RTTM Syntax Extension
Universal syntax highlighting support for RTTM (Rich Transcription Time Marked) files in Visual Studio Code.
Example
SPEAKER CMU_001 1 0.00 5.24 <NA> <NA> spk1 <NA> <NA>
LEXEME CMU_001 1 0.10 0.40 hello lex spk1 0.97 <NA>
NON-SPEECH CMU_001 1 5.24 1.50 <NA> music <NA> <NA> <NA>
NOSCORE CMU_001 1 0.00 0.20 <NA> <NA> <NA> <NA> <NA>
SPKR-INFO CMU_001 <NA> <NA> <NA> <NA> adult_male spk1 <NA> <NA>
# End of file example
About RTTM
RTTM (Rich Transcription Time Marked) is a standard text-based format used widely in audio annotation, speaker diarization, and automatic speech recognition (ASR).
An RTTM file consists of lines with exactly 10 space-separated fields:
| # |
Field Name |
Description |
Typical Values |
| 1 |
Type |
Segment type |
SPEAKER, LEXEME, NON-SPEECH, NOSCORE, etc. |
| 2 |
File ID |
File name (without extension) |
e.g., CMU_001 |
| 3 |
Channel ID |
Audio channel (usually 1) |
1, 2 |
| 4 |
Turn Onset |
Segment onset (in seconds) |
e.g., 130.430 |
| 5 |
Turn Duration |
Segment duration (in seconds) |
e.g., 2.350 |
| 6 |
Orthography |
Orthographic transcription or <NA> |
word text, <NA> |
| 7 |
Subtype (stype) |
Subtype of segment or <NA> |
male, female, child, lex, <NA> |
| 8 |
Speaker Name |
Speaker identifier |
e.g., spk1, <NA> |
| 9 |
Confidence Score |
Confidence probability (0–1) or <NA> |
0.97, <NA> |
| 10 |
Lookahead Time |
Lookahead/prediction time or <NA> |
e.g., 0.30, <NA> |
Fields marked as <NA> ("Not Applicable") represent optional or unspecified values.
Features
- 🎨 Full syntax highlighting for all RTTM segment types:
- Segment Types (
SPEAKER, LEXEME, NON-SPEECH, NOSCORE, etc.)
- Speaker Information (
male, female, child, etc.)
- Timestamps (
Turn Onset, Turn Duration)
- Confidence and Lookahead Times
- 📝 Comments (
#) support
- 🚀 Universal grammar supporting all RTTM-compatible tools
| |
