Images & conversion7 min read
Compressing images without visible loss
What a quality setting really controls, what chroma subsampling throws away, which SSIMULACRA 2 scores map to which settings, and why re-encoding hurts.
Omnvert Blog
Measured, hands-on guides on file conversion, PDF, 3D printing, fonts and developer tooling.
PDF & documents6 min read
Getting tables out of a PDF and into Excel
A PDF has no cells, only positioned glyphs. How extractors infer columns from ruling lines and x coordinates, where that fails, and how to verify the result.
3D printing & STL5 min read
How to reduce STL file size without ruining the print
Why STL files get large, how to read the triangle count straight from the file size, and what binary, re-export, decimation and 3MF each save in practice.
Developer tools7 min read
Inspect an EXE without running it: reading PE headers
What the PE headers, section table, import list and per-section entropy tell you about a Windows executable, and where static triage stops being enough.
Developer tools6 min read
JWT debugging safely: decoding is not verifying
JWT structure and base64url, the alg and kid headers, the none and algorithm-confusion attacks, expiry and clock skew, and how to debug a token without leaking it.
Privacy & security6 min read
PDF redaction that actually works
A black rectangle hides text without deleting it. Where PDFs keep the original content, how to remove it properly, and four ways to verify a redacted file.
Developer tools6 min read
Read a pcap without Wireshark: headers, flows and TLS
How pcap and pcapng are laid out, where snaplen truncation bites, how to turn packets into flows and endpoints, and what TLS still leaks in a capture.
Images & conversion7 min read
Removing an AI watermark from an image
What removing an AI badge actually involves, when inpainting gives a clean result, and why erasing the visible mark leaves the embedded provenance intact.
3D printing & STL7 min read
STL vs 3MF vs OBJ: picking the right 3D file format
What each format really stores, how many bytes the same mesh costs in all three, why units decide most arguments, and the edge cases that break silently.
Fonts & typography7 min read
Stylised game fonts are Unicode, not fonts
The fancy letters in Adopt Me and Roblox names are Unicode characters, not a typeface. Which blocks they come from, where they get rejected and what they cost.
Fonts & typography8 min read
Web font formats and subsetting: WOFF2 done properly
What really separates TTF, OTF and WOFF2, why subsetting saves the most, and how a Latin-only subset silently breaks Turkish text on a live page.
Images & conversion6 min read
WebP vs AVIF: which format wins, and where
Measured compression figures with sources, WebP's 8-bit and 4:2:0 ceilings, AVIF's encode cost and missing progressive rendering, and how to choose per content type.
PDF & documents6 min read
What decides OCR quality on a scanned PDF
OCR accuracy is set by the image, not the engine: resolution, binarisation, skew, compression artefacts, language data and layout. Plus how to judge the output.
Privacy & security7 min read
What photo metadata reveals: Exif, IPTC and XMP
The location, timestamps, serial numbers and embedded thumbnail inside a photo, what platforms really strip, and how to verify a file is actually clean.
Images & conversion
Compression, resizing, format choice and measuring quality loss.
3
PDF & documents
Merging, OCR, pulling tables into Excel and redaction.
2
3D printing & STL
STL file size, mesh repair and pre-slicer checks.
2
Developer tools
EXE/PE inspection, pcap analysis, JSON, JWT and encoding work.
3
Fonts & typography
Font formats, subsetting, web font loading and licences.
2
Privacy & security
Metadata hygiene, redaction mistakes and where files get processed.
2