# anjok07/ultimatevocalremovergui architecture

> Architecture diagram of the GitHub repository anjok07/ultimatevocalremovergui, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-09). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/anjok07/ultimatevocalremovergui
- Repository: https://github.com/anjok07/ultimatevocalremovergui

## Overview

UVR is a desktop audio source-separation application: users select audio and separation options, run a chosen model or ensemble, then receive separated stems and can apply audio adjustments. The graph preserves the documented separation product and the sampled GUI-to-engine and engine-to-model relationships. Several model families and post-processing capabilities are identified in the README or source index, but their detailed runtime wiring was not sampled; those links are shown only where documented or evidenced by the index.

Main entry points: [`separate.py`](https://github.com/anjok07/ultimatevocalremovergui/blob/master/separate.py) (Separation orchestration), [`UVR.py`](https://github.com/anjok07/ultimatevocalremovergui/blob/master/UVR.py) (UVR interface), [`demucs/apply.py`](https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/apply.py) (Demucs inference), [`demucs/hdemucs.py`](https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/hdemucs.py) (Demucs architectures)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_interface["Desktop interface"]
  node_gui["UVR interface<br/>[UVR.py]"]
  node_dragdrop["Drag and drop<br/>[TkinterDnD.py]"]
  node_errors["Error messages<br/>[error_handling.py]"]
end

subgraph group_engine["Separation engine"]
  node_separation["Separation orchestration<br/>[separate.py]"]
  node_demucs_apply["Demucs inference<br/>[apply.py]"]
  node_demucs_states["Model state loading<br/>[states.py]"]
  node_demucs_utils["Chunked inference<br/>[utils.py]"]
end

subgraph group_models["Model families"]
  node_demucs_arch["Demucs architectures<br/>[hdemucs.py]"]
  node_demucs_other["Demucs variants<br/>[tasnet.py]"]
  node_demucs_transformer["Demucs transformer<br/>[transformer.py]"]
  node_mdx["MDX-Net<br/>[mdxnet.py]"]
  node_vr["VR network<br/>[nets.py]"]
  node_tfctdf["TFC-TDF network<br/>[tfc_tdf_v3.py]"]
  node_model_files[("Model files")]
end

subgraph group_audio["Audio processing"]
  node_spec["Spectrogram conversion<br/>[spec_utils.py]"]
  node_results["Stem ensembling<br/>[results.py]"]
  node_audio_tools["Pitch and time shift<br/>[UVR.py]"]
end

node_user(("Audio user"))

node_user -->|"selects audio"| node_gui
node_gui -->|"starts separation"| node_separation
node_gui -->|"imports drag-drop"| node_dragdrop
node_gui -->|"uses error handling"| node_errors
node_separation -->|"imports inference"| node_demucs_apply
node_separation -->|"loads architecture"| node_demucs_arch
node_separation -.->|"imports model loading"| node_demucs_states
node_demucs_apply -->|"imports architecture"| node_demucs_arch
node_demucs_apply -->|"uses helpers"| node_demucs_utils
node_demucs_arch -->|"imports helpers"| node_demucs_utils
node_separation -->|"imports model variant"| node_demucs_other
node_separation -.->|"selects model variant"| node_demucs_transformer
node_separation -->|"imports model runner"| node_mdx
node_separation -->|"imports network"| node_vr
node_separation -->|"imports network"| node_tfctdf
node_separation -->|"imports audio conversion"| node_spec
node_separation -.->|"loads model assets"| node_model_files
node_gui -->|"applies time or pitch shift"| node_audio_tools
node_audio_tools -->|"calls augmentation"| node_spec
node_user -.->|"chooses model"| node_model_files
node_separation -.->|"ensembles outputs"| node_results

click node_gui "https://github.com/anjok07/ultimatevocalremovergui/blob/master/UVR.py"
click node_dragdrop "https://github.com/anjok07/ultimatevocalremovergui/blob/master/gui_data/tkinterdnd2/TkinterDnD.py"
click node_errors "https://github.com/anjok07/ultimatevocalremovergui/blob/master/gui_data/error_handling.py"
click node_separation "https://github.com/anjok07/ultimatevocalremovergui/blob/master/separate.py"
click node_demucs_apply "https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/apply.py"
click node_demucs_states "https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/states.py"
click node_demucs_utils "https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/utils.py"
click node_demucs_arch "https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/hdemucs.py"
click node_demucs_other "https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/tasnet.py"
click node_demucs_transformer "https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/transformer.py"
click node_mdx "https://github.com/anjok07/ultimatevocalremovergui/blob/master/lib_v5/mdxnet.py"
click node_vr "https://github.com/anjok07/ultimatevocalremovergui/blob/master/lib_v5/vr_network/nets.py"
click node_tfctdf "https://github.com/anjok07/ultimatevocalremovergui/blob/master/lib_v5/tfc_tdf_v3.py"
click node_model_files "https://github.com/anjok07/ultimatevocalremovergui/tree/master/models"
click node_spec "https://github.com/anjok07/ultimatevocalremovergui/blob/master/lib_v5/spec_utils.py"
click node_results "https://github.com/anjok07/ultimatevocalremovergui/blob/master/lib_v5/results.py"
click node_audio_tools "https://github.com/anjok07/ultimatevocalremovergui/blob/master/UVR.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_gui,node_dragdrop,node_errors,node_user toneBlue
class node_separation,node_demucs_apply,node_demucs_states,node_demucs_utils toneAmber
class node_demucs_arch,node_demucs_other,node_demucs_transformer,node_mdx,node_vr,node_tfctdf,node_model_files toneMint
class node_spec,node_results,node_audio_tools toneRose
```

## Components

### Desktop interface

- [UVR interface](https://github.com/anjok07/ultimatevocalremovergui/blob/master/UVR.py): `UVR.py`
- [Drag and drop](https://github.com/anjok07/ultimatevocalremovergui/blob/master/gui_data/tkinterdnd2/TkinterDnD.py): `gui_data/tkinterdnd2/TkinterDnD.py`
- [Error messages](https://github.com/anjok07/ultimatevocalremovergui/blob/master/gui_data/error_handling.py): `gui_data/error_handling.py`

### Separation engine

- [Separation orchestration](https://github.com/anjok07/ultimatevocalremovergui/blob/master/separate.py): `separate.py`
- [Demucs inference](https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/apply.py): `demucs/apply.py`
- [Model state loading](https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/states.py): `demucs/states.py`
- [Chunked inference](https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/utils.py): `demucs/utils.py`

### Model families

- [Demucs architectures](https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/hdemucs.py): `demucs/hdemucs.py`
- [Demucs variants](https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/tasnet.py): `demucs/tasnet.py`
- [Demucs transformer](https://github.com/anjok07/ultimatevocalremovergui/blob/master/demucs/transformer.py): `demucs/transformer.py`
- [MDX-Net](https://github.com/anjok07/ultimatevocalremovergui/blob/master/lib_v5/mdxnet.py): `lib_v5/mdxnet.py`
- [VR network](https://github.com/anjok07/ultimatevocalremovergui/blob/master/lib_v5/vr_network/nets.py): `lib_v5/vr_network/nets.py`
- [TFC-TDF network](https://github.com/anjok07/ultimatevocalremovergui/blob/master/lib_v5/tfc_tdf_v3.py): `lib_v5/tfc_tdf_v3.py`
- [Model files](https://github.com/anjok07/ultimatevocalremovergui/tree/master/models): `models`

### Audio processing

- [Spectrogram conversion](https://github.com/anjok07/ultimatevocalremovergui/blob/master/lib_v5/spec_utils.py): `lib_v5/spec_utils.py`
- [Stem ensembling](https://github.com/anjok07/ultimatevocalremovergui/blob/master/lib_v5/results.py): `lib_v5/results.py`
- [Pitch and time shift](https://github.com/anjok07/ultimatevocalremovergui/blob/master/UVR.py): `UVR.py`

### Other components

- **Audio user**

## Connections

- Audio user → UVR interface: selects audio (evidence: `README.md`)
- UVR interface → Separation orchestration: starts separation (evidence: `UVR.py`)
- UVR interface → Drag and drop: imports drag-drop (evidence: `UVR.py`)
- UVR interface → Error messages: uses error handling (evidence: `UVR.py`)
- Separation orchestration → Demucs inference: imports inference (evidence: `separate.py`)
- Separation orchestration → Demucs architectures: loads architecture (evidence: `separate.py`)
- Separation orchestration → Model state loading: imports model loading
- Demucs inference → Demucs architectures: imports architecture (evidence: `demucs/apply.py`)
- Demucs inference → Chunked inference: uses helpers (evidence: `demucs/apply.py`)
- Demucs architectures → Chunked inference: imports helpers (evidence: `demucs/hdemucs.py`)
- Separation orchestration → Demucs variants: imports model variant (evidence: `separate.py`)
- Separation orchestration → Demucs transformer: selects model variant
- Separation orchestration → MDX-Net: imports model runner (evidence: `separate.py`)
- Separation orchestration → VR network: imports network (evidence: `separate.py`)
- Separation orchestration → TFC-TDF network: imports network (evidence: `separate.py`)
- Separation orchestration → Spectrogram conversion: imports audio conversion (evidence: `separate.py`)
- Separation orchestration → Model files: loads model assets
- UVR interface → Pitch and time shift: applies time or pitch shift (evidence: `UVR.py`)
- Pitch and time shift → Spectrogram conversion: calls augmentation (evidence: `UVR.py`)
- Audio user → Model files: chooses model
- Separation orchestration → Stem ensembling: ensembles outputs
