The Artificial intelligence It continues with its irrepressible expansion to carry out impossible or extraordinarily laborious jobs in a matter of seconds. Now it’s the music industry’s turn to achieve something unthinkable a short while ago: creating a perfect karaoke track from the original song.
In 2022 we saw how the use of information systems exploded. Artificial Intelligence (AI) to carry out creative and artistic works that we simply considered impossible years and decades before. Although, to be honest, the vast majority of advances in focused on the generation of images and illustrations.
This situation ignited the debate about the scope of this technology and how it could impact the work of illustrators, designers and other graphic artists who could never achieve these results in those few minutes.
But apparently that was just the beginning and the power of these platforms is about to spread to other territories where even their intervention was difficult to conceive of a few years ago. As for example in the music industry with tasks unrelated to the production and conception of original melodic pieces.
And it is that someone has now developed a technology that uses precisely the AI to do in seconds a task that was previously complicated, imperfect and extremely complicated: remove the audio track of finished original songs.
Vocal Remover shows the power of Artificial Intelligence to create karaoke tracks
the platform of Vocal Remover It has the peculiarity that its access is free and free. Just enter the official website to try it.
There, the page asks the user to upload an audio file of a song, either in format MP3, FLAC or Opus. And in a matter of seconds the Artificial Intelligence will load the song, analyze it and completely remove the voice track.
Doing this manually is feasible, however, any audio engineer or specialist can confirm that even with the most advanced programs it is a non-immediate task that requires care and attention to each section of the melody.
The case with Vocal Remover is the opposite, it is enough to have the song file, upload it to the platform and in less than a minute it will upload the melody to the interface with two separate tracks: the audio track and the instrumentation track.
The results achieved by the AI are impressive. We did the test with a relatively complicated song due to the similarity between the frequencies of each instrument, the wind sections and the tone of the main voice: Lodo de Torreblanca:
In the end result it was not flawless in some sections, particularly in the chorus, but in most parts the karaoke track is perfect.
For obvious reasons we do not share the final result, but the platform allows you to download the resulting audio in MP3 format.