Guides
Formats, explained by people who convert them all day
What a format actually stores, why a conversion drops something, and how to check before and after. No filler.
What is HEIC?
HEIC is Apple’s common HEIF photo file: small, high-quality, and sometimes hard to open on Windows or older apps.
DOCX vs DOC: what changed?
DOCX is a zipped XML document package. DOC is Word’s older binary format with weaker interoperability and harder recovery.
WebP vs PNG: when to use each
WebP can be lossy or lossless and smaller. PNG is still the conservative choice for exact pixels and wide tooling.
JPG vs PNG: which should you use?
JPG is lossy and suited to photos. PNG is lossless and suited to screenshots, logos, transparency, and exact pixels.
AVIF vs WebP: image format differences
AVIF uses AV1 image coding and can compress very well. WebP is older, faster to adopt, and still easier in many tools.
PDF/A explained: archive PDF basics
PDF/A is a restricted PDF profile for long-term preservation. It embeds fonts and bans risky dependencies.
EPUB vs MOBI vs AZW3 for Kindle
Kindle sending now prefers EPUB. Devices still use Amazon formats internally, while MOBI is mostly legacy.
SRT vs VTT: subtitle format differences
SRT is simple subtitle text. VTT adds web cues, styling hooks, and metadata used by HTML video players.
EML vs MSG: email file differences
EML is RFC 5322 message text. MSG is Outlook’s MAPI data in an OLE compound file with Outlook-specific fields.
Parquet vs CSV: data format differences
CSV is row text for exchange. Parquet is typed, columnar, compressed storage for analytics and data lakes.
JSONL vs JSON: when to use each
JSON is one document. JSONL is newline-delimited JSON records for logs, streams, and append-only datasets.
YAML vs JSON: practical differences
JSON is strict data syntax for machines. YAML is more readable but has indentation, typing, and parser traps.
How to tell if a PDF is scanned
Learn quick tests for scanned PDFs: selectable text, embedded fonts, OCR layers, image-only pages, and command-line checks.
OCR vs text layer in a PDF
OCR creates text from page images. A PDF text layer stores selectable text that may be visible or hidden.
PDF to Word without losing formatting
PDF-to-Word layout loss comes from fixed-position pages. Diagnose text layers, fonts, columns, tables, and scans first.
Markdown to PDF with Pandoc and Typst
Use Pandoc for Markdown parsing and Typst for PDF layout when you want repeatable command-line documents.
Markdown to DOCX with Pandoc
Pandoc can make editable DOCX from Markdown, but reference documents, tables, images, and raw HTML decide the result.
How to send EPUB to Kindle
Send to Kindle accepts EPUB and converts it for delivery. MOBI and AZW are no longer accepted for sending.
How to export Outlook email to PDF
Outlook email to PDF depends on body format, headers, attachments, and whether the source is EML or MSG.
LibreOffice headless convert CLI guide
Use `soffice --headless --convert-to` for server-side Office conversion, with checks for filters, fonts, and profiles.
Handwriting to text OCR accuracy guide
Handwriting OCR depends on image quality, writing style, language, layout, and how much manual review you can afford.
CSV delimiter problems in Excel
CSV delimiter errors in Excel come from locale separators, quoting, RFC 4180 variation, and files that are not really CSV.
CSV UTF-8 BOM and Excel
Excel may assume ANSI for CSV on Windows. A UTF-8 BOM or import workflow prevents broken accented characters.
CSV to JSON with Python and pandas
Use pandas for CSV to JSON when you control delimiter, dtype, encoding, date handling, and output orientation.
JSON to CSV nested flattening guide
Nested JSON must be flattened before CSV. Choose record paths, metadata, separators, and array handling deliberately.