Aligning PDF/A and PDF/X LWG

Dear Aligning PDF/a and PDF/X Liaison Working Group,

PDF Week Fall 2023 is quickly approaching! We don’t want you to miss this opportunity. 

Not only is your working group meeting but at PDF Week Fall 2023, 14 PDF Association working groups will meet in-person alongside 5 of ISO TC 171 SC 2’s working groups. In addition, Tuesday, October 17, is the “Symposium on Advancing the PDF Imaging Model” where you can help shape the future of PDF’s imaging model.  

PDF Week is a great way to advance your interests, learn from others, network with colleagues and participate in ISO standards development. 

Be sure to register today and make your hotel reservation at the Omni San Francisco where all the meetings will be held. We have block of rooms reserved at a discounted rate; this link will be distributed to those who register for PDF Week.

Review the latest agenda and register today! 

Naturally, remote access will be provided to members who are unable to attend in-person.

We look forward to seeing you in San Francisco!

Betsy Fanning

Betsy A Fanning <betsy.fanning@pdfa.org> (July 25, 2023 22:42):

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=8501&h=56ffa78edd307765aad936dff377aed5ce884fee&sa=1809379872

This list is a service of the PDF Association. For questions contact info@pdfa.org.

External Sender: Use caution with links/attachments.

Dietrich von Seggern <d.seggern@callassoftware.com> (June 28, 2023 17:04):

Dietrich von Seggern | Managing Director

Find us:   callas software GmbH | Schönhauser Allee 6/7 | 10119 Berlin | Germany   Contact us:   Tel +49.30.44390310 | www.callassoftware.com   Management:   Dietrich von Seggern, Ulrich Frotscher, Olaf Drümmer   Registry:   Amtsgericht Charlottenburg, HRB 59615

On 28. Jun 2023, at 16:33, Mike Bremford <mike@bfo.co.uk> wrote:

As discussed on today's call, attached are a couple of PDF files:

streameol-good.pdf - this has a zero-length stream which looks like "stream\r\n\nEOF" - this is good (it has the EOL marker)

streameol-bad.pdf - this has a zero-length stream which looks like "stream\r\nEOF" - this is bad (it has no EOL marker)

We fail the bad one, so does veraPDF, so this was just my antenna twitching rather than anything to be actually worried about.

Cheers... Mike

--

-----------------------------------------------------
Mike Bremford - CTO                   mike@bfo.com
Big Faceless Organization             http://bfo.com

On Mon, 26 Jun 2023 at 14:20, Leonard Rosenthol <lrosenth@adobe.com> wrote:

Leonard Rosenthol <lrosenth@adobe.com> (June 26, 2023 15:19):

Added as https://github.com/ISOTC171SC2/19005-4/issues/9 for tracking.

 

Thanks!

 

Leonard

 

From: aligning-pdfa-pdfx-lwg@pdfa.org <aligning-pdfa-pdfx-lwg@pdfa.org> on behalf of Patrick Gallot <pgallot@datalogics.com>
Date: Wednesday, June 21, 2023 at 12:02 PM
To: Aligning PDF/A and PDF/X LWG <aligning-pdfa-pdfx-lwg@pdfa.org>
Subject: [aligning-pdfa-pdfx-lwg]PDF/A PDF/X Document Requirements

EXTERNAL: Use caution when clicking on links or opening attachments.

 

Patrick Gallot <pgallot@datalogics.com> (June 21, 2023 17:57):

Both PDF/A and PDF/X are aligned in requiring that the ‘requirements’ key not be present in conforming documents’ document catalog dictionary.

 

However, I am wondering if this is the correct approach with the gltf and step in PDF standards that are coming.

 

Would it make more sense to keep any document requirements and instead direct conforming processors to ignore or conform to particular requirements as appropriate ?

 

-Patrick

 

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7684&h=d6d3874594f7e9005990053c656b4b05123abd97&sa=1328609717

This list is a service of the PDF Association. For questions contact info@pdfa.org.

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7713&h=f616ecb038ff422ee0aeb43cc2110beb9fb8d9c2&sa=215228303

This list is a service of the PDF Association. For questions contact info@pdfa.org.

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7777&h=134c900f92e5f7ae84281369552087af079f2aed&sa=165368105

This list is a service of the PDF Association. For questions contact info@pdfa.org.

External Sender: Use caution with links/attachments.

Leonard Rosenthol <lrosenth@adobe.com> (June 7, 2023 19:21):

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7548&h=57ac1a2f757f37321312e791bf2515e72b3bbc93&sa=695763837

This list is a service of the PDF Association. For questions contact info@pdfa.org.

On 8. Jun 2023, at 08:04, Peter Wyatt <peter.wyatt@pdfa.org> wrote:

Peter Wyatt <peter.wyatt@pdfa.org> (June 8, 2023 8:02):

That seems very reasonable to me.

 

Can I suggest that we also add some light-weight informative note in PDF/A stating that PDF/X differs in regards to the “shall not be used to render content”? That way both sides of the equation are highlighted to an important difference.

 

 

From: aligning-pdfa-pdfx-lwg@pdfa.org <aligning-pdfa-pdfx-lwg@pdfa.org> On Behalf Of Leonard Rosenthol
Sent: Thursday, June 8, 2023 3:26 AM
To: Aligning PDF/A and PDF/X LWG <aligning-pdfa-pdfx-lwg@pdfa.org>
Subject: Re: [aligning-pdfa-pdfx-lwg]15930-9 revision question about "Any Data [...] not described [...] shall not be used to render content on a page" statement

 

Leonard Rosenthol <lrosenth@adobe.com> (June 7, 2023 19:21):

So PDF/X already tries to address this with the following text in the “Conforming Processor” section – including three examples (one of which specifically calls out 21812).

 

The rendering and other processing of conforming files shall be performed as defined in <<ISO32000-2>> subject to the additional restrictions specified by this document.

 

To the extent that <<ISO32000-2>> and this document permit more than one rendering of a conforming file, a conforming processor may use embedded job ticket or metadata information to control the rendering of the file more precisely.

 

EXAMPLE 1 (Trapping) If a PDF/X-6 conforming file specifies **pdf:Trapped**=__False__, a conforming processor might use job ticket information to determine details of how the file is to be trapped. If the file specifies**pdf:Trapped**=__True__, a conforming processor might ignore any trapping information in an embedded job ticket.

 

EXAMPLE 2 (Screening) A PDF/X-6 conforming processor can use embedded job ticket information to determine the screening to be used to render the file. A PDF/X-6 conforming processor can ignore screening information in the PDF/X-6 file. A conforming processor might use screening data from the PDF/X-6 file, from the job ticket, or from local system defaults.

 

EXAMPLE 3 (Print product metadata) A PDF/X-6 conforming processor can look for the presence of information that complies with ISO 21812-1^[7]^ to determine if there is any print product metadata that would influence the production of the file. A PDF/X-6 conforming can ignore any additional information in the PDF/X-6 file, not mentioned in this document.

 

 

For PDF/A, we have the same starting “shall” statement, but then go on to you should & may on other aspects.

 

The rendering and other processing of conforming files shall be performed as defined in <<ISO32000-2>> subject to the additional restrictions specified by this document. Features described in specifications that are not explicitly described in <<ISO32000-2>> should be ignored by conforming processors. Features described as deprecated in <<ISO32000-2>> may be ignored by conforming processors.

 

 

However, in PDF/A we also have a file format requirement (which is what Patrick started this with):

 

Any data contained in a conforming file that is not described in <<ISO32000-2>> or this document should be ignored by a conforming processor and shall not be used to render content on a page.

 

 

So the first part of that sentence is a duplication of the conforming processor statement and is a should.  It’s the second part the “shall not be used to render content on a page” where are indeed differing since (as noted above in the PDF/X processor section) PDF/X *does* allow such data to be used for rendering.

 

Recommendation: I’d like to get rid of the sentence from the file format requirements and leave only statements about this in conforming processors – that is where it belongs.   And I’d like to then leave the “shall not be used to render content on a page” bit in PDF/A *but* leave PDF/X alone as it’s statement about alternate renderings is just fine for it.

 

Thoughts?

 

Leonard

 

From: aligning-pdfa-pdfx-lwg@pdfa.org <aligning-pdfa-pdfx-lwg@pdfa.org> on behalf of Martin Bailey <Martin.Bailey@hybridsoftware.group>
Date: Wednesday, June 7, 2023 at 12:50 PM
To: Aligning PDF/A and PDF/X LWG <aligning-pdfa-pdfx-lwg@pdfa.org>
Subject: Re: [aligning-pdfa-pdfx-lwg]15930-9 revision question about "Any Data [...] not described [...] shall not be used to render content on a page" statement

EXTERNAL: Use caution when clicking on links or opening attachments.

 

Martin Bailey <Martin.Bailey@hybridsoftware.group> (June 7, 2023 18:49):

 

Leonard put his finger on something really important here. There are cases where the strictly conforming rendering of a PDF/X file is what is required. And there are situations where something different is required, but where the fact that the file conforms to PDF/X still provides significant value.

 

This includes things like technical spots, automatically adjusting optional content (e.g. with Processing Steps) or applying colour management for a virtual output device (or anything else that doesn't match the output intent.

 

Acknowledging that in the PDF/X standard, perhaps with an equivalent for PDF/A noting that non-conforming renderings have less widespread value in that context, would be a useful approach, I think.

 

Reading back through the thread this is pretty much what Patrick suggested.

 

Martin

---

From: aligning-pdfa-pdfx-lwg@pdfa.org <aligning-pdfa-pdfx-lwg@pdfa.org> on behalf of Dov Isaacs <dov@dovisaacs.com>
Sent: 07 June 2023 16:38
To: Aligning PDF/A and PDF/X LWG <aligning-pdfa-pdfx-lwg@pdfa.org>
Subject: Re: [aligning-pdfa-pdfx-lwg]15930-9 revision question about "Any Data [...] not described [...] shall not be used to render content on a page" statement

 

Dov Isaacs <dov@dovisaacs.com> (June 7, 2023 17:37):

 

I don’t think that any of this discussion, at least so far, has done anything to be problematic in this area, but we should keep in mind that we should not forget that we should allow for PDF files to be in conformance and tagged as both PDF/X-6 and PDF/A-4 within some reasonable constraints.

- Dov

 

From: aligning-pdfa-pdfx-lwg@pdfa.org <aligning-pdfa-pdfx-lwg@pdfa.org> On Behalf Of Leonard Rosenthol
Sent: Wednesday, June 7, 2023 8:28 AM
To: Aligning PDF/A and PDF/X LWG <aligning-pdfa-pdfx-lwg@pdfa.org>
Subject: Re: [aligning-pdfa-pdfx-lwg]15930-9 revision question about "Any Data [...] not described [...] shall not be used to render content on a page" statement

 

Leonard Rosenthol <lrosenth@adobe.com> (June 7, 2023 17:23):

 

> A lot of this probably goes to assumptions that soft-proofing is the same as a conforming PDF/X rendering.

 

This is a long standing debate in the community…It has to do with how the soft-proofing takes place.  

 

In theory, if the PDF/X file is rendered according to PDF and PDF/X rules *and then* color corrected for soft proofing – then the soft proof is out of scope and thus perfectly fine “as per spec”.   However, if color correction/conversion happens *during* the rendering process – then the output is clearly not compliant.

 

Leonard

 

From: aligning-pdfa-pdfx-lwg@pdfa.org <aligning-pdfa-pdfx-lwg@pdfa.org> on behalf of Peter Wyatt <peter.wyatt@pdfa.org>
Date: Tuesday, June 6, 2023 at 9:44 PM
To: Aligning PDF/A and PDF/X LWG <aligning-pdfa-pdfx-lwg@pdfa.org>
Subject: Re: [aligning-pdfa-pdfx-lwg]15930-9 revision question about "Any Data [...] not described [...] shall not be used to render content on a page" statement

EXTERNAL: Use caution when clicking on links or opening attachments.

 

Peter Wyatt <peter.wyatt@pdfa.org> (June 7, 2023 3:42):

 

In regards to ISO 19593-1:2018 (processing steps) and ISO 21812-1:2019 (PPM):

 

I’m not sure if anything in ISO 21812-1:2019 (PPM) actually changes the final rendered appearance of page graphics within a conforming PDF/X as it is all ignorable metadata about traditional post-press.  DPart/DPM, binding, glueing, folding, product intents, VDP recipient info, etc. do not impact the printed visual appearance in any way that is expressed with the PDF/X graphics. Even CIP4_ColorIntent coatings don’t change the meaning of the PDF/X graphics (my highlight in red): 

“The following restrictions shall be adhered to when mapping XJDF Intent Resources to PDF Metadata.

— CIP4_ColorIntent: Information about printing colour conditions is out of scope. This information shall be provided using the standard methods defined for PDF, e.g. output profiles.”

At most we might say that a soft proofing system may additionally visualize PPM information (e.g. overprint varnish, hole patterns) on top of the PDF/X reliable appearance soft proof.

 

However ISO 19593-1:2018 (processing steps) uses standard OCGs and these could impact PDF/X rendering appearance if not authored with appropriate visibility (e.g. white overprint over the top of everything, dimension lines not intended to be printed, etc). Thus ISO 19593-1 OCGs need to be correctly written to the PDF/X: “PDF/X-6 conforming file may also include other valid PDF features that do not affect final print reproduction of the compound entity” (quoting from ISO 15930-9). 

 

A lot of this probably goes to assumptions that soft-proofing is the same as a conforming PDF/X rendering.

 

 

From: aligning-pdfa-pdfx-lwg@pdfa.org <aligning-pdfa-pdfx-lwg@pdfa.org> On Behalf Of Patrick Gallot
Sent: Wednesday, June 7, 2023 3:39 AM
To: Aligning PDF/A and PDF/X LWG <aligning-pdfa-pdfx-lwg@pdfa.org>
Subject: Re: [aligning-pdfa-pdfx-lwg]15930-9 revision question about "Any Data [...] not described [...] shall not be used to render content on a page" statement

 

Patrick Gallot <pgallot@datalogics.com> (June 6, 2023 19:38):

 

Throwing this out there for thought (using deliberately non-normative language for now):

 

PDF/X-6 conforming files may contain additional data conforming to PDF/X-6 derived standards such as to ISO 21812 and ISO 19593 that could be used to render PDF/X document differently without being considered non-conformant, but the _canonical_ rendering of a PDF/X-6 document may only use features described in the PDF 2.0 spec as allowed by the  PDF/X-6 spec. 

 

A separate follow-up question: 

Is there a particular reason why the ‘Any Data […]’ sentence is in the ‘File Structure’ subsection and not in the ‘Conforming Processor’ subsection of the in the Conformance section for both specs ?

 

-Patrick

 

From: aligning-pdfa-pdfx-lwg@pdfa.org <aligning-pdfa-pdfx-lwg@pdfa.org> On Behalf Of Leonard Rosenthol
Sent: Tuesday, June 6, 2023 11:07 AM
To: Aligning PDF/A and PDF/X LWG <aligning-pdfa-pdfx-lwg@pdfa.org>
Subject: Re: [aligning-pdfa-pdfx-lwg]15930-9 revision question about "Any Data [...] not described [...] shall not be used to render content on a page" statement

 

Leonard Rosenthol <lrosenth@adobe.com> (June 6, 2023 18:02):

 

This is a tough one with respect to aligning…

 

Both groups are focused on “reliable rendering” – that is where the “X” (in terms of reliable exchange) started, after all.   

 

HOWEVER, as Dietrich notes below the community that uses PDF/X (aka professional print) also wants to be able to “go their own way” when necessary for their specific workflows while the community for PDF/A (aka archivists) would never allow that (aka “there is only one rendering”).

 

And the more I think about that, the more it worries me with respect to the whole “alignment” situation overall….since a bit part of this is to have a single file that works for both.  BUT that can only happen if the PDF “rendering” is identical in both cases…

 

Leonard

 

From: aligning-pdfa-pdfx-lwg@pdfa.org <aligning-pdfa-pdfx-lwg@pdfa.org> on behalf of Dietrich von Seggern <d.seggern@callassoftware.com>
Date: Tuesday, June 6, 2023 at 10:03 AM
To: Aligning PDF/A and PDF/X LWG <aligning-pdfa-pdfx-lwg@pdfa.org>
Subject: Re: [aligning-pdfa-pdfx-lwg]15930-9 revision question about "Any Data [...] not described [...] shall not be used to render content on a page" statement

EXTERNAL: Use caution when clicking on links or opening attachments.

 

Dietrich von Seggern <d.seggern@callassoftware.com> (June 6, 2023 15:59):

Another example would be Processing Steps metadata as defined in ISO 19593. A viewer may decide to display or not display layered objects depending on this metadata. 

 

But I think these all are workflow dependent features and the general requirement to not use them for rendering may stand for a “pure” or strict rendering mode. In a note we may refer to ISO 21812 and ISO 19593 as examples for when other modes may be preferable. 

 

Dietrich von Seggern | Managing Director

 

Find us:   callas software GmbH | Schönhauser Allee 6/7 | 10119 Berlin | Germany   Contact us:   Tel +49.30.44390310 | www.callassoftware.com   Management:   Dietrich von Seggern, Ulrich Frotscher, Olaf Drümmer   Registry:   Amtsgericht Charlottenburg, HRB 59615

 

 

 

 

On 6. Jun 2023, at 15:17, Martin Bailey <martin.bailey@hybridsoftware.group> wrote:

 

Martin Bailey <Martin.Bailey@hybridsoftware.group> (June 6, 2023 15:15):

 

It was that job ticket paragraph that I was thinking of when I suggested that the blanket requirement to ignore extra data may be inappropriate.

 

Martin

---

From: aligning-pdfa-pdfx-lwg@pdfa.org <aligning-pdfa-pdfx-lwg@pdfa.org> on behalf of Patrick Gallot <pgallot@datalogics.com>
Sent: 06 June 2023 01:28
To: Aligning PDF/A and PDF/X LWG <aligning-pdfa-pdfx-lwg@pdfa.org>
Subject: [aligning-pdfa-pdfx-lwg]15930-9 revision question about "Any Data [...] not described [...] shall not be used to render content on a page" statement

 

External Sender: Use caution with links/attachments.

 

Patrick Gallot <pgallot@datalogics.com> (June 6, 2023 2:23):

 

Hello Everybody,

We discussed this a bit at the kick-off meeting, but with regards to the full statement from PDF/A-4:

“Any data contained in a conforming file that is not described in [ISO32000-2] or this document should be ignored by a conforming processor and shall not be used to render content on a page.”

 

Would the inclusion of this  statement in (06) technical requirements section of the PDF/X-6 specification preclude, say, the use of CIP4_LABColorValue or CIP4_MediaColor in a CIP4_MediaIntent  (Table 21 of ISO21812) dictionary from ever being used to render a PDF/X page ?

 

I’m also considering this statement against the last four paragraphs of the preceding conformance section:

 

“To the extent that [ISO32000-2] and this document permit more than one rendering of a conforming file, a conforming processor may use embedded job ticket or metadata information to control the rendering of the file more precisely.”

 

And the 3 examples below that use the term “might ignore” or “can ignore”, which implies that _not_ ignoring the “embedded job ticket or metadata information” is an equally valid choice that conforming PDF/X-6 processor might make, but are not required to.

 

Or am I going astray with my reasoning?

 

-Patrick

 

Editorial note: Last sentence of Example 3  in section 05 seems to be missing  the word processor after the word conforming.

 

 

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7525&h=8385ac67b1b0784b4aac22431a3492de8aa77354&sa=1773002622

This list is a service of the PDF Association. For questions contact info@pdfa.org.

 

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7526&h=e83674118cdd2d12aabfc63f3e80b52a432cd4a7&sa=362619515

This list is a service of the PDF Association. For questions contact info@pdfa.org.

 

 

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7532&h=80fbacf33a97fc7927451b4dbdf430a328b5af43&sa=407118007

This list is a service of the PDF Association. For questions contact info@pdfa.org.

 

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7533&h=e2081a06676cae3bfd3da0a4d713405d89ec2d66&sa=1003150655

This list is a service of the PDF Association. For questions contact info@pdfa.org.

 

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7534&h=1a802e2a77c7984a7a80a1a90d860b11c7c07825&sa=1563431931

This list is a service of the PDF Association. For questions contact info@pdfa.org.

 

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7544&h=d24a7541c32180e5ac040d06c86aa404d3ff7efd&sa=1571302255

This list is a service of the PDF Association. For questions contact info@pdfa.org.

 

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7545&h=e937d03672cfcb15c256a22d06bb85ce07b06161&sa=1501418531

This list is a service of the PDF Association. For questions contact info@pdfa.org.

 

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7546&h=ed3f44a96d9a1cd74e38385ee7304aa9db52474b&sa=723906650

This list is a service of the PDF Association. For questions contact info@pdfa.org.

 

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7547&h=de2405408edef3c4d43f72cbc653d5a2e91c1dff&sa=2056997971

This list is a service of the PDF Association. For questions contact info@pdfa.org.

 

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7548&h=57ac1a2f757f37321312e791bf2515e72b3bbc93&sa=695763837

This list is a service of the PDF Association. For questions contact info@pdfa.org.

---

You are subscribed to the Aligning PDF/A and PDF/X LWG listserv.

Login to the member area at pdfa.org to change subscription options, or leave the Community directly via this link:

https://pdfa.org?confirm_unsubscribe=indeed&m=7549&h=493bc760c26e1f089530b2e6c3ba29075371ec35&sa=127677387

This list is a service of the PDF Association. For questions contact info@pdfa.org.