When standard PDF compression is not working because a portal keeps rejecting your file with a structural error, or because the file is simply too complex for the system to process, converting to images and rebuilding is often the most reliable fix. It sounds counterintuitive to go from PDF to images and back to PDF, but this technique strips out every layer of complexity in the original file and produces a simple, universally compatible PDF that passes almost any portal validator.
Why this technique works
A PDF file is not simply a collection of pages. It is a structured document format that can contain multiple layers of content: text layers, image layers, form fields, annotations, embedded fonts, JavaScript, embedded media, revision history, links, bookmarks, digital signatures, and various metadata blocks. Every piece of complexity adds overhead and gives portal validators something to reject.
A scanned PDF that has passed through multiple hands can accumulate a surprising amount of structural complexity. Someone annotates it in Acrobat. Someone else fills in a form field. A third person signs it digitally. Each of these actions adds a layer to the file. The file works fine when opened in a desktop PDF reader that understands all these layers, but when a web portal tries to process it, any element it cannot parse becomes a reason to reject the file.
When you convert the pages to images and rebuild as a new PDF, all of this complexity disappears. The resulting file contains only one thing: a sequence of image pages. There are no form fields to parse, no embedded fonts to load, no revision history to unwrap, and no scripts to execute. The portal validator has almost nothing to fail on.
When to use this technique
This approach is most useful in specific situations. It is not the first step for every PDF problem but it is often the right step when others have failed.
Use it when a portal gives a structural or compatibility error on a file that is within the size limit. Errors like "file cannot be processed," "unsupported file format," or "invalid PDF structure" when you are submitting a genuine PDF are signs of structural complexity. Converting to images resolves all of these.
Use it when a heavily annotated or form-based PDF needs to be submitted as a final read-only document. Flattening removes all interactive elements and produces a clean submission copy.
Use it when you need to definitively remove all metadata and revision history from a PDF. The image conversion strips everything except the visual content of each page.
Use it when a scanned PDF is larger than expected after standard compression. The image rebuild process applies fresh compression rather than trying to compress an already compressed file, and often produces a smaller result.
Do not use it when the PDF contains selectable text that needs to remain selectable. Converting to images removes all text layers. A recipient who needs to copy text from the document or use it with a screen reader will be unable to do so after this conversion. For documents where text searchability matters, use standard compression rather than image conversion.
How to do it
- Split the document first if it is long. For a multi-page document, especially one above 20 pages, splitting before converting makes the process more manageable. Use Split PDF to divide the document into sections, or process it as a whole if the page count is manageable.
- Convert pages to images using PDF to JPG for photographic content or documents where colour and gradients are important, or PDF to PNG for documents with text, line art, or sharp edges where you need the sharpest result.
- Choose the right resolution. For a document that will only be read on screen or through a portal, 150 DPI is sufficient and produces the smallest file. For a document that may be printed or zoomed in for close reading of small details like certificate numbers or stamps, 200 DPI is a reasonable balance. There is rarely a reason to go above 200 DPI for portal submissions.
- Rebuild the PDF using JPG to PDF. The resulting file will be a simple image-based PDF with no structural complexity.
- Verify the result by opening it and checking that all pages are present, correctly oriented, and that text is readable at 100% zoom. Pay particular attention to small text such as signature blocks, reference numbers, and dates.
Choosing between JPG and PNG for the intermediate images
The format you use for the intermediate images affects the quality and size of the final rebuilt PDF.
JPEG intermediate images produce smaller final PDFs but introduce compression artefacts around sharp edges and text. For documents with important small text, handwritten content, or precise graphical elements, these artefacts can make the final PDF harder to read. Use JPEG intermediate images when file size is the primary concern and the content is photographic or scanned at relatively low detail.
PNG intermediate images preserve sharp edges and text exactly but produce larger intermediate files. The final rebuilt PDF from PNG images will be larger than one from JPEG images. Use PNG when the document contains typed text, stamps, handwriting, or fine graphical detail where quality must be preserved.
Size expectations after the rebuild
The size of the rebuilt PDF depends primarily on the resolution and format you choose for the intermediate images. At 150 DPI with JPEG compression, a single A4 page of a standard certificate typically produces a file of 50KB to 150KB. A 10-page transcript at the same settings would be approximately 500KB to 1.5MB, well within the 2MB limit of most scholarship portals.
At 200 DPI with PNG intermediate images, the same single page would be 200KB to 500KB. For most portal submissions, the JPEG-based 150 DPI approach produces a file that is small enough while remaining clearly readable.
Limitations and considerations
The rebuilt PDF loses all interactive features. Form fields, clickable links, digital signatures, and bookmarks are all gone. If the document had a digital signature that proves authenticity, converting to images removes that signature. For submissions where the official digital signature is part of what makes the document valid, this technique may not be appropriate. Check whether the receiving institution requires or accepts electronically certified PDFs before flattening.
Text searchability is also lost. If a reviewer needs to search within the document by keyword, or if the document will be processed by optical character recognition for archival purposes, a text-based PDF is preferable. The image rebuild approach should be reserved for situations where readability is the only requirement.