PDF Flatten & Scan-Look Tool

You fill in a visa application, a benefits claim or a loan form, save it, upload it, and the portal says the file is invalid. Or it accepts the file, and the reviewer later reports that half the answers are blank. Both problems usually come from the same place: a fillable PDF is not one flat document but a page with a separate, interactive layer on top. This guide explains how that layer works, what “flattening” really does, when an image-only PDF is the better answer, and how to check a file before you send it.

How a fillable PDF is built

A PDF is a container of numbered objects: pages, fonts, images and the instructions that draw them. A fillable form adds an extra structure called the AcroForm, listed in the document’s catalogue. It holds every field in the document, and each field records its type (text, checkbox, radio button, list, or signature), its name and its current value.

The field itself is not what you see on the page. What you see is a widget annotation: a rectangle placed on a page, carrying an appearance stream, which is a small drawing of the field in its current state. When you type into a field, your PDF reader updates the value and redraws that appearance. The value and the drawing are stored separately, and nothing guarantees they agree. A form can also carry a flag called NeedAppearances, which tells the reader “ignore the stored drawings and redraw every field yourself”. Readers handle that request differently, which is the first reason the same form can look different in Adobe Reader, a web browser and a document system.

Some forms, especially older government ones, use a second technology called XFA (XML Forms Architecture). An XFA form describes its fields in XML, and a dynamic XFA form is drawn entirely by the reader. Adobe deprecated XFA, and the PDF 2.0 standard removed it. Browsers, phone apps and most document systems cannot draw dynamic XFA forms, and often show only a “Please wait…” message instead. If your form was built this way, you usually need Adobe Reader to fill it and print it to a new PDF.

What flattening does, and what it does not

Flattening takes each widget’s current appearance and draws it permanently into the page’s own content, the same layer as the printed text of the form. Then it removes the widget annotations and the field definitions. Your answers, ticks and selections stay exactly where they were, but there is no longer anything to click or type into. A flattened form behaves like any other document: it prints, merges, stamps and displays the same way everywhere.

In this tool’s Flatten mode, the result is still a normal vector PDF. Text remains text, so it stays crisp at any zoom, can be selected and searched, and the file usually stays about the same size. Where a field has no stored drawing, the tool draws its value in a standard font before flattening, and it tells you if any field could not be drawn.

Two common misunderstandings are worth clearing up:

• Flattening is not security. It stops casual edits through form fields, but anyone with a PDF editor can still change the page content, as with any PDF. If a document must not be altered, the receiving organisation should rely on a digital signature or its own copy, not on flattening.
• Flattening is not redaction. Drawing a black box over text in a PDF does not remove the text underneath, and flattening does not change that. Information that must be removed has to be deleted with a proper redaction tool, or the pages converted to images, which destroys anything not visible.

Why portals and document systems prefer flat PDFs

Case management systems, applicant tracking systems and university submission portals are built to store and show pages, not to run forms. Interactive fields cause predictable problems for them:

• Vanishing answers. When a system merges your file with others, adds a reference stamp or converts it for archiving, form fields can be dropped or redrawn blank, especially if they depend on NeedAppearances.
• Values that can change. An interactive field can be edited, and some forms contain scripts that recalculate totals or reformat dates when the file is opened. A reviewer may not see exactly what you submitted.
• Security rules. Some organisations block PDFs that contain form fields or JavaScript, because active content is a common route for malicious files.
• Archiving standards. Long-term formats such as PDF/A restrict interactive and dynamic content, so systems that convert uploads to PDF/A work best with flat input.

If a portal’s error message mentions “editable”, “fillable”, “interactive”, “form fields” or “protected content”, flattening is the first thing to try. If it asks for “scanned”, “image-based” or “non-searchable” documents, it wants an image-only PDF.

Vector flattening versus image-only PDFs

An image-only PDF is the most thorough kind of flattening. Each page is drawn at a chosen resolution and stored as a picture, and nothing else survives: no fields, no annotations, no hidden layers, no invisible text, no scripts, no attachments and no metadata. What the receiver gets is exactly what you saw in the preview.

The cost is real. Text becomes pixels, so it cannot be selected, searched or copied; links stop working; and the file is usually larger than a vector original. Resolution sets the balance between clarity and size. 150 dpi is sharp on screen and keeps files small. 200 dpi matches a typical office scan and is a safe default for most portals. 300 dpi is best when the document will be printed or run through text recognition, at roughly twice the file size of 200 dpi. If the result is too large for a portal’s size limit, choose a lower resolution or the Smaller file size setting, or run the finished file through a compressor built for exact limits.

Metadata and data you cannot see

PDFs carry information that never appears on the page. The document information dictionary stores a title, author, subject and keywords, plus the names of the programs that created and last saved the file. A separate XMP metadata stream can repeat all of that and add more, such as edit history from design software. Forms add their own leftovers: field names, default values and, in files saved by “incremental update”, earlier versions of the document that are still present after the visible content changed.

When the tool flattens a file, it rewrites the whole document, so earlier saved versions are dropped. With “Remove hidden document details” switched on, it also clears the information dictionary and the XMP stream, and deletes any objects the rewritten document no longer uses, including the old field definitions. Image-only output starts from an empty document and contains only the page images.

Text recognition (OCR) and accessibility

Optical character recognition turns pictures of text back into text. Many organisations run OCR on every upload so documents can be searched later, and some portals check that a file’s text can be read before they accept it. OCR accuracy depends on resolution, contrast, straightness and cleanliness. Engines generally work best at around 300 dpi with dark text on a clean, light background and pages that are not tilted. Every scanner-style effect in this tool makes recognition slightly harder. If you know the receiver will run OCR, use the Clean preset, or keep tilt, grain and softness low and choose 300 dpi.

Accessibility follows the same pattern. A vector PDF with real text can be read aloud by screen readers, and a well-made, tagged PDF also tells them the reading order and structure. An image-only PDF is silent to assistive technology unless OCR adds a text layer. If your document goes to people who use screen readers, or to a public body with accessibility duties, prefer Flatten mode.

The scan look: when it helps and when it does not

The Image-only mode can make pages look as if they came off an office scanner: greyscale or black-and-white output, stronger contrast, a slight tilt that can vary from page to page, paper grain, a little softness, an off-white paper tone and a shadow along the edge. These effects have honest uses:

• Making a typed page match the real scans it will be bundled with, so a document pack looks consistent.
• Meeting a portal’s requirement for image-based or scan-style PDFs.
• Checking how fine print, signatures or forms will hold up after being photocopied, faxed or scanned.
• Creating realistic sample documents for training material, mock-ups and software testing.

What the scan look cannot do is turn a typed document into a physically signed one. When an organisation asks for a document that was printed, signed by hand and scanned, it is asking for that process, often as evidence of who signed and when. A scan-look copy of a typed file is not that, and presenting it as one can break the organisation’s rules, a contract or the law. If you cannot print and sign, ask whether an electronic signature or a typed copy is acceptable. Many banks, employers and public bodies now accept them.

A checklist before you upload

1. Fill in the form completely and save it in your PDF reader before flattening. Flattening fixes whatever the fields contain at that moment.
2. Read the portal’s instructions for file type, size limit, page count and whether it wants searchable or image-based files.
3. Use Flatten mode first. It keeps the file small and the text searchable, and it solves most “editable PDF” errors.
4. Page through the preview, then open the finished file with “Open to check” and confirm every answer, tick and date is visible.
5. Switch to Image-only only if the portal still refuses the file or explicitly asks for image-based documents, and keep the effects light if the receiver will run OCR.
6. Check the file size against the portal’s limit before you upload.
7. Keep the original fillable file. Flattening is one-way, and you may need to correct an answer later.

Handled this way, a picky upload portal stops being a mystery: give it a file with one fixed layer, check that what you see is what you meant, and keep the original in case you need to change it.