Troubleshooting · How we test and maintain this guide
Start with the selected count
The ZIP action uses selected ready tracks. It does not include every input link, every title shown, or every ready result automatically after you change the checkboxes. With no selected ready audio, Build ZIP cannot create a useful music archive. Check the count beside the results before investigating the browser or changing formats.
A failed row is a record of an attempt, not an audio file waiting to be unpacked. An unchecked ready row is deliberately outside the archive. The CSV and JSON still record those rows so you can understand the full batch. Seeing their names in a manifest does not mean the associated music should appear as a ZIP entry.
Read the example in the right units
| In the controlled partial example | Count |
|---|---|
| Input result rows | 20 |
| Ready rows | 12 |
| Ready rows deliberately unchecked | 2 |
| Audio entries intended for the ZIP | 10 |
| Additional record entries | 2 |
| Expected total entries | 12 |
The two record entries are tracks.csv and manifest.json. Count audio files separately from them. In the downloadable partial-batch CSV, the inclusion flag identifies which ready rows belong in the selected archive. That dataset demonstrates bookkeeping with synthetic metadata; it is not a download of twenty songs. The separate format benchmark did build and reopen two real archives from self-generated audio signals.
A ZIP with twelve total entries in this example can be correct even though the input had twenty rows. A ZIP containing only the two records would not match the intended ten-track selection. That is a reason to preserve the archive and investigate, rather than editing the record to hide the difference.
Building and saving are separate actions
- After preparation, confirm at least one ready track is checked.
- Choose the filename style you want, then use Build ZIP.
- Wait for the Save ZIP link. Use it to start the browser's actual download.
- Check the browser's download list for a completed file and open its location.
- Extract the archive into an empty folder, then compare audio filenames with the CSV inside that same archive.
Changing selection or naming invalidates an archive that was prepared earlier. Build it again before saving the revised collection. A ZIP already on disk remains the old version; it cannot update itself when you uncheck a track in the browser. Similar names and automatic filename suffixes can make an old download easy to mistake for the latest one.
Do not assume that clicking a link means the download finished. A cancelled save, lack of disk space or a destination prompt can leave no complete archive. The tool cannot inspect your Downloads folder. If the browser shows a partial or failed download, resolve that stage before changing the selection or blaming the CSV.
Recover without losing ready files
If the workspace is still open, save an important ready track individually first. Export the current CSV for a record of the run. Recheck selection, rebuild once, and save under a distinguishable name or into a clean folder. If archive creation itself fails, try a smaller set or individual saves while the prepared results remain available.
If the tab was refreshed or closed, the old browser object used by Save ZIP is gone. A restored list of source links does not restore the archive. Reprepare only the files you still need, and avoid mixing the new export into an unchecked old folder. The ZIP verification guide explains comparing filenames, raw byte counts and hashes.
For support, retain the smallest non-sensitive example of the mismatch: expected selected count, actual audio count, archive size, visible build message and browser download status. Share music or manifests only when you have permission and are comfortable exposing their titles and links. A precise count discrepancy is more useful than the statement that a ZIP is empty.