How cosmicrealm/voicestudio works
Voice Studio is a local macOS voice-creation workspace. Users configure the local runtime and models, then create audio through built-in, cloned, designed, or multi-role workflows; they can review playback and export results. Script Studio normalizes role-tagged text and prepares synthesis input, while the app view model coordinates generation and persistent workspace state. Separate product branches cover DeepSeek-assisted rewriting, voice-library management, and audio merging. The README documents these capabilities; some implementations are unsampled, so only documented links are shown for them.
Main entry points: StudioViewModel.swift Studio coordinator · ContentView.swift Workspace navigation · ScriptStudioView.swift Script Studio · BackendJSONClient.swift Backend client
Components 23
Desktop application
- App shell
Sources/MacQwenVoice/MacQwenVoiceApp.swift - Workspace navigation
Sources/MacQwenVoice/ContentView.swift - Studio coordinator
Sources/MacQwenVoice/StudioViewModel.swift
Voice creation
- Premium voices
Sources/MacQwenVoice/Views/PremiumVoicesView.swift - Voice cloning
Sources/MacQwenVoice/Views/ClonedVoicesView.swift - Voice design
Sources/MacQwenVoice/Views/VoiceDesignView.swift - Saved voice library
Sources/MacQwenVoiceCore/Domain.swift
Script and dialogue
- Script Studio
Sources/MacQwenVoice/Views/ScriptStudioView.swift - Workflow and prompts
Sources/MacQwenVoiceCore/ScriptStudioWorkflow.swift - Role text parsing
Sources/MacQwenVoiceCore/SpeakerTaggedTextNormalizer.swift - Text segmentation
Sources/MacQwenVoiceCore/TextSegmenter.swift - Dialogue rewrite
Sources/MacQwenVoice/Views/ScriptRewriteView.swift - Generation results
Sources/MacQwenVoice/Views/GenerationResultsPanel.swift - Audio playback
Sources/MacQwenVoice/Views/AudioPlayButton.swift - Audio merge tools
Sources/MacQwenVoice/Views/VoiceToolsView.swift
Runtime and workspace
- Runtime management
Sources/MacQwenVoice/Views/ModelRuntimeView.swift - Backend client
Sources/MacQwenVoice/BackendJSONClient.swift - Local inference backend
backend/server.py - Runtime environment
Sources/MacQwenVoice/VoiceStudioRuntimeEnvironment.swift - Workspace database
Sources/MacQwenVoiceCore/AppDatabase.swift - Workspace paths
Sources/MacQwenVoiceCore/AppPaths.swift
Other components
- User
- DeepSeek service
Connections 30
- User → App shell: opens
README.md - App shell → Workspace navigation: shows
Sources/MacQwenVoice/MacQwenVoiceApp.swift - Workspace navigation → Studio coordinator: uses
Sources/MacQwenVoice/ContentView.swift - Workspace navigation → Script Studio: routes to
Sources/MacQwenVoice/ContentView.swift - Workspace navigation → Premium voices: routes to
Sources/MacQwenVoice/ContentView.swift - Workspace navigation → Voice cloning: routes to
Sources/MacQwenVoice/ContentView.swift - Workspace navigation → Voice design: routes to
Sources/MacQwenVoice/ContentView.swift - Workspace navigation → Dialogue rewrite: routes to
Sources/MacQwenVoice/ContentView.swift - Workspace navigation → Runtime management: routes to
Sources/MacQwenVoice/ContentView.swift - Workspace navigation → Audio merge tools: routes to
Sources/MacQwenVoice/ContentView.swift - Premium voices → Studio coordinator: requests generation
Sources/MacQwenVoice/Views/PremiumVoicesView.swift - Voice cloning → Studio coordinator: manages voices
Sources/MacQwenVoice/Views/ClonedVoicesView.swift - Voice design → Studio coordinator: requests design
Sources/MacQwenVoice/Views/VoiceDesignView.swift - Script Studio → Studio coordinator: submits script
Sources/MacQwenVoice/Views/ScriptStudioView.swift - Script Studio → Workflow and prompts: configures
Sources/MacQwenVoice/Views/ScriptStudioView.swift - Workflow and prompts → Role text parsing: supports role scripts
- Role text parsing → Text segmentation: segments text
Sources/MacQwenVoiceCore/SpeakerTaggedTextNormalizer.swift - Dialogue rewrite → Studio coordinator: requests rewrite
Sources/MacQwenVoice/Views/ScriptRewriteView.swift - Studio coordinator → DeepSeek service: sends rewrite request
README.md - Studio coordinator → Backend client: calls backend
Sources/MacQwenVoice/StudioViewModel.swift - Backend client → Local inference backend: launches script
Sources/MacQwenVoice/BackendJSONClient.swift - Backend client → Runtime environment: configures process
Sources/MacQwenVoice/BackendJSONClient.swift - Runtime management → Studio coordinator: manages models
Sources/MacQwenVoice/Views/ModelRuntimeView.swift - Studio coordinator → Workspace database: persists workspace
Sources/MacQwenVoice/StudioViewModel.swift - Studio coordinator → Workspace paths: uses paths
Sources/MacQwenVoice/StudioViewModel.swift - Studio coordinator → Saved voice library: stores voice profiles
- Studio coordinator → Generation results: provides results
Sources/MacQwenVoice/Views/GenerationResultsPanel.swift - Generation results → Audio playback: offers playback
Sources/MacQwenVoice/Views/GenerationResultsPanel.swift - Studio coordinator → Audio merge tools: merges audio
README.md - Studio coordinator → Audio playback: provides generated audio
Sources/MacQwenVoice/Views/AudioPlayButton.swift