Hi everyone,
I'm planning an online experiment where participants will respond to audio instructions from firefighters. These instructions will carry various tones and emotions, like neutral/abrupt/urgent. We're using AI-generated voices to avoid any bias in the recordings. We've already explored AI tools, including those from Microsoft and Google. However, very few allow us to change the tone of the voice. The best voice we've found so far is from https://murf.ai/text-to-speech. Two version with different tones are attached here. Unfortunately, recent feedback suggests room for improvement.
What are your thoughts on the voice we're currently using? I would welcome any feedback, suggestions, or recommendations for alternative tools or methods for our project. Thank you in advance for your assistance.
Dear Mr. Tong,
for creating TTS recordings and alter the voices used you could use llElevenLabs (hp: https://elevenlabs.io/). They allow you to control the:
- stability,
- clarity
and style exaggeration of the voices / recording. IMHO the results are very good and clear.
Also you could edit the recordings afterwards with free programs like Audacity. When adjusting the parameters of speed, frequency, and pitch, the audio output can be made to convey a greater sense of urgency.
Examples with your text are attached.
Sincerely,
Florian Herfurth
- Badges
- Science topic
- Similar topics
- Philosophy
- Esthetics
- Interpretation