MikelKlink
Advocate
MikelKlink
Advocate
Activity
Mar 26, 2025
This might be a hybrid AcroForm/XFA form, i.e. a PDF with two form definitions in different formats. Adobe software usually processes the XFA form entries while most other PDF processors process the AcroForm form entries.
... View more
‎Mar 25, 2025
11:04 PM
This sounds very much like an issue identified in Acrobat for Android: https://community.adobe.com/t5/acrobat-discussions/split-text-on-pdf-forms-from-android-app/m-p/15224490
... View more
‎Mar 25, 2025
03:43 PM
Thanks for sharing those additional files. They confirmed what I had assumed. The Authentisign signature form field flattening code is buggy and creates the broken content. So Docusign said it wasn't their problem. That's right. Their output is ok. Authentisign tried to help and said to flatten the document by printing in the chrome browser to pdf. Well, as their code damages the PDF, real help would be fixing their code for flattening signature widgets. Of course you can work around their bug by completely removing or flattening the Docusign signature before forwarding your documents to the next signer. But the next signer may also be bewildered by the missing digital signature... So that manual flattening approach merely would be a work-around while waiting for the Authentisign fix.
... View more
‎Mar 25, 2025
10:11 AM
I'm afraid you only shared screenshots. To analyze the issue the actual PDFs are needed.
... View more
‎Mar 25, 2025
09:28 AM
The error that makes Acrobat stop processing the page content streams and, therefore, not display the Authentisign additions, is quickly found: There is a cm (concatenate to transformation matrix) instruction with 4 of its 6 numeric parameters replaced by a "NaN" expression. This is invalid in PDFs. Furthermore, the effect of the cm instruction is to transform (scale/rotate/skew/translate) everything drawn thereafter in a way specified by those numbers. Thus, a responsible PDF viewer should stop here, inform the user about a non-recoverable error in the content stream, and not display the page at all, in particular in a signed document. Apparently, though, none of the viewers tested here handles this responsibly but instead each of them does display the page one way or another. The reason why the non-Acrobat viewers' apparent strategy of simply ignoring the issue works well here, is that after the transformation matrix change until its revocation by a restore-graphics-state instruction in a complicated way nothing is drawn. (I'm sure, though, that those viewers didn't check whether this is the case...) You only shared the final file of your example signing round trip, so I cannot be sure which software caused this. I assume, though, that Authentisign is the culprit. I assume so because this faulty cm instruction is between the additions by Docusign and those by Authentisign, and the content that is manipulated by it looks like a flattened widget annotation with a width and height of 0. This quite likely was the widget of the invisible (0x0) digital signature Docusign eventually applied. Authentisign needed to remove that digital signature, otherwise its changes to the document would have been disallowed altogether. To keep the visual appearance of digital signatures on the page after removal, Authentisign flattened the signature widget (i.e. tranformed it from an annotation into additional page content). Unfortunately, while doing so it didn't properly check the dimensions of the widget and blissfully divided by 0, getting NaN (Not-a-Number) values for the flattening transformation matrix. You should report this issue to Authentisign. As mentioned, though, this is only a (very reasonable!) assumption. To be 100% sure I'd need the intermediate file to compare with. ------ Some technical details: The page object in question is object 6. It has a content stream array with many partial streams. Among them is stream object 25 with this content: q
1.00 .00 .00 1.00 .00 816.00 cm
q
q
NaN NaN NaN NaN .00 .00 cm
1.00 .00 .00 1.00 .00 -816.00 cm
/e8b64a13-edb0-40e6-8242-3860f5dde0cc Do
Q
Q
Q The refered-to XObject e8b64a13-edb0-40e6-8242-3860f5dde0cc is in object 76 and has a 0x0 bounding box: 76 0 obj
<<
/Type /XObject
/Subtype /Form
/FormType 1
/Resources <<
/XObject <<
/n2 132 0 R
>>
>>
/BBox [0 0 0 0]
/Matrix [1.00 0 0 1.00 0 0]
/Length 9
>>
stream
™žÔ P„SFZ
endstream
endobj
... View more
‎Mar 25, 2025
07:15 AM
Do you by chance still have the original document and the intermediate one (with only the Docusign signature)? Analyzing "Adobe Test both signatures.pdf" one can find an error in the page contents, but without the other documents I cannot tell in which processing step they had been added.
... View more
‎Mar 25, 2025
06:26 AM
I'm afraid, I cannot help here. I've only used Telesec cards in combination with custom software with integrated support for them. In context with Acrobat I assume that you do need a PKCS#11 module but I don't have any hands-on experience to share.
... View more
‎Mar 25, 2025
02:52 AM
I'm afraid that indeed means that you're most likely subject to either some infection or some defect.
... View more
‎Mar 24, 2025
11:38 PM
The problem seems to be with Adobe. That's unlikely. Whenever Acrobat displays an error message that there is a problem on some page, this usually means that there is an error in the PDF. The reason why some PDF viewers do display the page as you expect it, is that PDF viewers generally try to fix some errors under the hood without reporting this, and that different viewers fix different sets of errors. Apparently the error in the file in question is fixed by the docusign viewer but not by Acrobat. Considering, though, that this "fixing" always includes some guessing how the broken PDF actually was meant to look like, different fixing viewers may fix the issue differently. This obviously is critical in particular in case of contracts which should look the same for all parties involved. That all being said, though, I'm afraid that without analyzing the PDF in question this is pure guesswork. Thus, can you share one such PDF? Maybe you can create some PDF with dummy data and try to recreate the issue together with that other realtor?
... View more
‎Mar 24, 2025
10:03 AM
There are some errors in your file. In particular the startxref value is incorrect and the xref keyword is not on a line by itself as it should be. (There may be other errors, too, but these leapt to the eye.) You should fix the PDF and try again. PS: The value entry in the ref field object is completely broken: /V <FEFF00540044002D0032002D0041002D003300310030002D00300033>ref])
... View more
‎Mar 24, 2025
06:51 AM
The file you shared does not seem to be a PDF file. The drive that file was on might have had a crash, or some program overwrote the original PDF contents with something else. Have you probably experimented with some file encryption tool? Or have you been subject to an unfinished ransomware attack that encrypted some of your files but then got stopped?
... View more
‎Mar 22, 2025
03:21 AM
This appears to be a bug Adobe has to fix in their product. If they cannot fix it quickly, you may consider using different PDF form filling software for the time being.
... View more
‎Mar 21, 2025
12:28 AM
Can you share an example file to analyze? If the file itself is identical to the one on which you originally have seen signed, the signature must still be there, but it's possible current PDF viewers don't show them. In particular the current Acrobat on Mac seems to be especially buggy.
... View more
‎Mar 13, 2025
06:21 AM
PDF/A requires that you have an OutputIntent if you use an uncalibrated color space. Thus, you will need to look out for an ICC profile (to put into your DestOutputProfile) with the characteristics you want. Unfortunately I'm not really into that subject and cannot recommend anything.
... View more
‎Mar 13, 2025
04:50 AM
1 Upvote
Indeed, just like @Tariq Ahmad Dar said, this is pure guesswork without inspecting the file itself. In addition to the options he gave there obviously also are the options of an error in the PDF or of a bug in one program or the other.
... View more
‎Mar 13, 2025
04:43 AM
Ah, ok, so I originally misinterpreted the meaning of your screen shots, and the second idea I added as an afterthough to my answer above should explain the cause.
... View more
‎Mar 13, 2025
03:44 AM
I could not reproduce your issue, here the rectangle is drawn in yellow color by Acrobat whether I zoom in or not. One thing is interesting, though: In your screenshot "Rectangle Acrobate Reader.JPG" the rectangle appears yellow in Acrobat while in your screenshot "Same Green.JPG" the zoomed-in rectangle appears green in Acrobat: What did you do differently there? Does the line become green while zooming? Or are the files "Rectangle.pdf" in the upper part and "Rectangle Uncompressed.pdf" in the lower part considerably different? Or when you said When I draw a rectangle in yellow, RGB FF FF 00, it appears green, but ONLY in Acrobate reader. did you probably not mean actually green but merely yellow with a slight greenish tint? That's due to the OutputIntent of your PDF (see the OutputIntents entry of the document Catalog). The information in such an OutputIntent can cause the colors to be transformed to match a certain profile. I removed the OutputIntents entry in your PDF and compared the Acrobat display with and without that entry side-by-side: In the upper part the "1.000 1.000 0.000 RG" due to the OutputIntent results in e0ff00 while in the lower part without OutputIntent you get ffff00.
... View more
‎Mar 13, 2025
02:01 AM
I remember having once seen issues in this context if an outline entry pointed to a high-level structure element (I don't remember if it was in a Dest or in a SE entry). When visually signing a tagged PDF, iText adds (or at least used to add) structure elements for the signature widget appearance, and if those added structure elements were underneath a structure element pointed to by an outline entry, Acrobat considered this a disallowed change.
... View more
‎Mar 12, 2025
02:32 PM
Hhmmm, PDF/A ist grundsätzlich eine Untermenge von PDF. Darum sollte ein PDF, dass aus Seiten einer normalen PDF-Datei und einer PDF/A-Datei besteht, als ganz normales PDF mit allen Seiten erkannt und verarbeitet werden. Ich würde deshalb annehmen, dass beim Zusammenführen der Seiten aus den unterschiedlichen Quellen noch etwas anderes schiefgelaufen ist.
... View more
‎Mar 12, 2025
03:34 AM
1 Upvote
Vermutlich ist die PDF-Datei beschädigt, und verschiedene PDF-Viewer reparieren sie in unterschiedlicher Weise mit unterschiedlichen Ergebnissen. Möglich wäre aber auch, dass es sich um ein Hybrid-Formular handelt, bei dem die (von Adobe Acrobat angezeigte) XFA-Ausprägung ungewöhnlicherweise kürzer als die (von den meisten anderen PDF-Viewern angezeigte) AcroForm-Ausprägung ist. Natürlich ist auch ein Bug bei der einen oder anderen Software nicht auszuschließen. Mit anderen Worten, ohne die Datei selber zur Analyse können wir hier nur raten.
... View more
‎Mar 10, 2025
12:59 PM
This is not uncommon. Most PDF viewers accept (and on the fly fix) a set of errors in PDFs they open. But it's not the same set for different viewers. So there are some broken PDFs one PDF viewer can still make sense of while there are some others another viewer fixes somehow.
... View more
‎Mar 10, 2025
04:35 AM
1 Upvote
Ok, as mentioned I could reproduce the issue on Android, not only with your PDF but with others, too. Also I couldn't find the error in your file that I knew to cause similar issues. So apparently this is an error in the current Acrobat for Android. The issue appears to be that Acrobat, while editing a field, displays both the current (edited) version of the field contents and the former (unchanged) version thereof. This is especially visible when the original contents had been filled in with some other software that formatted the text slightly differently, or when one edits at the start of the former contents.
... View more
‎Mar 10, 2025
01:39 AM
I can reproduce your problem on Android. The behavior reminds me of the behavior of some forms I saw earlier that had multiple widgets of the same field at the same position; when one clicked on them, one of the widgets got focus and was in editing mode while the other(s) still showed the original content in the same place. I'm going to inspect the internals of your PDF later to check whether I can identify the cause of the issue.
... View more
‎Mar 09, 2025
06:18 AM
Can you share the PDF form in question?
... View more
‎Mar 06, 2025
10:21 PM
I'm afraid without the file itself to analyze this is pure guesswork.
... View more
‎Mar 06, 2025
02:20 PM
I'm not aware of a way to do so with Acrobat alone. Also, as mentioned this is but one possible reason. Without the file itself, it is hard to tell.
... View more
‎Mar 06, 2025
02:09 PM
1 Upvote
The attachments are arranged in the EmbeddedFiles name tree structure. The names in such a name tree are expected to be sorted in lexical order, in your case 1.jpg, 10.jpg, 2.jpg. Actually, though, they are sorted in numerical order 1.jpg, 2.jpg, 10.jpg. This is wrong and PDF processors may run into issues because of this. My pdf is created with the pdf-lib framework. You may want to report a bug with that framework and ask them to create correctly ordered name trees.
... View more
‎Mar 06, 2025
01:37 PM
That's hard to tell. It could be hybrid form PDFs (i.e. PDFs with both an AcroForm form definition and a XFA form defintion) with the fill-ins only applied to the AcroForm part. In this case desktop Acrobat displays the XFA information while most other viewers will display the AcroForm information. But that's just one possibility...
... View more
‎Mar 02, 2025
11:25 PM
The permissions feature of PDF is a soft security feature, it requires PDF processors to cooperate and abstain from certain operations if certain flags in a document are not set. Apparently the tool you reference does not respect those flags. Even worse, it removes them. That permissions feature is coupled with the encryption feature, but as soon as a PDF processor can decrypt an encrypted document to display it, it technically can do anything with it. Adobe Acrobat does warn you about this when you apply permissions unless you have in a prior such warning clicked the "Don't show this warning again" check box.
... View more
‎Feb 23, 2025
10:26 PM
Are you using the service Adobe Acrobat Sign for signing or are you merely creating the PDF with Adobe Acrobat and then want to handle the sending yourself? The answer of @try67 refers to the latter option.
... View more