BDK
Assign speakers to subtitle lines. Review faster.
Turn dubbing prep
from line-by-line cleanup into batch review
BDK | Dubbing Prep for Film & Drama Localization
Upload video files, SRT subtitles, and character sheets. BDK helps detect speakers, assign subtitle lines to character roles, add audio cue notes, and calculate line counts so your team can deliver character-wise dubbing scripts faster.
- Free trial minutes included
- Upload files to get started
- Built for team review
Before dubbing starts, assigning subtitle lines to roles and preparing the script often take the most time
BDK automates the preparation work so your team can focus on higher-value review and dubbing decisions.
Import Materials
Automated Processing
Final Delivery
Less manual prep. Faster script delivery.
From line-by-line cleanup to batch preparation
Stop manually copying and formatting every line. Import multiple file types and let the system prepare the script in batches.
From single-line handling to full-series processing
Handle full multi-episode projects, from short dramas to long-form series, episode by episode, with automatic role-to-line matching and clear structure.
From manual alignment to AI-assisted generation
Automatically align music cues, subtitles, and line segments, then fill in dubbing notes with less manual effort.
From manual counting to one-click export
Export Excel scripts and summaries with clear line counts and task status for easier team tracking.
Core Capabilities for Dubbing Annotation
Flexible Media Intake
Batch import videos, subtitles, and character sheets to reduce setup work.
Speaker Recognition and Subtitle Role Assignment
Help identify who is speaking and apply system speaker labels to subtitle lines.
One-click Voice-to-Role Matching
Merge multiple detected speaker profiles into one confirmed character role.
Smart Character Metadata Extraction
Extract names, translated names, gender, and notes from character sheets.
Smart Audio Cue Handling
Add music, background voice, and other audio cue notes to make deliverables more complete.
Character-wise Script Export
Export scripts, character summaries, and workload statistics.
AI identifies speakers across the full series to speed up dubbing annotation
Upload full-series assets
Batch upload videos, subtitles, and character sheets; the system organizes materials by episode.
AI identifies speakers across episodes
The system recognizes speakers across episodes and organizes detected speaker profiles with their matching subtitle segments.
Map speaker profiles to roles
Map detected speaker profiles to confirmed character roles, and merge multiple system speakers into one role when needed.
Complete full-series annotation in one flow
After you confirm the mapping, the system annotates roles across the full series and can reuse the same mapping for newly added episodes.
Take on another series,
without adding another person to the team
BDK reduces the repetitive prep behind short-drama and film localization by handling subtitle role assignment and script preparation first, while annotators stay focused on the decisions that matter.
- Efficiency: For around 100 minutes of footage, a single annotator can potentially process more work per day.
- Collaboration: Clearer role ownership, exports, and workload statistics.
- Cost planning: Billing is based on video-minute credits, so tool costs can be estimated before taking on a project.
- Flexible additions: Add episodes later and support phased delivery for short dramas and long-form series.
Example: 100 minutes of series footage
At a common market rate of RMB 2.5/min for annotation, teams can free up more capacity for additional orders.
Built for teams handling overseas short-drama localization and film dubbing
FAQ
Is this an AI dubbing tool?
No. BDK does not replace human voice actors. It supports the pre-dubbing annotation workflow: identifying speakers, mapping subtitle lines to character roles, adding audio cue notes, and exporting scripts by character with character summaries.
What files do I need?
Usually you need video files and matching SRT subtitles. If you have a character Excel sheet, you can upload it so the system can extract role information.
Can speaker recognition be 100% accurate?
No. The system performs speaker recognition and full-series role annotation first, then annotators review key results. Final exports follow the manually confirmed dubbing roles.
What if one character is split into multiple speakers?
You can map multiple system-detected speakers to one confirmed character role. Even if the model splits a character into multiple speaker profiles, exports can still group them under the same role.
How many trial minutes are included?
New users currently receive 30 video minutes by default after registration. These minutes can be used directly to upload videos and create projects. The balance shown in the account after signup is authoritative.
How is usage priced?
The platform deducts billable minutes when new videos are uploaded. Billing uses 30-second units: 1-30 seconds counts as 0.5 minutes, and longer videos round up to the next 0.5-minute tier. Multiple videos are calculated separately and then summed.
How does top-up pricing work?
Default pricing: minute-based top-ups below 600 minutes are RMB 0.7/min, and 600+ minutes are RMB 0.5/min; amount-based top-ups below RMB 300 are RMB 0.7/min, and RMB 300+ are RMB 0.5/min. Credited minutes round up to the nearest 0.5 minute. The in-app top-up page shows the final price.
Will insufficient balance affect existing projects?
No. A low balance only prevents new video uploads. Existing projects can still be viewed, reviewed, processed for speaker recognition, have roles imported, and exported to Excel.