Create README.md

README.md

@@ -1,13 +1,21 @@
----
-title: Timescodeyt
-emoji: 📉
-colorFrom: pink
-colorTo: green
-sdk: gradio
-sdk_version: 5.45.0
-app_file: app.py
-pinned: false
-short_description: auto timescode
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# YouTube Timecode Generator (Gradio / Hugging Face Spaces)
+
+Generate YouTube-style timecodes from a long audio file.
+
+## Features
+- Auto-detect track starts using silence detection (pydub + ffmpeg)
+- Manual starts (seconds / mm:ss / hh:mm:ss)
+- Optional custom titles (one per line)
+- Exports ready-to-paste timecodes and downloadable `timecodes.txt`
+
+## How to use
+1. Upload your album/playlist audio (mp3/wav/flac).
+2. Choose **Auto** or **Manual**.
+   - Auto: tune **Silence threshold**, **Minimum silence**, **Seek step**.
+   - Manual: enter comma-separated starts.
+3. (Optional) Paste **Titles** (one per line).
+4. Click **Generate Timecodes** → copy result or download `timecodes.txt`.
+
+## Notes
+- This Space uses `ffmpeg` via `packages.txt`.
+- For loud continuous mixes or live recordings, Auto mode may need higher `min_silence` and a less negative `silence_threshold` (e.g., -26 dBFS).