Safe Use Guide
AutoOffload is built around one promise: your shoot is safe. The app enforces that in code — but a few simple habits make your whole workflow bulletproof. Please read this before you rely on AutoOffload for irreplaceable photos or video.
The golden rules
- Test first. Before trusting AutoOffload with a real shoot, run it once with a card you've already backed up, or with a few non-critical files. Confirm the copy lands where you expect and verifies cleanly.
- Keep your originals until AutoOffload reports success. Don't format or delete a card until you've seen a fully verified run and spot-checked the files at the destination.
- Keep an independent backup. AutoOffload makes copying to two destinations at once easy (a working drive and a backup). A single copy of an irreplaceable shoot is never enough — use the multi-destination feature, or back up the archive separately.
- Never wipe a card until you've manually verified the destination. "Verified" in the app means the bytes match. Still, open the folder and confirm your shots are there before you reach for Format in the camera.
- Don't disconnect a card or drive mid-offload. Let the run finish, pause it, or cancel it from inside the app first.
What AutoOffload guarantees
These are enforced by the app, not just promised:
- Your source card is read-only. AutoOffload only reads from cards and removable media — it never writes, renames, moves, or deletes anything on the source. A write that would land on a card is refused in code.
- Every file is verified. Each file is checksummed (BLAKE3) as it copies and re-read at the destination to confirm a match. A copy that doesn't match is deleted and reported — it never masquerades as a good file.
- Interruptions are safe. Files are written under a temporary name and only put into place after they've copied and verified, with the data flushed to disk first. A crash, power loss, force-quit, or cancel never leaves a half-written file pretending to be a real one. Leftover temp files from an interrupted run are cleaned up automatically.
- Nothing fails silently. A single bad file never aborts the rest of the card — it's recorded and reported, and the remaining files keep copying. Files that don't match any time window are reported, never quietly dropped.
- You get a record. Every run writes a per-file CSV manifest and a shareable HTML report alongside your files, so there's always proof of exactly what copied and what verified.
What happens if a run is interrupted
If you cancel, force-quit, lose power, or unplug a drive during an offload:
- No half-written file is left behind posing as complete.
- The partially-written file is removed; fully copied-and-verified files are kept.
- The manifest records what completed, even after a crash.
- You can simply run the offload again — AutoOffload recognizes files already copied (by checksum, not just name) and copies only what's missing.
Pulling the card itself mid-run is detected and recorded rather than crashing the app — but it's still best to pause or cancel first.
What "success" and "failure" mean
- Success / verified: every selected file was copied to every destination and its checksum confirmed. This is your green light to format the card.
- Failure / needs attention: one or more files didn't copy or didn't verify. The app lists them individually and keeps your cards mounted. Do not format the card — re-run the offload (it resumes only the missing files) or investigate the listed files first.
Long-term safety: re-verify your archive later
Verification at copy time proves the file was written correctly. Months later, drives can develop bit-rot or quietly corrupt files. AutoOffload writes a manifest you can use to re-verify a finished archive at any time — it re-reads every file and confirms it still matches, or flags exactly what changed. The app also reminds you when an archive hasn't been checked in a while.
This is the difference between "we copied it once" and "every shoot is provably still good." Build a periodic re-verify into your archive routine.
Next: Quick Start · Troubleshooting · FAQ