Voice Artists Needed for Kannada language | Studio-Quality NLP Training Project
Budget: $50 – $0 USD
We are seeking professional voice artists in Kannada to take part in a studio-quality NLP training project. The task involves recording 90 minutes of high-quality speech under strict technical guidelines. These recordings will be used to advance next-generation text-to-speech (TTS) models.
Highlights
Studio setup required: Treated recording environment with a high-quality microphone (no home ambient/background noise).
90 minutes of final audio: Clean, edited takes only — ready for model training.
AI training use only: Recordings will not be used for commercial purposes.
Remote project: Clear guidelines, flexible timeline, and smooth onboarding.
Budget: $50–$60 per hour (upon quality approval).
Essential Requirements
Audio Standards
Format: .WAV, 44.1 kHz, 24-bit depth, Mono output (no stereo).
Recording equipment frequency range: 20 Hz–20 kHz capture (below this will be rejected).
Amplitude: Must stay within –0.25 to +0.25 (outside range will be rejected).
Gain: Maintain between –6 dB and –9 dB.
Sound quality: No echo, room reverb, static noise, background sounds, music, EQ, or compression (raw audio only).
Clipping: Absolutely no hard or soft clipping.
Pronunciation: Clear, precise, and native Kannada accent.
Artist Requirements-
Ensure clear pronunciation throughout (no lisping, cut-off words, nasal congestion, or squeaks).
Maintain hydration and vocal care to avoid mouth noise and lip smacking.
Use proper mic distance to avoid popping/booming sounds.
Record in a noise-treated studio environment with minimal interference.
Keep a natural and steady pace — not too fast, not too slow.
How to Apply:
Please submit a 1-minute audio sample in Kannada recorded in your studio and meeting all the requirements above. The sample should reflect the same quality you will provide for the full 90 minutes.
We are looking forward to collaborating with talented Kannada voice artists on this exciting project!
Highlights
Studio setup required: Treated recording environment with a high-quality microphone (no home ambient/background noise).
90 minutes of final audio: Clean, edited takes only — ready for model training.
AI training use only: Recordings will not be used for commercial purposes.
Remote project: Clear guidelines, flexible timeline, and smooth onboarding.
Budget: $50–$60 per hour (upon quality approval).
Essential Requirements
Audio Standards
Format: .WAV, 44.1 kHz, 24-bit depth, Mono output (no stereo).
Recording equipment frequency range: 20 Hz–20 kHz capture (below this will be rejected).
Amplitude: Must stay within –0.25 to +0.25 (outside range will be rejected).
Gain: Maintain between –6 dB and –9 dB.
Sound quality: No echo, room reverb, static noise, background sounds, music, EQ, or compression (raw audio only).
Clipping: Absolutely no hard or soft clipping.
Pronunciation: Clear, precise, and native Kannada accent.
Artist Requirements-
Ensure clear pronunciation throughout (no lisping, cut-off words, nasal congestion, or squeaks).
Maintain hydration and vocal care to avoid mouth noise and lip smacking.
Use proper mic distance to avoid popping/booming sounds.
Record in a noise-treated studio environment with minimal interference.
Keep a natural and steady pace — not too fast, not too slow.
How to Apply:
Please submit a 1-minute audio sample in Kannada recorded in your studio and meeting all the requirements above. The sample should reflect the same quality you will provide for the full 90 minutes.
We are looking forward to collaborating with talented Kannada voice artists on this exciting project!