---
tags:
  - coreml
  - audio
  - music-source-separation
  - schism
license: mit
---

# DTTNet (All Stems)

CoreML model for [Schism](https://github.com/zakkeown/Schism) audio separation framework.

Dual Time-Frequency Transformer per-stem models (vocals, drums, bass, other)

## Details

| Property | Value |
|----------|-------|
| Expert IDs | dttnet-vocals, dttnet-drums, dttnet-bass, dttnet-other |
| Tasks | vocalSeparation, stemSeparation |
| Quality | high |
| Est. RTF | 0.4 |
| Peak Memory | 384 MB |

## Usage

Install to your local Schism model store:

```bash
schism models install dttnet
```

Or manually place in `~/.schism/models/dttnet/v1.0.0/`.