Transcription Workflows and Costs
Bookshift's transcription service converts your audio dictation into polished, ready-to-use text—optimized for fiction writers.
What Is Transcription
The transcription service is designed specifically for fiction authors who:
- Dictate their stories instead of typing
- Record brainstorming sessions or plot ideas
- Want to capture ideas while walking, driving, or away from the keyboard
What Makes It Different
Unlike generic transcription tools, Bookshift's service:
- Understands fiction - Recognizes dialogue tags, character names, narrative flow
- Applies proofreading - Multi-pass cleanup for clean output
- Handles explicit content - No need to self-censor your dictation
- Offers enhancement presets - Polish for dialogue, description, pacing
Supported Audio Formats
| Format | Support |
|---|---|
| MP3 | ✅ Supported |
| WAV | ✅ Supported |
| M4A | ✅ Supported |
| AAC | ✅ Supported |
| OGG | ✅ Supported |
| FLAC | ✅ Supported |
| AIFF | ✅ Supported |
Transcription Modes
Oneshot Mode (Default)
Best for: Standard dictation sessions under 30 minutes
How it works:
1. Download audio file
2. Transcribe using Whisper API
3. Apply proofreading pass 1
4. Apply proofreading pass 2
5. Return completed transcript
Processing time: Typically 2-5 minutes
Chunked Mode
Best for: Longer recordings requiring higher quality
How it works:
1. Preprocess audio (normalize, remove silence, speedup)
2. Split into manageable chunks
3. Transcribe chunks in parallel
4. Clean up each chunk
5. Aggregate and apply final cleanup
Processing time: Longer but higher quality
Chunked Mode Presets
| Preset | Speedup | Chunk Size | Best For |
|---|---|---|---|
| conservative | 1.35x | 5-min | Careful processing |
| aggressive | 2.0x | 3-min | Faster turnaround |
| hyper | 3.0x | 2.5-min | Maximum speed |
| no_speedup | 1.0x | 4-min | Original pacing |
| chunk_only | none | varies | Custom processing |
Costs
| Service | Credit Cost |
|---|---|
| Audio Transcription | Dynamic pricing shown before submission |
| Rough Notes to Full Outline (coming soon) | Coming soon |
Tip
Bookshift transcription is cheaper than any comparable service, and you can see the estimate before you submit.
Additional Features
Projects & Glossaries
Define context for your transcription:
- Project name - Helps the AI understand your work
- Character names - Ensures correct spelling
- Place names - Consistent throughout transcript
- Glossary terms - Custom words with preferred spellings
Enhancement Presets
After transcription, apply targeted polish:
| Preset | What It Does |
|---|---|
| dialogue | Smooths dialogue tags and speech patterns |
| description | Enhances descriptive passages |
| saidisms | Varies dialogue tags beyond "said" |
| pacing | Adjusts sentence rhythm |
Rough Notes to Full Outline (Coming Soon)
This workflow is not publicly available yet. It will convert recorded rough notes into a structured outline when released.
Output
Transcription delivers:
- Clean text - Ready to paste into your manuscript
- DOCX export - Optional Word document format
- Raw transcript - Unedited version for reference
Limitations
- Default mode is oneshot (chunked must be selected)
- Very long files may require chunked mode
- Background noise can affect accuracy
- No refunds on transcription failure (but issues are rare)
Best Practices
- Record in quiet environments - Reduces errors
- Speak clearly - Especially for character names
- Use your glossary - Define unusual terms beforehand
- Review and edit - Always check the output before using
- Chunk long sessions - Better quality for recordings over 30 minutes
Menu