Full technical info for any video file

A video file is a container holding several streams, and the report follows that structure. It opens with a General block about the file as a whole, then gives one block per track. Each line is a field name and a value, in the same layout MediaInfo users will recognise.

Two numbers confuse people more than any others. Overall bit rate in the General block counts everything in the file, while the bit rate in the Video block counts the picture alone, so the first is always the larger. And Frame rate mode: Variable is normal for phone footage; the frame rate checker measures the real minimum and maximum from the frame timings.

• Text is what people mean when they ask for “the MediaInfo”. Paste it into a forum thread, a bug report for an editing app, or a delivery check against a broadcaster’s spec sheet.
• JSON has the same values with stable keys, so a script can read them. It suits automated checks, such as confirming every file in a delivery is 25 fps and stereo.
• CSV opens in a spreadsheet as field and value columns, handy for comparing a handful of files by eye.
• The box tree (MP4 and MOV only) shows the boxes the file is made of, in order, with their sizes. It is where you look when a file will not stream until fully downloaded, or when you want to see exactly where a tag or a location is stored.

Reading the box tree

A few box names cover most questions. ftyp says which family the file claims to belong to (an iPhone MOV says qt). moov holds the movie header mvhd with the creation date and duration, one trak per stream, and user data in udta and meta, where titles, Apple’s com.apple.quicktime keys and location strings live. mdat is the media itself and is usually most of the file. If moov comes after mdat, a browser has to fetch the end of the file before it can start playing.

MediaInfo and ffprobe are the two tools professionals reach for, and both are excellent. This report does not replace them for every job; it removes the install step and adds explanations. Here is a fair comparison.

Because the text report comes from the same library as the desktop app, the values usually match what MediaInfo shows for the same file; small differences can appear between library versions. If you need to process hundreds of files or want every packet as data, the ffprobe guide shows the commands. If you want to understand one file, or check it on a phone or a locked-down work computer where nothing can be installed, the browser report is quicker.

Calculate hashes reads every byte and gives MD5, SHA-1 and SHA-256 fingerprints. If two files have the same SHA-256, they are identical, bit for bit. If a single byte differs, the hash changes completely.

• Record the hash when you receive a file, so you can later show the copy you are working on is unchanged. This is the first step in any careful video forensics work.
• Compare against a hash someone published, to confirm a download was not corrupted or swapped.
• Remember that editing a tag or stripping metadata changes the hash even though the picture and sound are untouched. Two files can look identical on screen and still hash differently.

Hashes are computed in your browser. Neither the file nor the hash is sent anywhere.

A full report is thorough, which means it can include things you would rather not post publicly: GPS coordinates from a phone, a camera serial number, the name of the editing software and the exact recording time. Look at Privacy first, which lists what the file reveals and rates each item High, Medium or Low.

If the report is for a support question about playback or encoding, the location and serial numbers are rarely relevant. You can delete those lines from the copied text, or clean the file with Remove all metadata, press Download new file, then Inspect new file and copy the report of the clean copy. The video and audio streams are carried over unchanged, so the technical details a helper needs stay the same. The metadata remover page explains what goes and what stays.