Why was my PDF refused?
This early version of Verso works with a limited set of PDFs. When a file is outside that set, Verso refuses it rather than risk damaging it, and your original stays exactly as it was.
Your PDF is processed in this browser tab and never uploaded, so the checks that refused it ran in this tab too. You can check this yourself.
Each section below explains one reason: what it means, why this version refuses it, and what you can do. The refusal message links to the matching section. When the message has a "Technical details" part, it shows a reason code there, and each section lists its codes at the end.
Most refusals come down to one rule. Before Verso changes a part of a PDF, testing has to show that the part comes through the change intact. Until that is shown for a feature, refusing the file is safer than returning one in which something was quietly lost or broken.
Some sections suggest making a copy of the file with another app. That may give you a file Verso accepts, but it is not a promise: Verso has not tested files made by other apps, and a new copy can be refused for a different reason. This page does not promise that a refused feature will be supported later either; where a change is planned, the section says so.
Files over the size limit
Each PDF can be up to 32 MB (33,554,432 bytes), or 25 MB on iPhone and iPad. Merge combines up to 32 MB in total, and every result is limited to 32 MB as well. The size is checked before anything is read, and the file selector shows the figure for the device you are using.
Why this version refuses it. Your device does the work. Verso reads the whole file into this tab's memory and builds the result there, so the file's size sets how much memory the job needs. A phone has far less room than a laptop, and on iPhone and iPad a tab that runs out of memory can be closed without any warning, which is why the limit there is lower. A result can also be refused after processing, if it would be larger than 32 MB.
What you can do. Choose a smaller file. Splitting the file first does not help, because Split reads the whole file too and has the same limit. On iPhone or iPad, a file between 25 MB and 32 MB is within the size limit on a computer. For Merge, combine fewer or smaller files.
Reason codes: size-limit, batch-size-limit, output-size-limit
Features inside the file
Before it changes a file, Verso looks through all of it for the features below and refuses the file if it finds any of them. The message names the feature it found. A file can contain several, and the message names one, so a copy without that feature can still be refused for another.
Merge and Split build a new document from the pages; Rotate, Protect and Unlock keep the file's structure and save it again. Either way, testing has not yet shown that these features come through intact, and a feature that is lost or broken often leaves no visible sign. Where a feature has a reason of its own, its section gives it.
Digital signatures or signature fields
A digital signature records who signed a document and lets a reader detect any later change. A signature field is the place a signature goes, whether or not it has been signed. Certification signatures, document timestamps and the validation data some signing apps add count too.
Why this version refuses it. Changing the file, even rotating one page, can make the signature invalid, and rewriting the file can remove the signature altogether.
What you can do. Use a copy of the document from before it was signed, and sign the result afterwards if you need to.
Reason code: feature:signatures
XFA forms
XFA is a form format in which the form is described in XML stored inside the PDF. Viewers that support XFA may draw the pages from that description rather than from the ordinary page content, so different viewers can show different things.
Why this version refuses it. Verso has no support for XFA, so a change could lose the form or leave it out of step with the pages, and the difference may show only in viewers that draw from XFA.
What you can do. If you only need the pages as they look, an app that can print or export the form to a new PDF may give you a file Verso accepts. Verso has not tested files made that way.
Reason code: feature:xfa
Interactive form fields
Fields you can type into, tick or choose from: text boxes, check boxes, radio buttons, lists and buttons. A form's answers, and how they look on the page, are stored separately from the page itself.
Why this version refuses it. Fields are listed once for the whole document and placed on particular pages. When pages are copied into a new document, or the file is saved again, fields can be left behind or stop working, and testing has not yet shown that the fields and their answers come through intact.
What you can do. If the form is finished and you no longer need to edit it, an app that can flatten forms (turn the fields into ordinary page content) may give you a file Verso accepts. Verso has not tested files made that way. Keep the original if you may need to change the answers.
Reason code: feature:forms
Tags or marked content
Tags describe a document's structure, such as headings, paragraphs, lists, tables, reading order and text alternatives for images, so screen readers and other assistive technology can present it properly. Marked content is the part of the page that the tags point to.
Why this version refuses it. If the pages change and the tags don't change with them, the file becomes harder or impossible to use with a screen reader, and nothing on screen shows it. This version accepts tags in the forms it can check, a structure tree that matches the marked content on each page: Rotate, Protect and Unlock keep them, and Split and Merge leave them out and say so. It refuses tags in other forms, such as tags that point at annotations other than links, or at links in forms it can't check, marked content for layers or never closed, a tree that doesn't match its pages, and tag data outside the tree.
What you can do. Keep the original for anyone who relies on assistive technology. Removing the tags to get past this refusal would make the document less accessible, so this page doesn't suggest it. For a file whose tags this version refuses, this version of Verso is not the right tool.
Reason code: feature:tagged-content
Document or object metadata streams
A block of descriptive data stored inside the file, for the whole document or for a single image or page: for example the title, the author, dates and the app that made it. It can also declare that the file meets a standard, such as PDF/A for archiving or PDF/UA for accessibility. Many apps add it automatically when they save a PDF.
Why this version refuses it. This version accepts metadata stored as a stream of its own, and reads the whole document's metadata when it is plain or Flate-compressed UTF-8 or UTF-16 XML. A declaration copied into a changed file could claim a standard the new file no longer meets, so Rotate, Protect and Unlock remove any such declaration and say so, and Split and Merge leave the document's metadata out and say so. It refuses metadata stored in another form or also used by another part of the file; document metadata it can't read that way, such as metadata compressed another way or holding a document type declaration; and document metadata that holds a PDF Declaration.
What you can do. If the app that made the file lets you leave metadata out when saving, a copy saved that way may be accepted. Verso has not tested files made that way.
Reason code: feature:metadata-streams
Output intents, archival and print profiles
An output intent says how the file's colours should be reproduced, usually by embedding a colour profile. PDF/A archival files and PDF/X print files often carry one.
Why this version refuses it. Changing the file can break the standard it was made to meet, and Verso does not check PDF/A or PDF/X conformance. Encryption is not allowed in PDF/A at all, so Protect could never keep a PDF/A file conformant.
What you can do. If the file must stay PDF/A or PDF/X, make the change with software that can check conformance afterwards.
Reason code: feature:output-intents
Embedded files or file attachments
Other files stored inside the PDF, such as a spreadsheet attached to a report, either for the whole document or pinned to a page behind an attachment icon.
Why this version refuses it. Files attached to the whole document are stored apart from the pages, so they can be left behind when pages are copied into a new document, and testing has not yet shown that attachments come through a re-save intact.
What you can do. If you need the attached files, save them from a PDF reader first. If you don't, an app that can remove attachments may give you a copy Verso accepts. Verso has not tested files made that way.
Reason code: feature:attachments
Associated-file relationships
Links that say how an embedded file relates to the document or to part of it: as its source, its data or another version of it. Electronic invoice formats such as Factur-X and ZUGFeRD use them to carry machine-readable invoice data next to the readable pages.
Why this version refuses it. The relationship is part of what the file means. For an electronic invoice, the attached data is what accounting software reads, so it has to stay in agreement with the pages.
What you can do. Keep the original. If you need a changed electronic invoice, create it again in the app that issued it, so the data and the pages still match.
Reason code: feature:associated-files
Optional content layers
Layers that a viewer can show or hide, common in technical drawings, maps and print artwork. Some layers can be hidden when the file opens.
Why this version refuses it. If the layer settings are not carried through, content that was hidden could become visible, or visible content could disappear.
What you can do. If the app that made the file can flatten or remove layers, a copy saved that way may be accepted. Verso has not tested files made that way. Check what a flattened copy shows: hidden layers can hold content you don't mean to share.
Reason code: feature:optional-content
Page annotations
Things placed on top of a page: comments, notes, highlights, drawings and stamps. Clickable links are stored as annotations too.
Why this version refuses it. Annotations belong to pages and often point at other parts of the document. When pages are copied or rewritten, those references can break. This version accepts links in the forms it can check: a link to a page of the same file, or to a web or email address in the forms it checks, with no visible border or drawing of its own. Rotate, Protect and Unlock keep them. Split and Merge keep them too, point each jump at its page in the new file, and remove only links to pages left out, saying how many. It refuses every other annotation, and links in other forms, such as a hidden link, a link with a visible border or a description of its own, a single link that covers a large part of the page, and links that overlap or touch and lead to different places.
What you can do. If you don't need the comments or links, an app that can remove them, or flatten comments into the page, may give you a copy Verso accepts. Verso has not tested files made that way.
Reason code: feature:annotations
Custom page labels
Page numbers that a viewer shows in place of the plain page position, such as i, ii and iii for a preface, or A-1 and A-2 for an appendix.
Why this version refuses it. This version accepts labels in the standard numbering styles (1, 2, 3; i, ii, iii; I, II, III; a, b, c; A, B, C), with an optional prefix and starting number, and labels that are only a prefix, such as Cover: Rotate, Protect and Unlock keep them, and Split and Merge leave them out and say so, because when pages are removed, reordered or combined the labels have to be worked out again. It refuses labels in any other form, such as another numbering style or a starting number below 1.
What you can do. If plain page numbers are fine, a copy without custom labels, saved by the app that made the file, may be accepted. Verso has not tested files made that way.
Reason code: feature:page-labels
Document actions or multimedia
Instructions the file gives a viewer: run a script, go to a page or a web address when the file opens or a link is used, submit a form, open another file, or play sound or video.
Why this version refuses it. Actions can refer to pages and objects that change when the file is rewritten, and some do things a reader would not expect. Verso does not pass them on unchecked. This version accepts the kinds it can check: going to a page of the file when it opens, and a bookmark or a link on the page that leads to a page of the file or to a web or email address in the forms it checks. Rotate, Protect and Unlock keep them. Split and Merge leave out the opening action and the bookmarks and say so, and keep links as the section on page annotations describes. It refuses every other action, such as a script, a form submission, opening another file or a chain of actions.
What you can do. If the file came from an app you control, a copy saved without scripts, actions or media may be accepted. Verso has not tested files made that way.
Reason code: feature:active-content
Content stored outside the file
Part of the content, such as an image or a page drawn from another PDF, is kept in a separate file that this PDF refers to, instead of inside the PDF itself.
Why this version refuses it. Verso works only with the file you chose and never fetches the files a PDF refers to, so it cannot check, or keep, content stored elsewhere.
What you can do. If you have the app that made the file, a copy exported with all content embedded may be accepted. Verso has not tested files made that way.
Reason code: feature:external-streams
Features this version doesn't support yet
Parts of a PDF outside the short list this version knows how to carry through a change. Examples include PDF portfolios, a base address for web links, data kept for capturing web pages again, how long a page is shown in a presentation, colour separation and page-boundary colour settings used in print production, and Type 3 fonts.
Why this version refuses it. Apart from data an app keeps under names of its own, anything not on that list is refused rather than dropped or copied unchecked. That app data is still checked for the features on this page; Rotate, Protect and Unlock keep it, and Split and Merge leave out what is stored for the whole document or its pages and say so.
What you can do. Choose a different PDF. If the file came from an app you control, a copy saved with simpler settings may be accepted. Verso has not tested files made that way.
Reason code: feature:unqualified-extensions
Passwords and protection
A password is needed, or it didn't work
A PDF that asks for a password to open is encrypted, and Verso needs the password to read it. You type it into the page; it is used in this browser tab and never uploaded.
If the password didn't work. Check that Caps Lock is off and that you are entering the password that opens the file. If the file opens in other apps without a password but Verso asks for one, it has only a permissions (owner) password; enter that one. Verso can't recover or get around a forgotten password, and does not try to.
If the password uses characters Verso can't use. The kind of protection Verso reads, and the kind Protect PDF writes, take passwords in a basic Western character set: letters, including most accented letters used in Western European languages, digits and common symbols. Emoji and letters from other alphabets can't be used, and only the first 32 characters count. Choose a password without them.
Reason codes: password-required, password-incorrect, password-characters
Protection other than AES-128
PDFs can be protected in several ways. This version opens one: AES-128 encryption with the standard password method (revision 4), with text and content encrypted the same way. Other kinds are refused, including AES-256, the older RC4 methods, certificate-based protection, and AES-128 files with unusual settings, such as metadata kept unencrypted inside the file.
Why this version refuses it. Verso's decryption has been tested for that one method only. Reading another with untested code could give a result that looks right but is not, so this version refuses instead.
What you can do. If you know the password and the file came from an app you control, a copy saved with AES-128 protection, or with no password, may be accepted. Verso has not tested files made that way.
Reason code: encryption-variant
How the file was saved
These refusals are about how the file is built inside, not what it shows. Two PDFs that look the same on screen can be saved in quite different ways.
Later edits saved on top of the original
Some apps save changes by adding them to the end of the file instead of rewriting it, often after a form is filled in, a comment is added or the file is signed. The file then holds the original version plus one or more layers of edits.
Why this version refuses it. Working out which layer is current needs checks of its own, because a mistake could bring back content that a later edit removed. This version makes those checks. When the file is unencrypted, every layer uses the classic index and the layers fit together exactly, it reads the file as its latest version, the one PDF readers show, and keeps only that version: earlier versions are not copied into the new PDF, though they stay in your original file. It refuses a file whose layers fit together in a way PDF readers are known to read differently, for example one where a later edit deletes something the latest version still uses, one whose index points at the wrong layer or skips one, or one with anything but line breaks and spaces between layers. It also refuses one whose edits use the compressed index, one that is password-protected and one that is signed.
What you can do. If this version refuses the file and the app you used can save a complete new copy (for example with Save As rather than Save), that copy may be accepted. Verso has not tested files made that way, and a new copy can gain compressed object streams, which this version also refuses.
Reason codes today: unsupported-structure for the files above that are refused, object-or-xref-streams when a later edit uses the compressed index, too-complex when the file holds more layers of edits or more objects than this version reads, and damaged when a layer is incomplete, is followed by stray bytes or doesn't match its own index. A signed file gets the code for signed files, and one whose earlier version holds encryption settings or fast-web-view data gets the code for those.
Compressed object streams and cross-reference streams
Since PDF 1.5, a file can pack its internal objects into compressed object streams, and store its index of those objects (the cross-reference table) as a compressed stream too. Many apps save this way to make files smaller. A hybrid file carries both the classic index and a compressed one.
Why this version refuses it. This version reads only the classic, uncompressed index, which it checks against every object in the file. The compressed forms need checks of their own, which it doesn't have yet.
What you can do. If the app that made the file can save for compatibility with PDF 1.4 or earlier, a copy saved that way may be accepted. Verso has not tested files made that way.
Reason code: object-or-xref-streams
Files saved for fast web viewing
A linearized file, often labelled "Fast Web View", is arranged so that a browser can show the first page before the whole file has downloaded.
Why this version refuses it. The arrangement depends on extra index data that must match the rest of the file exactly. This version checks that data: it accepts a linearized file in the common form the PDF standard describes when the data matches, and saves the result as an ordinary PDF without the fast-web arrangement. It refuses one that also uses a newer kind of index, one that is password-protected, one that uses the standard's rarer options, such as a second hint stream, and one whose data doesn't match. A linearized file with later edits saved on top is no longer arranged for fast web viewing, and PDF readers read it as an ordinary PDF: this version then checks the data against the original part of the file, apart from the page data the edits make out of date, and reads the file as the section on later edits explains.
What you can do. If the app that made the file has a Fast Web View or optimise-for-web setting, a copy saved with it turned off may be accepted. Verso has not tested files made that way.
Reason codes today: unsupported-structure for the linearized files above that are refused (object-or-xref-streams when the newer index is a compressed one), and linearized when fast-web-view data appears anywhere but at the start of the file.
PDF 2.0 and versions above 1.7
The first line of every PDF names the version of the format it follows. This version accepts PDF 1.0 to 1.7. PDF 2.0 (ISO 32000-2) is refused, and so is a first line Verso doesn't recognise.
Why this version refuses it. PDF 2.0 changed and added rules that this version has not been tested against.
What you can do. If the app that made the file can save as PDF 1.7 or earlier, a copy saved that way may be accepted. Verso has not tested files made that way.
Reason code: pdf-version
Other structures this version can't process yet
The file is built in a way this version doesn't handle that has no section of its own above, or it is a linearized file or a file with later edits saved on top that this version refuses, which the sections on files saved for fast web viewing and on later edits explain. Examples are a stream in a password-protected file whose length is stored in a separate object instead of next to the stream, and objects in the file that its index doesn't list as in use.
Why this version refuses it. It accepts only structures it has been tested with, and refuses the rest instead of guessing.
What you can do. Choose a different PDF. A copy saved by a different app, or with different settings, may be accepted. Verso has not tested files made that way.
Reason code: unsupported-structure
Files that couldn't be checked or read
Parts of the file couldn't be checked
Before changing a file, Verso reads the content of every page to look for the features listed above. Some content it can't read yet, such as some images stored inside the page text (inline images) or a compression method it doesn't decode. It also refuses pages that replace the plain colours with colour profiles of their own (default colour spaces), because viewers apply them differently. It also refuses pages whose size, rotation or resources are set in ways viewers read differently, such as an empty (null) value.
Why this version refuses it. If any part can't be checked, Verso can't be sure the file is free of features it would damage.
What you can do. Choose a different PDF. If the file came from an app you control, a copy exported with different image, colour or compression settings may be accepted. Verso has not tested files made that way.
Reason code: incomplete-inspection
Very long or complex files
Verso limits how many pages, objects and levels of nesting it reads in one file, and how much work one job may do. The same limits apply to the file a job would write.
Why this version refuses it. The limits keep one file from using up the memory of this tab. Beyond them, the browser could slow down or close the tab, so Verso stops early instead.
What you can do. Choose a shorter or simpler PDF. If the message is about the result, select fewer pages, or merge fewer or smaller files.
Reason codes: too-complex, too-many-pages
Damaged or unreadable files
Verso couldn't read the file as a PDF. It may be damaged, for example by a download that stopped early. It may not be a PDF at all, for example a different kind of file renamed to end in .pdf. Or it may have later edits saved on top of the original that are incomplete or in a form this version can't read; the section on later edits explains which ones it reads.
Why this version refuses it. A file that can't be read completely can't be changed safely.
What you can do. Check that the file opens in a PDF reader. If it was downloaded or sent to you, get a fresh copy. If it opens normally elsewhere, later edits saved on top of the original are one possible cause.
Reason code: damaged
Your browser couldn't read the file
The browser couldn't give Verso the file's contents. This can happen when the file is moved, renamed, deleted or changed after you chose it.
What you can do. Choose the file again.
Reason code: read-failed
Files with no pages
The file contains no pages, so there is nothing to work on.
What you can do. Choose a PDF with at least one page.
Reason code: empty
Other messages
Merge needs at least two PDFs
Merge combines files, so it needs two or more.
What you can do. Add another file.
Reason code: too-few-documents
Verso couldn't verify the result
Before it gives you a file, Verso reads the result back and checks it against what it meant to write. If that check fails, it produces nothing rather than a file that might be wrong.
What you can do. Try again. If it happens again with the same file, choose a different PDF.
Reason code: output-check-failed
Something went wrong
Verso hit an error it has no specific explanation for. Your file was not changed.
What you can do. Try again. If it keeps happening, reload the page.
Reason code: unknown
More about Verso
Why local? explains why Verso processes files in your browser, and how to check that nothing is uploaded. All tools lists what Verso can do today.