Open Speech and Language Resources



Thorsten Müller (German Emotional-TTS dataset)

Identifier: SLR110

Summary: Free EMOTIONAL single german speaker dataset (Neutral, Disgusted, Angry, Amused, Surprised, Sleepy, Drunk, Whispering) by Thorsten Müller (voice) and Dominik Kreutz (audio optimization) for TTS training

Category: Speech

License: Creative Commons (CC0) Licence

Downloads (use a mirror closer to you):
thorsten-emotional_v02.tgz [399M]   ( Thorsten emotional German speech and transcripts )   Mirrors: [US]   [EU]   [CN]  

About this resource:

I contribute my personal voice as a person believing in a world where all people are equal. No matter of gender, sexual orientation, religion, skin color and geocoordinates of birth location. A global world where everybody is warmly welcome on any place on this planet and open and free knowledge and education is available to everyone.

Speaker Thorsten Müller, 2021
(Thorsten Müller, 2021)

tl;dr Please don't use for evil!

Emotional dataset details:
  • recordings from male single-speaker native german speaker (Thorsten Müller)
  • audio optimized (Dominik Kreutz)
  • 300 identical phrases recorded in 8 DIFFERENT EMOTIONS = 2.400 recordings = 175 minutes audio material
  • ljspeech-1.1 directory structure
  • samplerate 22.050Hz
  • mono
  • normalized to -24dB
  • phrase length 59 - 148 chars
  • no silence at beginning/ending

Recording length (300 identical phrases) for each emotion
  • "neutral" (19 minutes)
  • "amused" (18 minutes)
  • "angry" (20 minutes)
  • "disgusted" (23 minutes)
  • "sleepy" (30 minutes)
  • "surprised" (18 minutes)
  • "drunk", recorded sober without being drunk, just pronounced it that way :-) (25 minutes)
  • "whispering" (22 minutes)

Additional links Please keep in mind that a am no professional speaker or voice actor. I'm just a guy sharing his voice.

External URL: https://github.com/thorstenMueller/deep-learning-german-tts/   Thorsten - Open German Voice Dataset