Set It and Forget It: Automate Transcription with Watch Folders

What a Watch Folder Does

A watch folder monitors a directory on your Mac. Drop an audio or video file into it, and Audio Note automatically picks it up, transcribes it, and saves the result. No clicking, no dragging, no manual start.

This sounds simple but it's surprisingly useful once it becomes part of your workflow. You stop thinking about transcription as a separate step and start treating it as something that just happens.

Setting Up

1. In Audio Note, go to the Watch Folder tab in the sidebar
2. Click Add Watch Folder
3. Pick a directory on your Mac
4. Choose the transcription model the folder should use
5. Click Start Watching

After setup, Audio Note watches that folder continuously. Any new audio or video file that appears in it gets added to the transcription queue.

You can set up multiple watch folders with different settings. For example, one folder can use a lighter model for quick turnaround, while another uses a larger model for important material.

Workflow Examples

Interview Workflow

1. Record an interview on your phone or voice recorder
2. Save the file to a cloud folder (iCloud, Dropbox, etc.) that syncs to your Mac
3. The file syncs to your watch folder
4. Audio Note transcribes it automatically

By the time you sit down at your Mac after the interview, the transcript is already waiting.

Podcast Production

1. Export your raw recording to the watch folder
2. Audio Note transcribes it automatically
3. Use the transcript to write show notes
4. Export as SRT to generate captions for the video version

Research and Archiving

1. Collect conference recordings, lecture captures, and interview files
2. Drop them into a watch folder
3. Walk away — Audio Note processes everything in the queue
4. Come back to searchable transcripts of all your material

Queue and Progress

The watch folder has its own queue separate from manual transcriptions. You can see which files are pending, which are in progress, and which are done. Completed transcripts appear in your main workspace alongside manually started ones.

If a transcription fails (corrupted file, unsupported format, etc.), it's marked in the queue with the error reason. You can retry or remove it.

Settings Worth Adjusting

Model Selection Per Folder

Each watch folder can use a different model. Clean recordings can use a lighter model; important or difficult recordings can use a larger one. Test with a sample file before standardizing.

Processing Priority

You can set processing priority in the app settings: Balanced (default), Prioritize Watched (watch folders jump the queue), or Prioritize Manual (manual transcriptions take priority).

File Cleanup

By default, completed files stay in the watch folder. You can enable auto-cleanup to move processed files to a separate "Processed" subfolder or delete them after confirmation.

Limitations to Know

• The watch folder detects new files, not modifications to existing ones. If you replace a file in the folder, rename it to trigger re-processing
• Network drives can be flaky. Use a local directory for reliable detection
• Processing speed depends on model, hardware, and audio length; queues full of long recordings require patience

Automation Checklist

Check	Recommendation
File source	Make sure recordings are fully synced locally before transcription starts
Naming convention	Include date, project, and speaker or topic for easier search
Model choice	Validate each folder's model with a sample file
Failure handling	Review queue errors so damaged files do not block the workflow
Privacy boundary	Check permissions on shared folders and cloud-sync directories