Contact person: Bilingual Asian Voice Recording

Phone:Show

Email:Show

Location: Hyderabad, India

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"The task is to capture clear, natural-sounding conversational audio that will be fed directly into an AI speech-training pipeline. Recordings must be in any two (or more) of these languages at a native or near-native level: Thai, Indonesian, Malay, Vietnamese, Japanese, Korean, or Chinese. A studio-quality condenser or dynamic microphone is essential so the data passes internal SNR and clarity checks without extra post-processing.

You will receive a short set of conversation prompts along with file-naming and channel-separation guidelines. Simply record both sides of each dialogue—either with a partner or by self-staging on two tracks—then deliver the raw, uncompressed WAV files plus a brief log sheet listing language pair, duration, and mic model used.

Acceptance criteria
• Speech is fluent and accent-authentic in both chosen languages
• No clipping, background hum, or room echo beyond −60 dB noise floor
• Each prompt fully covered, total runtime per language pair 25–30 minutes
• Files and logs follow the provided naming convention exactly

Once the submitted audio passes our quality gate, we sign off and release the milestone immediately, enabling you to move on to additional language pairs if you wish." (client-provided description)