Inspect what a file actually is

Inspect reads the opening bytes of a file, names the format those bytes actually describe, and shows you the hex — so a PNG named .jpg is caught before something else refuses to open it. For a PNG it walks the chunk list and recomputes every CRC-32 checksum; for a JPEG it walks the marker chain in file order. Nothing is written and nothing is uploaded: the file is read inside your own tab and released when you leave.

How it works

  1. Drop the file The picker here accepts any file, not only images — refusing a file because of its extension would be the mistake this tool exists to catch.
  2. Read the verdict The detected format appears with a confidence of certain, likely, possible or unsure, plus the size, dimensions, extension, the type your browser reported and the first eight bytes. If the bytes disagree with the extension, a warning says what the file really is.
  3. Walk the structure For a PNG every chunk is listed with its offset, type, length and whether its checksum passes. For a JPEG every marker is listed with its offset and length. Other formats get the hex view only.
  4. Move around the bytes Select a structure row to jump the hex view to that offset, or type an offset as a decimal or as 0x. Arrow keys, page up and page down move the view, and you can show 16, 32, 64 or 128 rows at a time.
  5. Copy the report Copy report puts the detection, the facts, the structure table and a dump of the first 256 bytes on the clipboard as plain text.

Questions

Does it show the internal structure of every format?
No. file4edit walks the internal structure of two formats: PNG, as its chunk list with checksums, and JPEG, as its marker chain. Every other format is identified from its signature and given the hex view, with no structure table. Signature detection covers PNG, JPEG, GIF, WebP, PDF, BMP, ICO, TIFF, AVIF, HEIC and SVG.
Can it tell me whether a file is damaged?
For a PNG, yes. Each chunk carries a CRC-32 over its type and its data, and file4edit recomputes all of them — a failed checksum means the file is corrupted. A structure that runs past the end of the file is reported as truncated. For other formats there is no checksum to check.
What happens with a very large file?
Files up to 96 MB are read whole. Above that, only the first 4 MB are read into memory, and the structure walk and the checksum checks stop there — a hex view of a gigabyte would not be useful, and holding a gigabyte in a tab is how the tab dies.

Every tool