Audio Separation and Cleaning Project
Budget: ₹37,500 – ₹75,000 INR
Project: Split Two Speakers into Separate Clean Audio Tracks
I have approximately 14 hours of audio recordings containing two speakers talking in the same file. I need someone to:
Task Requirements:
Separate the audio so that each speaker has their own isolated audio file
Ensure high accuracy when assigning each speaker’s voice to the correct file
Reduce or remove overlapping audio bleed where possible
Maintain or improve clarity, intelligibility, and volume consistency
Export final results in WAV format
Use clear naming conventions such as:
Speaker 1.wav and Speaker 2.wav
You may use AI tools or traditional audio editing tools to complete this task — whichever allows you to produce the most accurate results.
Audio Details:
Total audio files: 32
Format: Stereo WAV
Length breakdown:
14 files ≈ 41 minutes each
14 files ≈ 21 minutes each
4 files ≈ 6 minutes each
Each file contains exactly two speakers
Important Note (Please Read Carefully):
You will first receive a 41-minute sample file.
This is not a short trial — it is a complete script.
* Your job is to fully separate the audio into two complete speaker files, covering the entire 41 minutes.
How We Verify Your Work:
Once submitted, we will upload:
The original sample audio
The two separated audio files you send back
Our software will then:
Compare both files against the original
Calculate the exact percentage accuracy
Detect any missing portions, incorrect speaker assignments, or incomplete sections
**So please make sure you complete the sample fully and accurately.
Incomplete or incorrect separation will be automatically flagged.
Ideal Candidate:
Experience with voice isolation, podcast editing, or speaker diarization
I have approximately 14 hours of audio recordings containing two speakers talking in the same file. I need someone to:
Task Requirements:
Separate the audio so that each speaker has their own isolated audio file
Ensure high accuracy when assigning each speaker’s voice to the correct file
Reduce or remove overlapping audio bleed where possible
Maintain or improve clarity, intelligibility, and volume consistency
Export final results in WAV format
Use clear naming conventions such as:
Speaker 1.wav and Speaker 2.wav
You may use AI tools or traditional audio editing tools to complete this task — whichever allows you to produce the most accurate results.
Audio Details:
Total audio files: 32
Format: Stereo WAV
Length breakdown:
14 files ≈ 41 minutes each
14 files ≈ 21 minutes each
4 files ≈ 6 minutes each
Each file contains exactly two speakers
Important Note (Please Read Carefully):
You will first receive a 41-minute sample file.
This is not a short trial — it is a complete script.
* Your job is to fully separate the audio into two complete speaker files, covering the entire 41 minutes.
How We Verify Your Work:
Once submitted, we will upload:
The original sample audio
The two separated audio files you send back
Our software will then:
Compare both files against the original
Calculate the exact percentage accuracy
Detect any missing portions, incorrect speaker assignments, or incomplete sections
**So please make sure you complete the sample fully and accurately.
Incomplete or incorrect separation will be automatically flagged.
Ideal Candidate:
Experience with voice isolation, podcast editing, or speaker diarization