聲音去除器
Vocal Remover & Stem Splitter
用人工智能提取人声和乐器
SongDonkey is online vocal remover that extracts or isolates the vocal parts of an audio track or music recording. Separate stems (drums, bass, piano, etc.) with our vocal AI splitter. Make your own karaoke version of any song by removing voice or instruments.
How to Split Stems
- Click start and upload track from your PC, Mac or mobile device (use .mp3 or .wav file no longer than 10 minutes).
- Choose how many stems you want track to be separated.
- Listen to preview.
- Make a payment and download splitted tracks.
Or drop files here
或将文件放在这里
Only .mp3 and .wav files
只有 .mp3 和 .wav 文件
High Quality
总是快
No Registration
没有注册
$0.5/Song
How Our Online AI Vocal Remover Works
AI-powered vocal remover utilize machine learning techniques to identify and remove vocals from music tracks:
- It is trained on large datasets of songs with and without vocals. The AI models learn to recognize the unique characteristics of human voices.
- Common training techniques involve convolutional and recurrent neural networks analyzing spectrogram representations of the audio data. When fed a new track, the AI model can break it down into individual components and isolate the vocal parts based on what it learned during training.
- The vocal separation happens in a more musically-meaningful way compared to classic DSP approaches. The AI can better distinguish between vocals and instruments.
- Once isolated, the vocal parts are attenuated or removed while retaining the rest of the instrumentation. The AI filling in gaps left by the vocals.
- Additional post-processing is applied to smooth out artifacts and make the final instrumental mix sound seamless. The AI vocal removal is adjustable, allowing the user to control how much of the vocals get removed.
- The models continue to learn and improve over time as they process more music data. The separation and processing quality gets better.