Universal 3 Pro Transcribes Whispered ASMR Accurately (Full Transcript)

Demo shows Universal 3 Pro transcribing low-volume whisper audio and, with prompting, identifying ASMR sounds like tapping and breathing.
Download Transcript (DOCX)
Speakers
add Add new speaker

[00:00:00] Speaker 1: Hi cutie, cutie, cutie, cutie. Hi cutie, cutie, cutie, cutie. Hi cutie, cutie, cutie, cutie. I don't know about you, but ASMR makes me really uncomfortable. But this actually helped me discover something really interesting. Universal 3 Pro can really transcribe whispering audio. I can lower the volume on this ASMR recording, and even though it's just whispering, watch this. Universal 3 Pro transcribes it perfectly, but it gets even better. Even though the audio is low, it catches every word and it's super accurate. If I upload the file and then add this prompt to capture whispering, tapping, breathing, mouth sounds, and so on, it transcribes that too. Check this out. You can see it actually identifies and transcribes those areas really well. Hi cutie, cutie, cutie, cutie. Hi cutie, cutie, cutie, cutie. Universal 3 Pro's prompting capabilities expands use cases into fields that weren't previously possible, even with audio files that aren't exactly normal. So if you haven't tried it already, head down to our playground and try Universal 3 Pro today.

ai AI Insights
Arow Summary
A speaker discusses how ASMR audio can be uncomfortable but demonstrates that Universal 3 Pro can accurately transcribe whispered, low-volume ASMR recordings. They show that even with reduced volume, the model captures every word, and with an added prompt it can also identify and transcribe non-speech ASMR elements like tapping, breathing, and mouth sounds. The message concludes by highlighting that prompting expands transcription use cases for unusual audio and invites users to try Universal 3 Pro in the playground.
Arow Title
Universal 3 Pro Accurately Transcribes Whispered ASMR Audio
Arow Keywords
Universal 3 Pro Remove
ASMR Remove
transcription Remove
whispering audio Remove
low-volume audio Remove
prompting Remove
audio recognition Remove
tapping sounds Remove
breathing Remove
mouth sounds Remove
playground demo Remove
Arow Key Takeaways
  • Universal 3 Pro can transcribe whispered, low-volume audio with high accuracy.
  • Prompting can help capture and label non-verbal sound events (tapping, breathing, mouth sounds).
  • These capabilities expand transcription to atypical or 'non-normal' audio files.
  • Users are encouraged to test the feature in the product playground.
Arow Sentiments
Positive: The tone is promotional and enthusiastic, emphasizing improved accuracy and expanded capabilities, with curiosity and mild discomfort about ASMR used as a hook.
Arow Enter your query
{{ secondsToHumanTime(time) }}
Back
Forward
{{ Math.round(speed * 100) / 100 }}x
{{ secondsToHumanTime(duration) }}
close
New speaker
Add speaker
close
Edit speaker
Save changes
close
Share Transcript