Why your PDF is still huge after compression

You ran compress twice and the file shrank from 18 MB to 17 MB. Email still bounces. That usually means the PDF was already squeezed, or every page is a full-resolution scan where further lossy compression hits a quality floor fast. Compression tools re-encode embedded images and trim font data; they cannot invent smaller scans without visible blur. Start from the original export, delete pages you do not need, rescan text forms at lower DPI, then run one pass at ebook or screen on Compress PDF. For provider attachment caps and preset choice, read Compress PDF for email attachments first. Free accounts get five PDF operations per day, 25 MB per file. When bytes still will not budge, Split PDF sends only the pages that matter.

Why compression sometimes barely changes file size

PDF size is mostly embedded images, font subsets, and page count. Vector text is tiny. A ten-page contract exported from Word might sit under 500 KB before you touch compress.

A ten-page phone scan might be 15 MB because each page is a 300 DPI color photo inside the PDF shell. Compress re-encodes those photos. If they were already saved at JPEG quality 60, another pass saves kilobytes, not megabytes.

Some PDFs carry duplicate objects from bad merges: the same logo embedded on every page instead of once. Generic compressors may not deduplicate; the file stays fat.

Already-compressed PDFs

Running compress on a file you compressed yesterday often shows minimal gain. Each pass throws away more detail for diminishing returns. Work from the pre-compress original when you still have it.

“Optimize PDF” in one tool and “Compress PDF” in another may use similar algorithms. Chaining both surprises people who expect 50% each time.

Scanned pages dominate the byte count

Scan apps default to photo mode at 300 DPI color. Black text on white paper does not need color or 300 DPI. Grayscale at 150 DPI can drop a page from 1.5 MB to 200 KB before online compress runs.

Open the PDF and zoom a paragraph. If the page looks like a photograph of paper, compress is fighting image weight, not text.

Fonts, layers, and hidden bloat

Design PDFs from Illustrator or CAD exports embed fonts and keep editable layers. Email attachments rarely need layers. Export a flattened PDF from the source app when possible.

Embedded font subsets add hundreds of kilobytes per family. Usually not the main problem unless the file is otherwise tiny.

Metadata and attached files (XML, spreadsheets) hide in some enterprise exports. Inspect with a desktop PDF optimizer if size makes no sense for page count.

Duplicate pages from bad merges

Merged two packets and kept both cover pages? Each duplicate full-page scan doubles weight for that sheet. Edit PDF pages deletes extras before compress.

Merge PDFs for tax returns and bank statements stresses ordering; the same merge step is where duplicates sneak in.

Transparency and high-bit images

Slides with partial transparency rasterize to large bitmaps on export. Charts exported as huge PNG inserts balloon PDFs. Fix in the source deck, not only at compress time.

What to try when compress PDF did almost nothing

Stop re-compressing the same output. Return to the newest source file: Word export, camera scan, or plotter PDF.

Delete pages. A forty-page manual where the recipient needs three pages should be Split PDF first, not screen preset on the whole file.

Change capture settings. Rescan forms at grayscale 150 DPI. Export slides as PDF without speaker notes pages.

One disciplined compress pass

Pick ebook for mixed text and photos on Compress PDF. Check readability at 100% zoom. If still over cap, try screen once.

Document the file size after each step in a note so you do not loop five times on the same bytes.

Rescan and re-export beats endless compress

A 300 DPI color scan of a black-and-white form is the classic “PDF not compressing” case. Rescan smaller; compress once lightly.

From Word or Google Docs, use Save as PDF instead of Print to PDF through an old driver. Driver PDFs sometimes embed huge preview bitmaps per page.

Walkthrough: second pass with Toolsy Compress PDF

Upload the original PDF if you have it. If you only have the twice-compressed copy, results will disappoint; hunt the source first.

Choose ebook. Download. Compare megabytes in file properties.

Open two pages: dense text and a photo. Soft text means step up to printer or fix the scan.

When screen preset is appropriate

Photo catalogs, mood boards, and scan-only packets without small type tolerate screen. Footnotes and 8 pt legal text do not.

Match preset to the strictest reader in the chain: legal, finance, or the government portal.

Honesty about the 25 MB cap

Toolsy free tier accepts one PDF up to 25 MB per job. Multi-gigabyte archives need desktop tools or split workflows. Compress PDF for email attachments covers honest limits alongside Gmail caps.

Related fixes when size will not drop enough

Compression has a floor. Splitting, deleting, and rescans move that floor.

Split PDF extracts the range you must send. Edit PDF pages removes blanks and duplicates. Merge PDF after compressing pieces separately sometimes beats one giant pass.

When bloat is standalone photos before they entered the PDF, Compress images before email fixes upstream JPG weight.

Email vs portal targets

Gmail tolerates roughly 25 MB per message. Job portals want 2–5 MB total. A PDF that “compressed enough” for Gmail still fails a strict portal. Target the lower number.

Link to cloud storage when policy allows and attach a smaller excerpt PDF.

Text extraction instead of smaller scans

If the goal is search or RAG, shrinking scan PDFs is the wrong job. PDF to Markdown step by step addresses text pipelines. Compression stays in the send-and-store lane.

When file size is not the real problem

Sometimes upload fails for permissions, filename characters, or DLP rules while size looks fine.

Password-protected PDFs may fail portals without asking for size. Unlock or export an unprotected copy.

Corrupt PDFs show normal size but break on upload. Re-export from source or print to a fresh PDF on desktop as last resort.

Viewer vs server limits

Your desktop opens a 30 MB PDF fine. The server rejects at 10 MB inbound. Ask the recipient for the limit instead of compressing blindly.

Mobile mail apps download whole attachments on cellular. Smaller files help recipients even when server caps allow more.

Limits, privacy, and expectations

Compress discards image detail you may need for print later. Keep an archival master uncompressed on disk or cloud.

Toolsy processes uploads for the job and does not keep them after response. Sensitive scans follow the same trust rules as any online PDF tool.

Free: five PDF jobs per day. Pro removes the cap. Neither removes the 25 MB per-file limit.

Frequently asked questions

Why is my PDF still so big after compression?

Most often every page is a high-resolution scan or photo that was already JPEG-compressed inside the PDF. Another pass cannot shrink much without blur. Delete unneeded pages, rescan text at lower DPI in grayscale, or split out only the section you must send. Compress once from the best source file you have.

Why does compress PDF not make the file smaller?

The PDF may be mostly vector text with little to compress. Or images were pre-optimized at quality 50 and have no fat left. Merged duplicates and driver-printed PDFs also resist shrinkage. Inspect page content at zoom: photo pages need rescans; text exports need fewer pages, not stronger compress.

Can I compress a PDF that was already compressed?

You can, but expect small gains and more artifacts. Each lossy pass softens scans. Prefer one compress from the original export. If you already ran compress twice, go back to the scan or Word file instead of a third online pass.

How do I reduce PDF file size when online compress fails?

Rescan at 150 DPI grayscale for forms. Edit PDF pages to drop blanks. Split PDF to send a page range. Re-export from Word or slides instead of compressing a print-to-PDF file. See Compress PDF for email attachments for preset order.

Do scanned PDFs compress worse than Word PDFs?

Yes. Word PDFs store text as vectors; scans store photos of pages. Compress targets photos. A five-page scan might stay large until you change capture settings, not until you click compress five times.

Why did my PDF grow after compression?

Rare, but possible when the tool re-embeds images differently or adds objects. More often you compared against a mislabeled file or a different page count. Check properties on both files and page count in the viewer.

What is the smallest I can get a scan PDF for email?

Depends on readable text. Many scan packets reach 3–8 MB at ebook for email. Strict 1 MB portals need grayscale 150 DPI scans plus screen. Legal fine print may force a cloud link instead of attachment.

Will compressing hurt signatures and stamps?

Aggressive screen on scan PDFs can soften stamp edges and handwritten signatures. Use printer or ebook when those marks must survive printing. Test zoom on the signature page before you send.

Is it normal for compress to only save one megabyte?

Yes on an already-optimized 15 MB scan PDF. One megabyte might be enough to clear a 10 MB portal cap if you were at 11 MB. If you needed half the size, rescans or splits are the realistic path.

When should I stop compressing and split the PDF instead?

Stop when text or stamps look soft or file size barely moves between presets. Split delivers the pages the recipient asked for without destroying quality on the rest. Keep the full PDF on cloud storage; attach the excerpt.

If email limits are the blocker, start with Compress PDF for email attachments before diagnosing a stubborn file.