Hello there, back again with another excellent article for you! This time we will be looking at pretty new software that I call pure black magic! In this Moises AI review, you will find out how to separate tracks in a song with this awesome algorithm.
As you may already know, separating tracks in a song or isolation instruments in a song was always a big thing. However, from my own experience, I had never seen any software or methodology that really could achieve this.
There were a lot of promises made by different companies, but usually, you would end up with some kind of heavily EQed audio files. Depending on your luck, you would be able to have a usable one, but most of the time, the results were pretty much the same.
On top of that, being someone who spent a lot of time on internet forums, I had come across a lot of people asking how to separate tracks in a song. But usually, they were let down and even ridiculed. Personally, I also felt like they were asking for something impossible and never believed this could be really achieved.
Why Would You Separate Tracks in a Song?
There are various reasons to separate or isolate tracks in a song. Most of the time, people want to remove vocals or specific tracks in a composition so that they can use the remaining parts of the music in what they want.
Removing Vocals from a Song (Making Karaoke Backing Tracks)
One of the main reasons for musicians to isolate tracks in a song is to remove vocals. Most of the time, singers would like to practice and sing along with their favourite tunes.
But if they are not capable of remaking the backing tracks on a computer, not playing with a band or do not play any instruments, they tend to search the internet and try to find karaoke backing tracks.
Even though there are countless karaoke backing tracks available on the internet for singers, most of them don’t exactly sound like the originals or they are not made in the correct pitch that those singers would be able to sing along with.
Isolating Instruments from a Song
There are also many musicians and instrument players who would like to accompany their favourite songs. Normally, they would prefer separated original tracks so that they can mute the channel they want and then practice and play their parts.
However, unless you have the original stems of a song, it would be almost impossible to find the exact song with desired removed instrument parts. Even though you can find countless backing tracks on the internet for various instruments, it’s not always possible to find what you want.
In my opinion as a bedroom guitarist, it’s very crucial for someone who wants to develop their playing further. Having isolated tracks also helps musicians practice their parts in a way more efficient and correct way as they can focus really well on specific parts.
It also helps one to understand how to fill up the space in a mix with their desired instruments. To me, this almost feels like having one of the greatest music education opportunities at your disposal.
How to Separate Tracks from a Song?
As you know, there are countless materials on the internet on this particular topic. Most of them mention or list a couple of software or methodologies that you can attempt to isolate tracks from a song. But in my opinion, generally, they are pretty outdated.
Luckily, with the recent developments in artificial intelligence and machine learning algorithms, we have now what I call true black magic, Moises AI!
In order to introduce you to this wonderful piece of software, I have prepared all the necessary information as well as an example cover song that I made using Moises AI. Read on 🙂
What is Moises AI?
Moises AI (also known as Moises System Inc.) is an American music-tech company back in 2019. As you can understand from the foundation year of the company, Moises AI is pretty new yet in my opinion is the best in this business.
Their main focus is to bring ground-breaking developments in track separation and isolation algorithms. Even though they are a pretty fresh company, they were awarded as the best app in the personal growth category! And they deserve it!
Who Can Use Moises AI?
Moises AI has been developed by musicians for musicians. So basically anyone who is a musician can benefit from Moises AI. This is what separates Moises AI from its competitors.
As I have mentioned in the beginning, you may be able to find similar products, approaches or methodologies when it comes to isolation tracks from a song. However, most of them focus on a single musicians group such as vocalists (there are software called vocal removers).
So whether you are a singer, guitarist, bassist, flautist, trombonist, DJ, music producer or drummer, you can use Moises AI without an issue.
You can download Moises AI via the below links for both Apple and Android devices. Note that, you can also go to their website and use it on your desktop or laptop computer via your preferred browser.
Moises AI Features
Moises AI comes with a set of very assertive and promising features such as AI Audio Separation, Smart Metronome and Audio Speed Changer, Pitch Changer and AI Key Detection and Chord Detection.
Let’s quickly have a look at these awesome features down below.
AI Audio Separation
One of the main features of Moises AI is the AI-driven audio tracks separation from a song. All you need to do is to upload your favourite song to Moises AI using their mobile app or web app, choose the separation type then wait for Moises AI to do its magic!
Once it’s done, you will have a mixer view where you can adjust volumes of each channel, mute/solo individual tracks and play/sing along with your favourite songs!
*This feature is the core of Moises AI and is available on both mobile and web apps.
Smart Metronome and Audio Speed Changer
Another cool feature of Moises AI is the smart metronome and audio speed changer feature! Right after you have uploaded your favourite song to Moises AI and get the individual tracks, Moises AI will also create a metronome track! It saves you a lot of time in figuring out the correct tempo of a song.
You can then also alter the tempo of the song and believe me it works like a charm! Even when you slow down the song, it’s still crystal clear audio quality!
*This feature is only available on the mobile app version.
Pitch Changer and AI Key Detection
Moises AI also comes with a pitch changer as well as a groundbreaking AI driver key detection feature! For instance, you want to remove the vocals from your favourite song, and then sing along with it. However, the original key of the song is not a good match with your vocal range.
Moises AI gets you covered! You can alter the pitch of the song and continue with your singing performance!
*This feature is only available on the mobile app version.
Moises AI also comes with a new beta feature that detects the chords of a song! This AI-driven feature easily finds the chords in a song and shows you along with it.
If you are having difficulties with find chords on your own or you would like a little bit of support, you can benefit from this feature.
*This feature is available on both mobile and web apps.
As I have listed under each subheading, some of the features are only available on the mobile app. To summarise, smart metronome, pitch detection and changer, count-in, loop and trim features are listed as mobile-only features.
You will see from the first image above that I have marked some sections with red coloured rectangle. These are where you can access those features.
Is Moises AI Free?
Moises AI comes in free and premium versions. The free version can be used to test the algorithm and some of the features. However, most of the crucial and useful features come with premium only
For instance, the number of uploads, song length, and retention period of uploaded tracks is pretty limited in the free version.
In order to clearly show the difference between free and premium versions, I have prepared a comparison table below.
Moises AI Free vs Premium Comparison Chart
|Number of Uploads||5 songs/month||Unlimited|
|Retention Period of Uploaded Tracks||Stored for 3 days||Unlimited|
|Additional Instrument Separation||No||Piano/Keyboard Separation|
|Download Format||320kbps MP3||WAV or 320kbps MP3|
|Song Length||Up to 5 minutes||Up to 20 minutes|
|Mastering (Downloadable Wave file)||No||Yes|
|Song Processing||Standard||Fast Processing|
|Smart Metronome||1 minute||Unlimited|
|Chord Detection||1 minute||Unlimited|
|AI Mastering||5 songs/month||Unlimited|
When I first tried it without much expectation, I was already blown away and almost went for premium on my first day. But somehow, I postponed buying it. Then I was able to catch a very cool “Black Friday” discount and went for an annual premium subscription.
Is Moises AI Worth Buying?
I must say, you have to buy the premium version, it’s worth every penny! According to their website, if you pay annually, it will cost you $3.33/month but if you prefer paying only for a month without an annual subscription, it will cost you $3.99. Either way, it’s extremely affordable.
Even when you buy Moises AI premium for a month, you can have so many songs separated and made ready for download.
If you are an instrument player or a singer, Moises AI is one of the best investments for your personal growth and development as a musician.
As many musicians are already aware, playing and transcribing music by ear is one of the best methods to improve your skills. Normally, you would have to slow down a song and then slowly practice all those lines of music. But sometimes due to the mix of the song, not every channel is clearly audible.
With Moises AI separated tracks, you can really focus on a track and practice it without any disturbance of other tracks on the song.
As I have also mentioned in the beginning, being able to isolate all these tracks also helps significantly if you are into recording. One of my favourite activities with Moises AI is to isolate all tracks of a song, then try to fill up the guitar tracks with my own playing. Almost feels like a lesson!
Does Moises AI Really Work?
The question of the day! To be honest, before I had even tried it, I wasn’t expecting much. I only expected some heavily EQed tracks as a result of this AI-driven audio separation. But after I tried, I was really impressed already.
Having said that, in the beginning (mid-2021), there were a limited number of separation types with mostly some main instruments plus “other” classification. Generally, the other track would end up having almost no audio or nothing with real value.
I was also not happy with their file naming systems as every song I uploaded would produce exact same files names such as guitar.mp3, bass.mp3…etc. This would cause a lot of confusion when you downloaded them to your computer.
However, wonderful people at Moises AI have fixed everything and also released a lot of cool features that I have mentioned in the article.
In my last attempts, I have started getting excellent results with their new track separation algorithms. I must say I’m blown away by the results!
Separating Tracks from a Song with Moises AI
In order to prove that Moises AI really works and is worth every penny, here I’m sharing a guitar cover attempt for you! As you know, I’m a bedroom guitarist and I occasionally play guitar covers and publish them on my Youtube channel. So far, I was always trying to manipulate the original songs by EQing some of the frequencies and then attempting to fill these gaps with my guitar playing.
Since I have discovered Moises AI, the first thing I do is upload the song I would like to play to the Moises AI web app and do the first check.
As I have stated, with the latest updates Moises AI gives me exactly what I am after! In order to demonstrate this, I have picked up a very cool song by Özlem Tekin. For those who have never heard of her, she is a pretty successful Turkish musician who released various albums in so many genres.
This particular song is from her last album Kargalar called Sebepsiz Savaş. Sebepsiz Savaş was actually in her debut album Kime Ne back in 1995. But on this album, they made a very heavy cover version.
As far as I know, she was working with one of the best guitar players in Turkey, Özhan Deneç. I’m pretty sure he influenced all the arrangements and heavy guitar tracks.
Note that, I do not own any rights to these songs. I’m just using it as an example and a way of educating my visitors.
I prefer using Moises AI as it gives me everything I need very quickly. As you can see from the first image below, you first need to click on “Track Separation” and then choose the “Separation Type“.
This one I have picked is my latest favourite and just separates everything so well from a guitarist’s perspective.
Upon clicking on upload, it will first upload your song to Moises AI platform then it will process it.
After it’s done, you will have this mixer view where you can adjust the volumes of individual channels or solo/mute tracks.
What I normally do is to first check the separation and then click “Download All” to get all the files in a zipped archive.
Next, I import these tracks into Reaper and adjust the volume levels roughly. In parallel, I would also use the separated guitar track to examine the guitar parts and practice with more focus easily.
At this point, I was also planning how to fill up all the space left by huge guitar tones. So I decided to record 4 guitar tracks with different panning and volume levels.
I also recorded the short solo and the ending support riff on separate tracks.
Özlem Tekin Sebepsiz Savaş Gutar Cover #madewithmoises
Here are the results! All the guitars you will be hearing in this video are played by me! Nothing comes from the original track in terms of guitar tones!
I used my Epiphone SG Custom Ebony, plugged into ENGL E570 All Tube Preamp (Lead II, Contour active) and then ENGL Cabloader (4x12 ENGL V30, 6L6 Power Amp, 57 microphone model). I also used Valhalla reverb and delay effects as you will hear in the video.
I hope you like my playing this cover song!
I hope you like this article and have found this helpful and inspiring for your next cover projects. Many thanks for visiting my blog and supporting me so far! I will hopefully see you in the next content here!