A repeatable PDF compression test: compare size, quality, and function
A smaller file is not automatically a better result. A useful compression test holds the source document and test conditions constant, records measurable changes, and checks whether the output still works for its intended purpose.
Build a representative test set
Do not judge a compression mode using only one convenient PDF. Choose disposable or non-confidential samples that represent the documents you actually handle. Keep an untouched copy of every source and give each test file a stable identifier.
- A text-led PDF exported from an office application.
- An image-heavy scan with small print, stamps, or signatures.
- A presentation or report containing photographs and diagrams.
- A PDF with searchable text, links, bookmarks, or form fields.
- A file already optimized by another application.
Control and record the test conditions
Run every mode from the same original file rather than compressing a previous result. Record the SleekPDF mode, browser, device, test date, source size, output size, page count, and processing time. Close unnecessary tabs when comparing processing time because browser memory and device load can affect it.
Calculate size reduction as (source size − output size) ÷ source size × 100. A negative result means the generated file is larger; keep the original in that case.
Choose the first mode from the document risk
This table is a starting rule, not a promised outcome. Mixed documents may need page-by-page inspection because one PDF can contain digital text, scans, photographs, and interactive features together.
| PDF type or goal | Start with | Validate before use |
|---|---|---|
| Digital text, links, or structure matter | Lossless | Text selection, search, links, bookmarks, and page count |
| Ordinary scan or image-heavy report | Balanced | Small text, photographs, stamps, QR codes, and orientation |
| Strict size limit still not met | Maximum | Every critical detail against the untouched original |
| Digital signature or complex form | Original or Lossless | Signature status, fields, calculations, and recipient requirements |
Score more than file size
Mark a result as unsuitable if any essential feature fails, even when the size reduction is large. For an important submission, open the output in a second PDF viewer and test any machine-readable element with the real receiving workflow when possible.
- Size: source bytes, output bytes, and percentage saved.
- Integrity: output opens, page count matches, and page order is unchanged.
- Readability: names, numbers, footnotes, diagrams, and fine print remain clear at 100 percent zoom.
- Machine use: searchable text, copy and paste, links, form fields, QR codes, and barcodes still work where required.
- Visual fidelity: photographs, colors, thin lines, and page orientation remain acceptable.
- Time: processing duration is recorded as context, not as a universal performance promise.
Report results without overclaiming
State the document type, test conditions, mode, and checks performed beside every result. Do not generalize one scan into a claim that all PDFs will shrink by the same percentage. Compression depends on the images, fonts, streams, metadata, and prior optimization inside each source.
SDMV uses this protocol as the basis for future SleekPDF benchmark updates. Results will be published only when the sample files, conditions, and validation method can be described clearly enough for another person to repeat the comparison.
How to merge PDFs without losing page order or important quality
Ready to try it on a copy of your document?
Compress PDF online