AI is being talked about a LOT, but when it comes to the tools that will bring great results RIGHT NOW, there are only a few I think actually work currently.. Therefore, I made a list of my favorite AI-based tools that already deliver amazing results and will help you as a musician as soon as you start using them:
4 THOUGHTS FROM ME
1. Develop better chord progressions and lyrics with CHATGPT. As shown previously in this video, CHATGPT can be really helpful to you as a musician when making chord progressions and/or writing lyrics. For example, you can ask for melodies from scratch in the mood you like (“write a happy chord progression in A minor”) or ask for multiple iterations until you find something you want (“I don’t like it, please send me another one”). In addition, you can ask for variations of your chord progression. For example, “if my chord progression is Am, Em, Dm, and Fmaj, could you make some variations that go along with that”. It’s important to understand how to properly ask ChatGPT to write those chords, but what I love about it is that it often gives me variations that I wouldn’t think of right away and would consider as ‘out of the box’. This way, you can not only create your own chord progression, but also expand by generating alternative progressions that you can place in your breakdown.
2. Convert your poorly recorded vocals to different artists or test if your song is vocal-worthy with LALALS.COM. I’m a terrible singer and LALALS made my voice (1) sound like Drake and (2) usable for a song (check the result of my song over here, even though the lyrics are terrible). To test it, I did a small test with my phone’s microphone to test a crappy mic, and even then it made the vocal ‘usable’ for a song, essentially enabling you to sing and deep fake your voice into someone you love. “But Leo, what about Copyright?” When using Lalals, your intention should be to test if the vocal works in a song and what kind of vocal would be adequate for it, so you might not be using it in a song. For example, pick a vocal from a vocal pack then test it as a male or female, and if it works, boom, you already know what to look for. Or, change that voice to a different artist that doesn’t sound perfect to avoid copyright issues, and again, you have something unique to work with. It’s a paid service if you want to download it, but you can always record the input with OBS, or support them by subscribing (as I do). Second, you can create your own male and a female voice and check if a vocal works as if it was her or him. If it does, then you can later contact someone similar without worrying if it will or not work since I’ve already pre-tested that kind of voice. In any way, this will save you hours when trying to find the perfect tone for your song, if the song works with a vocal or even to do remixes of songs in other artists’ voices.
3. Enhance your mastering with tools like Soothe2 and Gullfoss, or explore a really decent auto-mastering tool with Ozone by Izotope. Mastering tools that involve AI are getting significantly better and I hate to admit that I can’t live without them anymore. They’re so great that I use them in almost all my masters nowadays. What I love about them is that they can help you do some extra mixing in your song that you haven’t thought about or just forgot to fix (or didn’t know how to). For example, in a recent master, I was doing for a client, the song had a lot of kick and bass, very few mids and highs, and I fixed the master with 2 Gullfoss, 1 soothe and Ozone. “Leo, but that’s not AI, is it?” Yes it is, especially because these tools dynamically adjust and apply fixes to your track specifically according to a very limited set of parameters you give to them, and it’s impressive how it always makes it better. Therefore, if you already know how to master, Gullfoss and Soothe2 will make your masters even better (wait for Black Friday for better deals on them). If you don’t know how to master, Ozone, for example, has the Mastering Assistant, which understands your song and masters it instantly for you, and I’ve heard two Armada producers say that’s how they master their songs. The best way to buy the ozone bundle is through crossgrades on PluginBoutique by buying a cheap bundle and then crossgrading, which can save you 50-60%. Email me if interested and I’ll help you out.
4. Convert text to speech with Play.Ht and make amazing and impacting catchphrases and ambiences for your songs. Melodic Techno is full of words and phrases of impact like what Anyma’s has done in many of their songs, like Eternity. However, how can you make them? With Play.Ht, you can use different kinds of male and female vocals, with different intonations and intentions, and create several spoken word vocals to use in your songs. For example, you can use ChatGPT to create a speech or text, and then convert that speech with Play.HT into voice, then apply reverbs, delays, chorus, vocoders, and several other FX to make your sound more like Melodic Techno. But, what if you don’t make melodic techno, like me? Well, then you can wash these vocals even more and use them as a background ambiance for your songs, or you can use them as ‘monologues’ like what I did in my recent song ‘Looking at Me’. In any case, instead of using the same old ambiances and or using samples that a lot of other producers might have already used, create your own unique speeches.
1 QUESTION FOR YOU
What other AI Tools can you use?
1. Break songs into stems with LALAL.AI: With Lalals.ai, you can break any song you want into Melodics, Bass, Drums and Vocals. It’s not perfect, but it’s really useful if you want to analyze that song!
2. Clean your stems and songs with iZotope RX: If you receive vocals that have too much noise, or if you have a loop that you really like that creates a little (and annoying) click at the end of every 8 bar, you can clean all this and a lot more with iZotope RX.
3. Make some really useful drums with Emergent Drums 2: If you want to make your own drums, you can use Emergent Drums to make your own and it has some amazing results, especially for Closed Hats, Percs, Claps, Snares and Glitch noises. They are soon releasing another tool called ‘Deep Sampler’, in which you can import a sample you love and the AI will generate 4 sounds based on the one you imported, which is really cool when you want to change the sounds you always use.
