Memory cards · job record · RDG-2025-2057
Full-Size Files, and Not One of Them Plays.
Overnight the card turned against him. A one-man video business in Tilehurst sent it in after filming a marquee wedding at a farm west of Theale, describing it plainly: three files from the end of the evening simply won't import
. What he said next framed the whole job — the sizes are all still there in the listing, so the footage must be on the card
— and the couple flew out a fortnight later.
Same fault on yours? Tell us what it is doing.
0800 6890668
What actually caused it.
He had read it correctly. A recording is written in two parts. Frames go onto the card one after another as they are captured. The index is written only at the end, as a file closes in good order: a table of where each frame begins and how long it lasts. Pull the card, run the battery flat or run out of space, and the frames stay put while that table never gets written. An editor has nothing else to navigate a clip by, which is why a file packed with video but carrying no map looks exactly like an empty one on screen.
What this one needed on the bench.
How every job runs →| Bench tools | Its part in this job | Why it is here |
|---|---|---|
| Klennet Carver | Assembled the frames that remained into files his editing suite would accept | It knows the internal build of a video file; blunter tools hand back fragments that play nowhere |
| PC-3000 Flash | Copied the card sector by sector before anything else was tried | Reads the NAND direct, past the controller, against a maker-ID library kept current |
| UFS Explorer Professional Recovery | Took the undamaged clips, and the folder layout, off that copy | Opens volumes other tools refuse: APFS, ReFS, XFS, ZFS and Btrfs |
What the bench did.
Copy first, then diagnose
No diagnosis began until the copy existed. Somebody else's deadline can change how fast a job is run; it never licenses anyone to gamble with the only original in the building. Work from there on ran against the image alone, while the card sat in an antistatic bag in case a second pass was ever wanted.
Put the working clips somewhere safe first
Nearly the whole evening lifted straight off the copy and was put somewhere safe before anything harder began. Only then were the three difficult files opened and worked through. His account held up: all three held frame data, two had an index half-written, the third had none. That dictated the order of everything afterwards.
Take the pattern from clips that came out undamaged
Video is seldom laid down in one continuous run, and it is the scattering that defeats one-click software — whatever it returns will play nowhere. The pattern came from undamaged clips filmed earlier the same evening on the same body with the settings unchanged. Each broken file was then reconstructed around the frames it still held, and tested in the only way that counts: played from the first frame to the last.
What went back.
The first two came back entire, each watched from start to finish on the bench before it left here. The third plays cleanly until a few seconds from the end; those frames were still queued inside the camera when the card came out, and never reached the media at all. The edit was with the couple before their flight.
If this sounds like yours, start here.
Other card jobs in the log.
If that sounds like yours, stop there.
Post it in as it is, switched off. Looking costs nothing, and you get one fixed figure in writing before any chargeable work begins — £250 + VAT for a card or stick, £300 + VAT for a single drive, from £500 + VAT for RAID and NAS. On most jobs, nothing is charged unless your data comes back.