen,

All i have to do now is understand which profile to use, lol, we just use the basic Press Quality or High Press Quality settings, these PDF/x setting are quite mind boggling tbh.

Considering we use 4 colour and spot, transparency and sometimes piccys that are in RGB format, i have no idea which one will work well with our requirements. It was hard enough to figure out our old problem of why transparency's were being turned into boxes which would slightly move and cause an edge mis-aligning, this was down to not using PDF 1.4 which has the ability to keep the transparency's in one piece and not freaking out.

Alot of the websites that try and explain these actions of what these PDFs do are quite complicated to understand, why did they set out so many different options or variants of PDFs?

I came from doing things by hand, learning as much as i could when i got told computers are taking over my job, alot of my knowledge has come from self taught methods due to managers not wanting to pay out the thousands of pounds it takes to get someone in to try and explain what and how these different programs and options actually do.

Some more learning is needed as we can never know enough, esp in the world of the digital age of print and all these new settings, effects and PDF settings i suppose lol.

Many thanks again.

Phil.

 

[Dov Isaacs]

Adobe Recommendations!

Adobe Recommendations!

Phil,

At Adobe, we most strongly recommend use of the PDF/X-4 joboptions for exporting PDF from InDesign or saving PDF from Illustrator or Photoshop when printing is required from such PDF.

Note that the words export and save were used here. For Adobe applications, creation of PDF by distillation of PostScript from a print operation (or save as PostScript) is strongly discouraged. Such PostScript is optimized only for direct printing and not for PDF creation. Furthermore it does not maintain color management or live transparency.

PDF/X-4 maintains the live transparency and color management of your original document and artwork, providing for color management and proper blending of transparency at the RIP, where it belongs!

Having said that, our industry does have its share of Luddite print service providers who are stuck in the 20th century either by virtue of obsolete software, obsolete training, or both. You should always check with your print service provider first to determine whether in fact they support modern, reliable PDF print publishing workflows or whether they require preflattened, CMYK-only PDF/X-1a. If that is their requirement, you either have to go with that or find an alternative print service provider.

- Dov

PS: Although the most recent Ghent Workgroup standards cited in another response here are based on PDF/X-4, they don't allow for anything other than DeviceCMYK, grayscale, and spot colors. Until such time as those standards are augmented to support full ICC color management, they are not endorsed or recommended by Adobe.

 

[Lennyr]

One other thing. Know what type of press the job will be printed on. The default settings are for US Web. If the job is printing on a sheetfed press, and you leave the color profile set to web you will have jobs where the proofs are hard to match on press.

 

[Phish]

Phil,

At Adobe, we most strongly recommend use of the PDF/X-4 joboptions for exporting PDF from InDesign or saving PDF from Illustrator or Photoshop when printing is required from such PDF.

Note that the words export and save were used here. For Adobe applications, creation of PDF by distillation of PostScript from a print operation (or save as PostScript) is strongly discouraged. Such PostScript is optimized only for direct printing and not for PDF creation. Furthermore it does not maintain color management or live transparency.

PDF/X-4 maintains the live transparency and color management of your original document and artwork, providing for color management and proper blending of transparency at the RIP, where it belongs!

Having said that, our industry does have its share of Luddite print service providers who are stuck in the 20th century either by virtue of obsolete software, obsolete training, or both. You should always check with your print service provider first to determine whether in fact they support modern, reliable PDF print publishing workflows or whether they require preflattened, CMYK-only PDF/X-1a. If that is their requirement, you either have to go with that or find an alternative print service provider.

- Dov

PS: Although the most recent Ghent Workgroup standards cited in another response here are based on PDF/X-4, they don't allow for anything other than DeviceCMYK, grayscale, and spot colors. Until such time as those standards are augmented to support full ICC color management, they are not endorsed or recommended by Adobe.

Many thanks for your advice Dov,


So your saying that when i go to print our documents from Illustrator, we should export them via the PDF/x-4 option rather than than print it to a postscript separated file and then distill it under whatever settings we want? I have been told that printing our files to postscript first is the safest option and now, from how i read your article, your saying this advice is wrong? We have had trouble with transparencies and from what you have said makes sense and could solve a problem many people seem to suffer from.

Many thanks again

Phil.

 

[Dov Isaacs]

“Safety” is a Relative Term!

“Safety” is a Relative Term!

The real question is who told you “that printing our files to postscript first is the safest option” and why they told you that. If the reason is that they have really antiquated RIP and PDF workflow software and/or don't know how to use whatever they have, that could be right. On the other hand, you should be strongly suspicious of such advice. With modern RIP and PDF workflow, PDF via distillation of PostScript is highly counterproductive, often yielding quite inferior results. The PDF/X-4 standard, the applications that support it, as well as modern RIP and PDF workflow software that support the PDF/X-4 standard have a proven reason for their existence, the resolution of a myriad of problems associated with transparency that plagued the industry until these workflows became available!

 

[Dov Isaacs]

Absolutely agree. The output color space is exceptionally important here. That having been said, many current generation RIPs will properly convert from one to another using the output color space as the basis of converting to the proper color space during the RIP process.

 

[fire11]

Dov,

We've needed to modernize our recommendations above X-1a for some but have not done so because of the myriad different types of files we see. For all the print service providers reading this forum, can you recommend a test file so we can evaluate how our "modern RIPs" will perform? Since reading your post this morning, I created a simple test file in InDesign with some live transparencies, spot colors and ICC profiles and then exported the file both as an X-1a and again as an X-4 file, processed both files in our modern PDF Print Engine-equipped RIP and can clearly see a difference. Before making a change in our PDF creation recommendations, I'd like to do some more thorough testing. Can you make some recommendations for us or point us to a good test file?

Thanks for your expert advice!

 

[Phish]

I was told by a valuable source, let me explain our procedure of how we print in our studio.

After creating a document in Illustrator, checking everything is correct visually as we have never been shown any of these preflight stuff, we will print a pre separated PDF of all the colours needed in our jobs, to a postscript file that we then distill via distiller under settings that have been set to our Signtronic screen machine spec.

We then check our PDF, once distilled, to make sure everything is ok then send it to a hotfolder that is then Ripped with RTI RipKit, which then creates 1bit Tiffs to be printed via our Signtronic Screen maker, basically a CTP screen maker.

If we were to use PDF/x-4 export option, this would create a composite PDF which may not be compatible with our RIP due to the age of the RIP and how recent the PDF/x-4 is?

We do not use a workflow, everything we do is manual, even trapping, however i am looking at this also which i am discussing in a separate post on another thread.

Phil

 

[abc]

@fire11
here's a set of test files you can use to test and validate your workflow and output rips - and they're free
Newly-Released Testing File Suite from the GWG Insures Up-to-Date Functionality of PDF Workflows - Ghent Workgroup

 

[Dov Isaacs]

Test Files!

Test Files!

Phil,

I wish that it was that simple to recommend one or simply a few universally-accepted test files to give you confidence in this workflow.

There are some companies and organizations that provide what they claim to be comprehensive test files for PDF/X-4 compliance. However, they often require expert analysis to determine what the output really means and in one particular case, the Altona 2.0 test suite, the file presents some “patches” that will give different results with different implementations due to a dispute as to interpretation of the PDF transparency specification. The two pieces of good news are the dispute is currently being resolved by the ISO 32000 PDF and ISO TC 130 PDF/X committees and that the situation in dispute is a boundary condition that you and/or your customers would be very unlikely to get into with normal use of any Adobe or supporting applications in a PDF/X-4 workflow.

That having been said, my advice, for what it is worth, would be to satisfy yourself in terms of this workflow by getting the source files for the gnarliest jobs you have run (with typical transparency and color problems), regenerate PDF/X-4 files for same, and try printing them. The proof in in the tasting!

- Dov

 

[Dov Isaacs]

Ghent Workgroup Standards & Tests of Limited Value!

Ghent Workgroup Standards & Tests of Limited Value!

The Ghent Workgroup test files are a good starting point, but have a very important and major limitation.

These test files are in support of the latest Ghent Workgroup Standards, which although are based on PDF/X-4, strictly prohibit use of any ICC color management. In other words, you are forced to convert all color managed objects such as those in Lab or RGB color spaces to CMYK when creating this subset of PDF/X-4. In other words, the current GWG standards are akin to PDF/X-1a but with transparency and a few other non-color-related features allowed.

It is for that reason that although Adobe is a member of the Ghent Workgroup (I am actually the representative from Adobe to the GWG), we officially recommend against use of the current GWG standards as opposed to full, color-managed PDF/X-4. When (or if) the Ghent Workgroup gets over its irrational fear of anything other than CMYK and spot colors (as handed down by Moses at Mt. Sinai), Adobe will reconsider endorsing such standards.

That having been said, download and try those test files, but since they don't test any ICC color management, they won't give you the full picture, so to speak!

- Dov

 

[Tommyjt]

I agree with Dov almost 100%. The only two things that I would change for my shop are:

1) I would not downsample images at all. PDF export downsampling is inferior to Photoshop. If I wanted to keep the file size down I would do so in Photoshop and place the downsampled image.

2) I would use ZIP compression instead of JPEG. Most images will be probably be okay but I have seen some nasty JPEG artifacts in my day.

I know these two items will significantly increase the file size of the final PDF, but I did say for "my shop". Quality is the most important factor here.

Tom

 

[Dov Isaacs]

Yes, But ...

Yes, But ...

Tom,

Be aware that image downsampling in PDF export (InDesign) and PDF save (Illustrator and Photoshop) uses the exact same code that Photoshop uses assuming you use the default “bicubic sampling” option in the .joboptions. We also use the same algorithm and code in the Adobe PDF Print Engine. Thus, you don't gain anything by doing downsampling in Photoshop prior to image placement in Illustrator or InDesign that you won't get in the PDF save/export from Illustrator or InDesign.

That having been said, there are at two advantages, at least for printing, if you don't downsample at all when creating PDF:

The first is that cascading operations, i.e. downsampling either in Photoshop or in PDF creation and then during the RIP process (contrary to some popular beliefs, no matter what resolution your imagery has, it is indeed resampled during the RIP's rendering process in order to provide for separations and screening) is inherently more lossy than downsampling/resampling only once during the RIP process. Such lossiness may be not be perceptible, but it does happen even if not harmful.

The second advantage is that of flexibility for re-purposing. If you don't downsample the raster images and you wish to print on a larger size substrate, such as for a poster - even going from letter size to ledger not to mention blow-ups to trade-show size posters - maintaining all the raster data will definitely result in better output than having to let the RIP artificially upsample the imagery. And it is certainly much less work than having to go back to the original document and regenerate the PDF for a higher resolution (i.e., not downsampling at all or downsampling to a higher resolution).

It is also true that JPEG compression can result in some pretty nasty image artifacts, especially when used on content that would have better been represented as vector and/or text content. However, the default compression modes in the PDF/X-4 .joboptions call for Automatic (JPEG) compression at Maximum quality. Unfortunately, few users know what that really means. When specified, Automatic (JPEG) compression results in selection of compression format for each individual image based upon examination of each image in terms of its attributes. For images that appear to be content better represented as vector and/or text including most computer screen shots, ZIP compression is used for such images. Other images which appear to be photographic images for which JPEG compression was really designed for are compressed as JPEG and in the case of the default PDF/X-4 .joboption, are compressed with maximum quality, lowest compression settings yielding results that have more compression than the .ZIP compression of the same file, but with pretty much imperceptible lossiness. We certainly would never recommend cascading cycles of saving, editing, resaving, etc. with the JPEG format, no matter what the content or JPEG compression settings, but as the last step in a PDF creation workflow, the default settings of Automatic (JPEG) at Maximum quality are a very safe bet. Yes, in theory, always using ZIP compression would yield absolutely lossless imagery, but in practice, since the availability of PDF creation provided the Automatic (JPEG), we have never seen a real problem in printing with its use including use in print facilities with the highest quality standards.

- Dov

 

[Dov Isaacs]

Phil,

I am obviously not familiar with your particular RIP and what it accepts or not. Nor will I pass judgement in terms of whether it is economical based on your time and volume of work to even consider its replacement. However, that having been said, what you describe is a hell of a lot more work in terms of steps and things that can go wrong than is necessary using more modern composite PDF workflow where the composite PDF file, typically a PDF/X-4 file, would be fed to the RIP which would automatically handle trapping, imposition, and separations for you as well as sending and monitoring the plate-making process with CtP technology.

- Dov

 

[toronar]

Tom,

Be aware that image downsampling in PDF export (InDesign) and PDF save (Illustrator and Photoshop) uses the exact same code that Photoshop uses assuming you use the default “bicubic sampling” option in the .joboptions. We also use the same algorithm and code in the Adobe PDF Print Engine. Thus, you don't gain anything by doing downsampling in Photoshop prior to image placement in Illustrator or InDesign that you won't get in the PDF save/export from Illustrator or InDesign.

I am sorry Dov, but I (and reality) begs to differ:
Photoshop does downsampling in a different way than InDesign and Distiller.
See Quality issues with Distiller Downsampling for examples and an explaination.

This was reported to Adobe as a bug, and the official response was:

First of all, Distiller never switches from bicubic to subsampling.
The code simply cannot do that. However, I think there is a need
for clarification about how Distiller performs bicubic downsampling.
For bicubic downsampling Distiller can only do 2:1 downsampling.
That's how the code is designed. Therefore, if you try to downsample
from for example 400dpi to 150dpi, Distiller first downsamples the
image with "Average" downsampling to two times the target resulution
which is 300dpi. Then it applies bicubic downsampling from 300 to
150dpi.
This also means that if you try to downsample with bicubic from
400 to 300 dpi, then Distiller won't do bicubic, but it switches
to Average downsampling to go from 400dpi to 300dpi in one step.

Photoshop on the other hand does it differently with much better results.

As others have said already, I strongly advise to downsample only for factors of 2:1, use Photoshop to downsample as it provides better results, or simply do not downsample at all.

 

[Stephen Marsh]

If an image is being resampled down in resolution, there is a very strong chance that it could benefit from sharpening. Sharpening to account for the softness introduced in the resampling process and most likely to account for the reproduction process as well. A bicubic sharper resize that sharpens as part of the resize can help with the resample, however it is not enough to account for the print process (in some cases this can also make things worse, see below).

If an image has a repeating pattern in it, it can sometimes result in aliasing issues or "interpolation moiré" when it is reduced in resolution. This is another case where careful application of image reduction in Photoshop evaluated by a human can be beneficial (this topic on aliasing interpolation artefacts deserves a separate topic thread to do it justice).


Stephen Marsh