1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
|
# producer
One-click audio file mastering. Give it a raw recording; it denoises, enhances, applies a warm voice chain, and delivers a loudness-normalized master.
```bash ./producer <filename>.wav ```
## Requirements
- curl
- ffmpeg
- NVidia GPU (optional)
## The sound
The default `audiobook` profile targets a close-narration master: high-pass at 80 Hz, a gentle mud cut, low-shelf warmth, serial compression (2:1 then 3:1), de-essing, restrained presence/air, breath ducking, then normalization to RMS -20 dB with a -3 dB true-peak ceiling.
`--profile podcast` switches to broadcast loudness: -16 LUFS integrated, -1.5 dBTP ceiling, 48 kHz, slightly brighter EQ. `--profile radio` targets the deep, warm broadcast voice: +3 dB low shelf at 100 Hz, a tighter mud cut, heavy serial compression, dynamic resonance control (`soothe`) that ducks boxy 200-450 Hz and harsh 2.5-6 kHz peaks only while they stick out, and asymmetric tape saturation (`tape`) for the even-harmonic analog sheen.
## Engines
| Stage | Choices | Notes |
| ----------- | ------------------------------------------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `--denoise` | `dfn3` (default), `zipenhancer`, `spectral`, `off` | [DeepFilterNet3](https://github.com/Rikorose/DeepFilterNet): 48 kHz full-band, faithful, fast on CPU, CUDA optional. [ZipEnhancer](https://github.com/gyj1201/zipEnhancer) (ICASSP 2025 SOTA, PESQ 3.69): 16 kHz native, bandwidth is restored after, slightly softer highs. `spectral`: pure-DSP decision-directed Wiener gate over a **file-global noise profile** (measured from the whole file's quiet frames) — deep ~20-35 dB suppression of steady noise and tape hiss in sentence gaps, deterministic, no model download, zero gain wobble. Strength defaults to 0.9: the small dry blend flattens residual denoiser gain wobble; use `--denoise-strength 1.0` for maximum suppression. |
| `--enhance` | `off` (default), `mossformer2`, `resemble` | [MossFormer2_SE_48K](https://github.com/modelscope/ClearerVoice-Studio) (ClearVoice): full-band studio restoration, GPU-strong. [Resemble Enhance](https://github.com/resemble-ai/resemble-enhance): generative restoration for badly damaged audio, runs in an isolated venv; can alter voice timbre, so it is opt-in. |
Engine dependencies (torch ~2.5 GB on CUDA, model weights ~8 MB) are installed lazily on first use into `lib/` (`lib/models`, `lib/venvs`); zipenhancer additionally pulls modelscope plus any helpers its import chain needs at first use. Prefer `--denoise spectral` or `--denoise off` for zero heavy downloads. The DeepFilterNet post filter (`--denoise-pf`, or `pf = true` under `[denoise]` in `config.toml`) adds extra noise reduction but can eat soft speech on clean recordings, so it is off by default.
## Usage
```bash
./producer in.wav [-o out.wav] [--profile audiobook|podcast|radio]
[--denoise dfn3|zipenhancer|spectral|off] [--denoise-strength 0-1] [--denoise-pf] [--enhance off|mossformer2|resemble]
[--enhance-strength 0-1] [--no-dsp] [--no-levelling] [--hpf-hz N] [--mud N] [--warmth N]
[--soothe N] [--compress N] [--tape N] [--deess N] [--presence N] [--air N] [--breath N]
[--target N] [--ceiling N] [--format wav|flac|mp3] [--sample-rateN] [--bit-depth 16|24|32]
[--device auto|cuda|cpu] [--engine-chunk SEC] [--engine-overlap SEC]
[--batch] [--report] [--dry-run] [-v]
```
## Troubleshooting
If the gain wobbles, it's likely the denoiser. Try `--denoise-strength 0.5` or `--denoise zipenhancer`.
## License
0BSD
|
