How Speaker Labeling Works
Good Tape uses advanced diarization to automatically detect different voices and assign labels, such as Speaker 1 and Speaker 2.
Enhanced Accuracy: You can now specify the number of speakers during the upload process to help the system distinguish between voices more accurately.
Professional Feature: Once you upgrade to a Professional plan, you can enable speaker labels via the toggle switch in the top-right corner of your transcript.
Naming: You can edit the names of the speakers at any time by clicking directly on the labels themselves within the editor.
Pro-Tips for Better Results
Speaker labeling is most effective when the following conditions are met:
Distinct Voices: The AI performs best when speakers have clearly different vocal tones.
High Audio Quality: Clear recordings without background hum or distance from the microphone lead to better results.
No Overlapping: Try to avoid instances where people speak over each other, as this can confuse the detection system.
Improving & Troubleshooting Labels
If the labels aren't as accurate as you'd like, you can try these steps:
Open your transcript.
Click the three dots (...) menu in the top right corner.
Select "Regenerate speaker labels" to have the AI take another pass at the audio.
Important Limitations to Keep in Mind
Pre-Upgrade Transcripts: Transcripts created on a free account before upgrading will not have labeling enabled. To get labels for those files, please re-upload them after your Professional subscription is active.
Similar Voices: In recordings where many people sound similar, the labels may be less accurate initially.
Audio Issues: If the file has poor sound or significant background noise, the AI may struggle to differentiate between speakers.