diff options
Diffstat (limited to 'README.md')
| -rw-r--r-- | README.md | 51 |
1 files changed, 13 insertions, 38 deletions
@@ -9,8 +9,9 @@ Original project: [https://github.com/WhiskeyCoder/Qwen3-Audiobook-Converter](ht The converter sends text extracted from your books to a locally running Qwen3-TTS server and assembles the returned audio into a single audiobook file. -- Supported input: `.txt`, `.pdf`, `.epub` +- Input: `.txt`, `.pdf`, or `.epub` - Output: `.mp3` or `.m4b` +- Output a single mp3 or one per chapter - Two voice modes: - Custom voice: pre-built speakers - Voice clone: clone a voice from a `.wav` reference audio file @@ -89,47 +90,21 @@ CUSTOM_VOICE_INSTRUCT = "Speak naturally and clearly, as if reading a dramatic b python audiobook_converter.py --voice-clone --voice-sample path/to/reference.wav ``` -The reference .wav should be ~10-15 seconds with a minimum of 3 seconds and maximum of 60 seconds. Longer is not better. ~15 seconds is ideal. +The reference `.wav` should be ~10-15 seconds (3 second minimum, 60 second maximum; ~15 seconds is ideal). -Whisper will be used automatically to transcribe the reference audio (`faster_whisper` or `whisper`). If no Whisper backend is installed, it falls back to x-vector-only cloning. +Whisper (`faster_whisper` or `whisper`) is used automatically to transcribe the reference audio; without a Whisper backend it falls back to x-vector-only cloning. Override with `--voice-sample-text "..."` or skip transcription with `--no-transcription`. -To skip automatic transcription explicitly, pass `--no-transcription`. This should be worse, but in my experience may give a preferable flatter tone to certain voices. +### Options -You can override whisper by passing your own transcription with `--voice-sample-text "What the reference audio says"` +| Flag | Description | +| --------------------------- | ----------------------------------------------------------------------------------------- | +| `--speed <n>` | Playback speed, pitch-preserving (`1.0` = normal). A normal-speed copy is also output. | +| `--format {mp3,m4b}` | Output format (default `mp3`). `m4b` uses AAC audio and is always a single file. | +| `--single-file` | Merge all chapters into a single mp3 (default: one mp3 per chapter). | +| `--voice-sample-text "..."` | Transcript of the reference audio (voice clone only). | +| `--no-transcription` | Skip auto-transcription of the reference audio (voice clone only). | -### Playback speed - -Adjust the speed of the final audiobook without changing pitch (uses ffmpeg `atempo` before encoding). The normal-speed audiobook file is also preserved in the output directory. - -```bash -python audiobook_converter.py --speed 0.9 -``` - -### Output format (mp3 / m4b) - -Use `--format` to choose the output container. The default is `mp3`; `m4b` uses AAC audio (ffmpeg `aac`). - -```bash -python audiobook_converter.py --format m4b -``` - -### Chapters (EPUB) - -Books with chapters (e.g. EPUB) are converted to **one file per chapter** by default. Files are named `output/<Book>_01_<Chapter>.mp3`, `output/<Book>_02_<Chapter>.mp3`, and so on. - -To merge all chapters into a single file instead, pass `--single-file`: - -```bash -python audiobook_converter.py --single-file -``` - -When the source has chapters and the output is a single `m4b`, chapter markers are embedded so listeners can skip between chapters: - -```bash -python audiobook_converter.py --format m4b --single-file -``` - -TXT and PDF files have no chapter structure and always produce a single file. +Books with chapters (e.g. EPUB) are converted to **one mp3 per chapter** by default, named `output/<Book>_01_<Chapter>.mp3`, `output/<Book>_02_<Chapter>.mp3`, etc.; use `--single-file` to merge them. `m4b` output is always a single file with chapter markers embedded so listeners can skip between chapters. TXT and PDF files have no chapter structure and always produce a single file. The `chunks/` folder is scratch space for the current book only — it is emptied before and after every conversion, so an interrupted run never affects the next one. |