Dropout Detector (Beta)
Dropout Detector is a desktop tool for scanning raw production WAV files after ingest and finding mic dropouts that are painful to spot by hand. It is aimed at the nightmare version of audio QC: a loose connection or signal collapse hidden somewhere inside a long shoot day, visible only as a tiny dip in a waveform if you know exactly where to look.
Point it at a folder of Zoom F8n WAVs, run the scan, and export a timecoded CSV showing which file, channel, track name, timecode, and duration need attention in post.
DropoutDetector-Mac.dmg
DropoutDetector-Setup-Windows.exe
If this saves you time on a job, you can support future Hotshot tools here.
Why this exists
A fast ingest check for a failure mode that is easy to miss until it is expensive.
A mic dropout is not always obvious. On the Zoom F8n issue this tool was built around, the affected channel does not become clean digital silence. Instead, the signal collapses to an extremely low level before recovering. That can be almost invisible in a long waveform view, and standard silence detection can miss it completely.
Dropout Detector is designed to sit near the start of post. After copying the shoot-day audio, run it against the original WAV folder before editorial is deep into the cut. If a lav dropped for half a second, the report gives you the exact channel and timecode early enough to check alternate mics, choose another take, request ADR, or flag the issue before the problem disappears into the timeline.
The tool scans recursively, so you can choose the top-level folder for a session or shoot day rather than opening each file manually. Each result includes the file name, channel number, iXML track name, start timecode, end timecode, and duration.
How it detects dropouts
The detector combines amplitude checks with context from the mix channels so quiet gaps between takes do not turn into noise.
Amplitude analysis
The app analyses active channels in 20 millisecond windows. A window is treated as dead when its peak amplitude falls at or below 0.1% of full scale. That threshold is tuned for the F8n dropout floor, where the failed signal is extremely low but not absolute zero.
Mix reference gating
Channels 1 and 2 are treated as the F8n stereo mix reference. A dropout is only reported when the target input channel is dead while the mix reference is active. This helps avoid false positives during pauses, room tone gaps, or moments where nothing was actually being captured.
Unused channel filtering
If a channel is dead for more than 95% of its analysis windows, the app treats it as unused and skips it. That keeps an unpatched or muted input from flooding the report.
Minimum duration
A dropout must last at least 0.3 seconds before it appears in the results. Shorter blips are ignored so the report stays focused on events likely to matter in the edit.
| Parameter | Default | Purpose |
|---|---|---|
| Window | 20 ms | Chunk size used for audio analysis |
| Dead threshold | 0.001 | Peak level at or below this is treated as dead |
| Mix reference | Ch 1 and Ch 2 | Confirms that recording activity is present |
| Minimum duration | 0.3 seconds | Filters out short artefacts |
| Dead channel cutoff | Over 95% | Skips channels that appear unused |
| Default FPS | 25 | Formats report timecodes as HH:MM:SS:FF |
Using the tool
Run a scan, review the table, then export a CSV for Resolve, Excel, Numbers, or your post notes.
- Open Dropout Detector.
- Click Browse and choose the folder containing your raw WAV files.
- Leave FPS at 25 unless the project was shot at a different frame rate.
- Click Run Scan.
- Watch the scan log for clean files, files with dropouts, and skipped unused channels.
- Sort the results by start timecode, duration, track name, file, or channel.
- Click Export CSV to save the full report.
What the CSV contains
| Field | Why it matters |
|---|---|
| File name | Identifies the source WAV containing the dropout |
| Channel | Shows which channel failed |
| Track name | Uses iXML metadata to show the named mic, such as a lav or boom |
| Start TC / End TC | Lets you jump directly to the event in Resolve or another NLE |
| Duration | Helps sort tiny checks from longer audible problems |
Installation
Install the app for your platform, then launch it from Applications, Spotlight, the Start menu, or your desktop shortcut.
Mac
- Download
DropoutDetector-Mac.dmg. - Open the DMG.
- Drag Dropout Detector to your Applications folder.
- Launch it from Applications or Spotlight.
Windows
- Download
DropoutDetector-Setup-Windows.exe. - Run the installer and follow the prompts.
- Launch Dropout Detector from the Start menu or desktop shortcut.
Troubleshooting
A few things to check if the beta gives unexpected results.
No dropouts found, but you know there were some
Check that you are scanning the original F8n WAV files, not an exported or processed copy. The detector is calibrated for the raw recorder output.
Too many false positives
Confirm the files are standard F8n polyphonic WAVs with the stereo mix on channels 1 and 2. The beta uses those channels as its activity reference.
A channel you expected is skipped
The app skips channels that appear dead for more than 95% of the scan. If the channel should have been active, check whether the mic was muted, unplugged, or not armed at the recorder.
Timecodes do not match Resolve
Make sure the FPS in Dropout Detector matches the Resolve project. The app reads absolute BWF time reference metadata and formats it at the FPS you set.
The beta does not behave on your files
Please email ash@hotshotmedia.co.nz with what happened, what recorder or file type you used, and whether the issue was a missed dropout, a false positive, or an install problem.