Free tool · works now
Speaker Talk-Time Analyzer
Measure speaker talk time, airtime distribution, and monologue durations from WebVTT and SRT subtitles.
Runs 100% in your browser. Paste WebVTT or SRT subtitle files to calculate airtime shares, turn frequencies, monologue durations, and export metric ledgers.
# Meeting Speaker Talk-Time Analysis **Total Monitored Speech:** 2m 34s · **Speakers Detected:** 3 ## Conversational Distribution | Speaker | Airtime | Share | Turns | Avg Turn | Longest Turn | | :--- | :--- | :--- | :--- | :--- | :--- | | **Marcus Chen** | 1m 14s | 48.1% | 2 | 37s | 49s | | **Elena Vance** | 0m 56s | 36.4% | 2 | 28s | 30s | | **David Kim** | 0m 24s | 15.6% | 1 | 24s | 24s |
Why measuring conversational turn-taking matters for teams
In high-stakes corporate communication, meeting dynamics determine organizational health. Research across organizational psychology and executive coaching demonstrates that meeting outcomes closely correlate with turn-taking equity. When an executive or team lead occupies 80% of the airtime in a collaborative problem-solving session, psychological safety diminishes and team members withhold critical dissenting opinions.
Similarly, in candidate debriefs, user discovery interviews, and 1-on-1 performance check-ins, conversational ratios serve as an objective diagnostic metric. An interviewer should never speak for more than 25% of a discovery call. However, without concrete data, human perception of speaking time is notoriously unreliable. Our browser-based analyzer parses subtitle cues to provide an objective, mathematical breakdown of dialogue time.
Mathematical cue delta parsing and conversational analytics
Our analyzer processes standard WebVTT and SRT subtitle files to generate structured facilitation insights:
- Timestamp delta calculation: Parses millisecond timing tokens (
HH:MM:SS.mmm) for both start and end boundaries, calculating exact speech duration for each turn. - Speaker voice tag extraction: Supports standard WebVTT voice markup (
<v Elena Vance>) as well as traditional colon-delimited speaker transcripts (Marcus: text). - Airtime share percentages: Aggregates total conversational seconds across all participants and computes normalized percentage shares.
- Turn frequency and average length: Tracks how often each participant spoke and calculates average monologue duration to distinguish frequent contributors from occasional speakers.
- Monologue flags: Identifies continuous speaking blocks exceeding 90 seconds, highlighting moments where one participant held the floor without collaborative dialogue.
Zero network footprint and absolute transcript confidentiality
Meeting transcripts frequently contain sensitive personnel reviews, confidential merger discussions, and proprietary product strategies. Uploading subtitle files to cloud transcription platforms risks regulatory non-compliance and intellectual property leakage. This analyzer operates 100% locally within your browser tab using zero external network requests.
In Oyma, privacy-first meeting recording is a foundational architecture. Our Mac app intercepts meeting audio locally through ScreenCaptureKit and runs Whisper transcription on your Apple Silicon chip with zero recording bots. Discover our botless meeting recording guide and explore how our local AI summaries capture key decisions.
Frequently asked questions
Questions
How does this tool calculate talk time from subtitle files?
The engine parses standard timestamp cue ranges (HH:MM:SS.mmm --> HH:MM:SS.mmm) and extracts speaker names from voice tags (<v Speaker>) or leading dialogue headers (Speaker: text). It sums the millisecond durations for each speaker to calculate exact speaking totals and percentages.
Why is tracking speaker talk-time distribution valuable for teams?
In executive meetings, 1-on-1s, and customer discovery interviews, conversation balance is crucial. Tracking speaking airtime reveals whether a manager dominated a 1-on-1 (aim for 30% manager / 70% direct report) or whether user research interviewers allowed customers sufficient room to speak.
What triggers an extended monologue alert?
The tool flags any single uninterrupted speaking turn that lasts longer than 90 seconds. These alerts highlight opportunities for meeting facilitators to invite participation from quieter team members.
Does this tool transmit my confidential meeting transcripts anywhere?
No. All subtitle cue parsing, duration calculations, and report generation execute locally in your browser memory. No meeting transcripts leave your computer.
How does meeting recording and transcription work in the desktop app?
The desktop app captures system audio cleanly using macOS ScreenCaptureKit and performs live, on-device Whisper speech-to-text with zero recording bots and zero cloud audio processing.
Capture private, botless meeting notes on your Mac
Record Zoom, Google Meet, and Teams calls with on-device Whisper transcription and zero cloud audio transmission in Oyma.