For today’s installment we’ll be looking at an interesting tool designed to help you with capturing cleaner vocal recordings.
OmniSale GmbH, the company behind the groundbreaking stem separation AI, LALAL.AI, has expanded upon its selection of available services by adding a noise reduction tool designed to remove background noise or music from audio and video recordings.
With their latest addition, LALAL.AI has become the first 8-stem splitter on the market and has expanded its service beyond just the music world. The LALAL.AI Voice Cleaner allows you to obtain cleaner vocal or standard voice recordings all through a window in your browser.
Where Can Voice Cleaner Be Used?
Whether you’re a musician, content creator, sound engineer, transcriber, streamer, or journalist Voice Cleaner can help you. Unfortunately, noise has a way of sneaking up on the best of us, and, outside of recording in a well-treated room or controlled environment, you don’t always have a say in when or how loud it strikes.
Thankfully, we now have access to the tool we need to make our audio-related lives easier.
Various creators and developers can use LALAL.AI Vocal Cleaner for better audio results:
- Musicians
Has a noisy environment spoiled a great vocal recording? You can use the new feature to save your recording, your voice, and your time. - Transcribers
Trying to decipher speech when it is surrounded by sound FX, music, and ambient sounds can make an already challenging job cumbersome. Now you have the tool you need to make text-to-speech easier. - Journalists/Podcasters
Great-sounding interviews can be tricky, especially when you don’t have the freedom to choose where they are conducted. At least now you can fix minor inconveniences after the fact. - Content Creators
Dancing around copyright issues can be tiring, and a single minor mistake can be detrimental. However, the vocal cleaner can aid you by removing any copyrighted music in a recording.
Using LALAL.AI Voice Cleaner
Outside of the free price package (detailed below), you can preview the Vocal Cleaner before signing up. On the site’s homepage, you will find a section to select the type of stem separation you want and a button to upload your file/files. The Voice and Noise stem separation type selects the new feature.

Unfortunately, you cannot download previews, so if you have not signed up already, you will have a series of plans to choose from below your audio previews:

Once you have signed in or purchased a creator’s package, the preview display will change, giving you the option to download the voice and noise stems. In addition, the stems are emailed to you.

Service Quality
After processing a variety of songs, movie trailers, clips, and random voice recordings riddled with ambient noise, I can say that the new feature does its job extremely well. The Vocal Cleaner separates the audio file into two stems: voice and noise.
Depending on the type of noise present, the results do vary. Natural ambient sounds and noises are separated from the vocal stem brilliantly with almost no trace to minimal traces of noise. Abstract noises, like those designed for special FX, weren’t removed as cleanly.
You may find if you are using the vocal cleaner to separate noise and voices for transcription for a film that traces of the FX are prominent within the vocal stem.
However, the AI did an impressive job removing plosives, sibilance, poppysmic, breathing, and ambient noise for music-related audio. Overall, if you have a vocal track needing minor cleaning up, you have found the tool to use.
The only downside I would have to point out is that the natural “dead air” that helps vocal recordings sound authentic is removed with the rest of the noise.
Price Packages
OmniSale offers three standard price packages for using their AI. Two of which are one-time payments. The three packages are as follows:
Price Package |
No. Of Tracks/Overall Time Limit |
File Size |
Audio & Video Support |
Supported Formats |
Price |
Free |
3 tracks/ 10 mins |
10 mb/ file |
Audio |
MP3, OGG, WAV |
Free |
Lite |
10 tracks/ 90 mins |
5 GB/ file |
Audio & Video |
MP3, OGG, WAV, FLAC, AVI, MP4, MKV, AIFF, and AAC |
$15 |
Plus |
30 tracks/ 300 mins |
5 GB/ file |
Audio & Video |
MP3, OGG, WAV, FLAC, AVI, MP4, MKV, AIFF, and AAC |
$30 (40% off at the time of review) |
And for business purposes, a separate trio of price packages are on offer with API access to seamlessly integrate LALAL.AI’s services with your own. The business price packages are as follows:
Price Package |
No. Of Tracks/Overall Time Limit |
File Size |
Audio & Video Support |
Supported Formats |
Price |
Pack 1 |
1000 mins |
5 GB/ file |
Audio & Video |
MP3, OGG, WAV, FLAC, AVI, MP4, MKV, AIFF, and AAC |
$100 |
Pack 2 |
3000 mins |
5 GB/ file |
Audio & Video |
MP3, OGG, WAV, FLAC, AVI, MP4, MKV, AIFF, and AAC |
$200 |
Pack 3 |
5000 mins |
5 GB/ file |
Audio & Video |
MP3, OGG, WAV, FLAC, AVI, MP4, MKV, AIFF, and AAC |
$300 |
Conclusion
The added feature is brilliant. There are a few areas that could be improved upon, like the separation of designed sound FX from the vocal stem when working with film audio, but at the end of the day, the beauty behind what makes LALAL.AI great is the fact that the more it’s used, the better it gets.
What I find most appealing about LALAL.AI is that you lose nothing in trying its features. If you are curious about whether it can help you, you can easily process an audio file and preview it before making your final decision.
Of course, perfection is impossible to achieve, but that doesn’t mean we can’t attempt to come close.

A part-time sound engineering lecturer and musician based in Pretoria, South Africa. He has had a passion for all things music since he was big enough to sit on his cousin’s bed and strum away on an old guitar. All while his uncle, in a room below him, stood with a broom in hand and drummed along on the ceiling. When he isn’t teaching others the basics of sound theory, how to record, and how to mix, you can find him sitting with a guitar fiddling away, completely lost to the world.