Today we’ll give you a step-by-step guide on how to use LALAL.AI to remove vocals from musical files. Additionally, we’ll show you what it is, how it works, and try it out for ourselves!
People have been extracting vocals from music for a long time. In the beginning, it could be mainly for karaoke purposes, but nowadays, there are many other uses for vocal-less tracks. From background music for social media to movie translations, we found the perfect solution. Let’s dive in:
What is LALAL.AI?
LALAL.AI is an AI-based stem extractor for musicians, producers, creative professionals, and enthusiasts. The website allows you to remove vocals and instrumental tracks with superb quality almost instantaneously. It is perfect to create an instrumental track when you don’t have access to the stems or multitracks.
Have you ever wondered how to remove the vocals from a song without artifacts? Isn’t it frustrating when you remove the vocals but the instrumental sounds thin and with several phase issues? LALAL.AI offers to solve those problems with its online service. Whether you need an instrumental track to use as the background of a video, or you’re planning a karaoke night with friends, LALAL.AI can help you.
Furthermore, the service is beating the competition thanks to its advanced AI. And while there is still no way to extract vocals from a song perfectly, LALAL.AI offers the best solution on the market. The technology behind the service acts incredibly even on the most complex vocal melodies, and the resulting sound is impressively clean. Take a look:
How Does LALAL.AI Work?
LALAL.AI is web-only, so you don’t even need to download an app for your phone or computer. It supports files of all formats, sizes, and lengths, so as long as it has audio, you can upload it. Additionally, the service offers an entry pack to extract the audio from three songs for free here.
At LALAL.AI, vocals and instrumentals are separated by a neural network. Like any artificial intelligence, the more operations it performs, the smarter it gets. So, the team at LALAL.AI spent long months feeding the neural networks with hundreds of thousands of songs. All this before it was even ready for the general public. It was massive work, but now the algorithms learn with every new song that users upload. That means the service gets better every day, as producers, singers, and karaoke lovers from around the globe use it.
The quality of the separated stems depends heavily on the source file. If you upload an Mp3 file, your stems will likely sound blurry and distorted. However, a lossless file such as WAV or FLAC will have a greater chance of sounding clearer and more defined. Here’s a quick comparison using a snippet from a live performance by Yellow Boulevard, recorded at my studio:
Stems from a 24 bit/48kHz WAV file:
Stems from a 128 kbps Mp3 file:
It’s clearly noticeable that the Mp3 files carry way less quality than the WAV files. But the service provided by LALAL.AI is stellar; live recordings are a lot harder to isolate than studio ones. In conclusion, if you’re planning on using their services for professional scenarios, do the following:
- Use lossless files whenever possible
- Subscribe to one of LALAL.AI’s Creator Packages
How Do You Use LALAL.AI to Remove Vocals?
Using LALAL.AI is extremely simple and easy. All you have to do is enter their website. There, you’ll find a drag and drop interface on the home page in which you can upload the song. The AI will process the data for a few seconds, and you’ll get two downloadable files: one with the instrumental and another only with the vocals.
LALAL.AI lets you grab your stems in five easy steps. We’ll walk you through the process here:
1. Open the website
Upon opening the website, the first thing you’ll see is the upload interface.
2. Select the Processing Level
Under the “Select File” button, you’ll find three processing levels: Mild, Normal, and Aggressive. Mild will give you some bleed between stems since the processing is lighter. Normal is the default processing option and will suit most needs. Finally, Aggressive uses all of the algorithm’s processing power for extra clean results. Additionally, you can choose between using LALAL.AI’s new algorithm, which is recommended, or the older version.
3. Click ‘Select File’
Upon clicking the button, a new window will open your Documents folder. Then, find the files you want to process and click “Upload.”
4. Wait while the AI works
Although it depends on the file’s size and length, it usually doesn’t take a minute.
5. Download your Stems
To download the entire audio, you need to log in to the website. Even so, you can download and listen t0 the previews from this page.
Conclusion
It’s exciting to see advancements in technology, especially those making our lives as producers and musicians easier. For now, LALAL.AI is focusing on extracting vocals and instrumentals, but we can only hope for the day when it can also extract individual instruments. Imagine how much better it would be for musicians to learn and practice if they could listen to their reference instruments without the rest of the track. Until then, LALAL.AI is hands down the best stem extractor on the market.
Pedro Nascente is an artist, record producer, and mix engineer, currently operating his own studio and working with his band, Yellow Boulevard. Believing that music should convey experiences and feelings, Pedro is known for applying design thinking to his workflow to achieve different sounds and deliver the right messages.