How to merge PDFs safely
Safe merging means: order you control, page counts verified after the job, and originals untouched until you confirm the output. Merge into a new file — never 'save over' the only copy.
It matters because phones and scanners produce one PDF per page; contracts, applications and expense reports need one file in the right order with nothing missing.
Mistakes: merging encrypted files without unlocking first, mixing page sizes without checking (printers scale oddly), and trusting filenames instead of opening the merged result.
Example: scan_01..03.pdf (1 page each) → merged.pdf must report 3 pages. If it reports 2, a file was corrupt — redo, don't send.
Try: Merge PDF · Batch Merge Queues
Reduce PDF size without killing quality
PDF weight comes from images inside: a scanned page at 300 DPI color can be 1 MB; the same page downsampled to 150 DPI grayscale is ~100 KB and still reads perfectly.
Shrink in order: downsample images to the needed DPI (150 screen, 300 print), drop to grayscale where color adds nothing, then re-save to compact object streams.
Mistakes: re-compressing an already optimized PDF expecting miracles, downsampling signatures/stamps into illegibility, and emailing 50 MB files instead of linking cloud storage.
Example: 10 scanned pages, 24 MB → 150 DPI grayscale re-save ≈ 2.5 MB. Same words, one-tenth the attachment.
Try: Compress PDF (local) · Size Analyzer
PDF or image? choosing a format
PDF keeps vector text selectable and prints identically everywhere; images (JPG/PNG) are universal but freeze text into pixels. Choose by what the reader must do with it.
Send PDF for forms, contracts and printing. Send images for quick chat viewing, social posts, or when the recipient's phone has no PDF reader.
Mistakes: photographing a document when a scan-to-PDF keeps it sharp and small, embedding giant photos in a 'PDF' that is just a wrapper, and rendering text pages to JPG (blurry zoom, huge files).
Example: a 12 MP whiteboard photo (4 MB JPG) → A4 PDF ≈ 300 KB, prints sharp. A one-page flyer → JPG preview for chat, PDF for print.
Try: Images → PDF · PDF → JPG (render)
OCR a scanned PDF locally
A scanned PDF is photos of pages: nothing to select, search, or copy. OCR reads the pixels and rebuilds the words — on your device with Tesseract, nothing uploaded.
It matters because the alternative is uploading sensitive scans to a server queue. Local OCR keeps contracts, IDs, and medical pages on your machine.
Mistakes: expecting handwriting miracles (print works, cursive often fails), OCRing tiny renders (bigger pixels read better), and trusting proper nouns without a glance.
Example: a 1-page scan → English OCR → HELLO OCR 123 recovered as selectable TXT in seconds, verified by eye.
Try: OCR Scanned PDF · PDF → Text
Redact PDFs for real
Real redaction has two halves: cover the words AND remove the text layer underneath. Black boxes alone leave every character searchable and copyable — that is how leaks happen.
It matters for contracts, court filings, and medical records: the tool burns boxes then rebuilds the pages as images, so the output holds zero extractable text. Verify by searching the result.
Mistakes: drawing annotations instead of burning (they peel off), redacting the scan but sharing the original, and partial-word covers leaving readable fragments.
Example: a page containing proof redacted for proof → runs covered, the output searches clean for the term, page count intact.
Try: Redact PDF · Unlock PDF (known password)