Settings Overview
Loki Studio has three settings areas, each with a different scope:
| Settings Tab |
Norse Name |
Scope |
| Application Settings |
Yggdrasil |
Global - affects all profiles |
| Power User Settings |
Ratatoskr |
Global - YouTube and processing defaults |
| Profile Settings |
Norns |
Per-profile - thumbnails, upload, personalities |
Application Settings (Yggdrasil)
Global settings for API keys, LLM providers, transcription, and directories.
Directories Tab
Base Video Directory
- Path: Parent folder where all game/series subdirectories are created
- Browse: Open folder picker
- Create: Create directory if it doesn't exist
Example: E:\Videos\Gaming\ - profiles will create subdirectories here
Data Directory (AI Models)
- Path: Where AI models (Whisper, LLM) are stored
- Default: C:\Users\[username]\AppData\Local\LokiStudio
- Reset: Restore default location
LLM Providers Tab
See the dedicated AI Setup Guide for detailed configuration.
LLM Mode
- Grimnir (Built-in LLM): Pro feature - runs locally
- Bring Your Own (Third-Party): Use external providers
Grimnir Settings
- Model: Select from downloaded models
- Max Tokens: Token budget for reasoning (1K-200K)
- Download Models: Install new LLM models
BYO Provider Settings
- Provider: Ollama, LM Studio, OpenAI, Claude, ChatGPT
- Base URL: API endpoint (for local providers)
- API Key: Authentication key (for cloud providers)
- Model: Model name/ID
Transcription Tab
See the dedicated Transcription Guide for detailed settings.
Basic Settings
- Model: Whisper model size (tiny to large-v3-turbo)
- Device: auto, cuda, or cpu
- Compute Type: float16, int8_float16, int8, float32
Quality Settings
- Beam Size: 1-10 (default: 10) - higher = more accurate
- Patience: 1.0-3.0 (default: 2.0) - alternative consideration time
- Compression Threshold: Default 3.0
- No-Speech Threshold: Default 0.40
- Log Prob Threshold: Default -1.0
Voice Activity Detection
- VAD Type: Auto, Off, Energy, Silero, TEN
- VAD Threshold: 0.0-1.0 (default: 0.35)
Output Options
- Generate Subtitles (.srt): Create subtitle files
- Word-level Timestamps: Enable for karaoke captions
- Initial Prompt: Context hint for the model
Performance
- Batch Size: Segments per batch (default: 24)
- Intra Threads: Threads per operation (0 = auto)
- Inter Threads: Parallel operations (default: 1)
Captions Tab
Display Settings
- Style: Sliding Window, One Word Pop, Full Segment
- Max Words: Words shown at once (default: 6)
- Font Size: Caption text size in pixels (default: 32)
- Max Width: Maximum caption width as % (default: 80)
Colors
- Highlight: Currently spoken word (default: yellow)
- Pending: Upcoming words (default: gray)
- Spoken: Already spoken words (default: light gray)
Position & Timing
- Position: Bottom, Top, Center, corners
- Padding: Distance from edge in pixels (default: 80)
- Linger Time: How long words stay after spoken (default: 0.5s)
- Show Background: Semi-transparent background behind text
- CC On By Default: Enable captions automatically
Advanced Tab
LLM Parameters
- Temperature: Creativity level (default: 0.70)
- Max Tokens: Output length limit (default: 2,000)
- Timeout: Request timeout in seconds (default: 60)
Logging & Error Handling
- Enable Logging: Write logs to file
- Log Level: DEBUG, INFO, WARNING, ERROR, CRITICAL
- Retry Attempts: Retries on failure (default: 3)
- Retry Backoff: Seconds between retries (default: 5)
Performance
- Waveform Sample Rate: 16000 Hz recommended
- Waveform Quality: 50% balanced
Workflow
- Enable File Variant System: Support for video variants
Power User Settings (Ratatoskr)
Advanced settings for fine-grained control. Defaults are optimized for most users.
YouTube Upload Defaults
- Allow embedding: Let other sites embed your videos (default: on)
- Show view count: Display view count publicly (default: on)
- License: Standard YouTube License or Creative Commons
- Audio Language: Primary audio language (e.g., English US)
- AI-generated content: Disclose AI usage
- Paid Promotion: No paid promotion / Has paid promotion
Video Processing
- Upload Chunk Size: Size in MB for chunked uploads
- Max Retry Attempts: Retries on upload failure (default: 3)
- Retry Delay: Seconds between retries (default: 5)
- Upload Timeout: Minutes before timeout (default: 60)
- Verify upload: Confirm upload completed (default: on)
- Original filename fallback: Use filename as title if no metadata
Caption & Subtitle Settings
- Caption Format: SRT (SubRip), VTT, SBV
- Caption Language: Language code for captions
- Auto-sync timing: Automatically adjust caption timing
- Mark as CC: Mark as Closed Captions for accessibility
Notifications & Alerts
- Desktop notifications: Show system notifications (default: on)
- Sound on complete: Play sound when upload finishes
- Quota warnings: Alert when approaching YouTube quota (default: on)
- Quota threshold: Warning threshold percentage (default: 80%)
Profile Settings (Norns)
Per-profile settings for thumbnails, upload, and content.
Individual Configs Tab
Each profile has its own configuration:
Thumbnail Customization
- Title Color: Main text color
- Subtitle Color: Secondary text color
- Outline Color: Text outline/stroke color
- Font: Font family (default: Impact)
- Title Font Scale: Title size multiplier (default: 4.0)
- Subtitle Font Scale: Subtitle size multiplier (default: 3.0)
- Logo Max Size: Maximum logo dimension in pixels (default: 400)
- Logo Position: top-right, center, corners, etc.
- Title/Subtitle Text: Default text (empty = auto-generate)
Flair Settings
- Use Flair: Enable flair overlay
- Flair File: Image file path
- Flair Position: Placement (bottom-left, etc.)
- Flair Max Size: Maximum dimension (default: 600)
Frame Selection
- Timestamp Range: Min/Max seconds for frame extraction
- Frame Selection: Random (Fast) or Brightest
- Brightness Interval: Frames to skip when finding brightest
YouTube Upload Settings
- Playlist Name: Default playlist for this profile
- Days Between: Scheduling interval
- Release Time: Default publish time (HH:MM)
- Privacy Status: public, unlisted, private
- Made for Kids: COPPA compliance setting
- YouTube Category: Gaming, Education, etc.
- Description Header: Text added to start of descriptions
- Description Footer: Text added to end of descriptions
Default Settings Tab
Configure defaults for new profiles:
- Save Defaults: Save current settings as defaults
- Apply to All: Apply these defaults to all existing profiles
Profile Management
- Add New Config: Create a new profile
- Duplicate: Clone an existing profile
- Delete: Remove a profile
- Refresh List: Reload profile list