Freelance Voice Talent – R&D Project (US and Canada English Speakers)
Project Overview
Be part of the launch of a revolutionary interactive calendar! Our client is building a diverse voice dataset for their device's voice activation feature, and authentic accent representation is essential to ensuring the system performs accurately across the full spectrum of American English speakers.
We are specifically seeking voice talent who naturally and authentically speak with one of the following accent profiles as their genuine, everyday speaking voice:
• Indian American English
• African American English
• Hispanic American English
These are specific linguistic requirements of the dataset. The voice activation system must be trained on authentic phonetic diversity to function accurately for all users — performed or affected accents cannot be accepted for this reason.
Applicants should self-identify as naturally speaking with one of the above accent profiles.
The Task
This is a straightforward, high-impact task. If selected, you will record 40 short phrases as individual audio files. The work should take no more than 2 hours to complete.
Requirements
• Accent:
Must be a natural, authentic speaker of one of the three accent profiles listed above
• Equipment:
Must have access to a high-quality recording setup
• Preferred: Large-diaphragm condenser microphone (e.g. Neumann TLM102, Audio Technica AT4040, or equivalent)
• Environment:
Recordings must be made in a quiet, acoustically treated space with zero background noise or echo
Technical Specifications
Submitted files must meet the following criteria to be accepted:
• Format:
WAV (Mono)
• Sample Rate:
16 kHz
• Bit Depth:
16-bit
• Delivery:
40 separate audio files (one phrase per file), clearly labelled using the provided naming format
Recording Guidelines
• Distance:
Maintain exactly 1 metre between your mouth and the microphone
• Tone:
Natural, conversational speaking voice
• Pacing:
Normal speed — please do not over-articulate or "perform" the line
All details can be found on the project page here
Privacy Notice
All data collected is strictly for internal device testing and algorithm improvement. Your voice will not be used for promotional purposes or public broadcast.
About Twine
Twine is a leading freelance marketplace connecting top freelancers, consultants, and contractors with companies needing creative and tech expertise. Trusted by Fortune 500 companies and innovative startups alike, Twine enables companies to scale their teams globally.
Our Mission
Twine's mission is to empower creators and businesses to thrive in an AI-driven, freelance-first world.
Apply tot his job
Apply To this Job
Be part of the launch of a revolutionary interactive calendar! Our client is building a diverse voice dataset for their device's voice activation feature, and authentic accent representation is essential to ensuring the system performs accurately across the full spectrum of American English speakers.
We are specifically seeking voice talent who naturally and authentically speak with one of the following accent profiles as their genuine, everyday speaking voice:
• Indian American English
• African American English
• Hispanic American English
These are specific linguistic requirements of the dataset. The voice activation system must be trained on authentic phonetic diversity to function accurately for all users — performed or affected accents cannot be accepted for this reason.
Applicants should self-identify as naturally speaking with one of the above accent profiles.
The Task
This is a straightforward, high-impact task. If selected, you will record 40 short phrases as individual audio files. The work should take no more than 2 hours to complete.
Requirements
• Accent:
Must be a natural, authentic speaker of one of the three accent profiles listed above
• Equipment:
Must have access to a high-quality recording setup
• Preferred: Large-diaphragm condenser microphone (e.g. Neumann TLM102, Audio Technica AT4040, or equivalent)
• Environment:
Recordings must be made in a quiet, acoustically treated space with zero background noise or echo
Technical Specifications
Submitted files must meet the following criteria to be accepted:
• Format:
WAV (Mono)
• Sample Rate:
16 kHz
• Bit Depth:
16-bit
• Delivery:
40 separate audio files (one phrase per file), clearly labelled using the provided naming format
Recording Guidelines
• Distance:
Maintain exactly 1 metre between your mouth and the microphone
• Tone:
Natural, conversational speaking voice
• Pacing:
Normal speed — please do not over-articulate or "perform" the line
All details can be found on the project page here
Privacy Notice
All data collected is strictly for internal device testing and algorithm improvement. Your voice will not be used for promotional purposes or public broadcast.
About Twine
Twine is a leading freelance marketplace connecting top freelancers, consultants, and contractors with companies needing creative and tech expertise. Trusted by Fortune 500 companies and innovative startups alike, Twine enables companies to scale their teams globally.
Our Mission
Twine's mission is to empower creators and businesses to thrive in an AI-driven, freelance-first world.
Apply tot his job
Apply To this Job