.jpg)
Main image courtesy of Unit Rcrds.
There’s been much chatter in recent years about AI and its role in creative industries. Writing, art, cinema, and of course music have all been affected by the introduction of AI tools, for better or worse.
Regardless of your views on the efficacy and quality of artificial intelligence, there is clearly a future for such technologies in the creative process of music production. But it’s not as simple as “Siri, make me some music.” It takes time and a little know-how to make these AI models work for you. So today we’ll be taking a look at five essential tips and tools for incorporating AI in your music production creative process.
Note: Laws and regulations around AI music are still iffy. Songs and other material generated by the tools outlined in this guide are best used as guides or inspiration rather than for profit.
Before you start adding AI to your creative process, it would serve you well to know how other creatives use AI, even outside of music production. In general, AI can save you a lot of time by doing repetitive tasks, allowing you to focus on other parts of your work. Here’s a few tasks that AI can be a great help with:
If you’re ever in need of concepts for a new product or project, AI like ChatGPT can generate a whole list of ideas from a single prompt.
Anyone who records videos will tell you that editing can be a tedious job. Tools like Veed.io take some of the hassle out by automatically removing pauses and mistakes.
Adding to or subtracting from one of your photos can be a daunting task for those not skilled in photo editing. Something like DALL-E makes it easy to edit photos in a variety of ways.
If you need to make striking advertisements or social media posts for your business, Microsoft Designer or a similar AI model can create well designed graphics based on the vision you outline in your prompt.
You’ll want to know how AI music tools work before you start to integrate them into your workflow. At a basic level, AI consumes a large amount of music data, such as songs, instrumentals, and lyrics, and analyzes this data to create algorithms. These algorithms are the blueprint that the AI will use to generate new music, incorporating the complexities of the material it was trained on.
Many AIs also use neural networks, a way of computing that mimics the human brain. This allows AI to learn, interpret, and replicate music similarly to how humans do, just at a much faster rate.
The more you tool around with AI and learn about certain patterns and quirks in its output, the better you’ll be able to shape it to your use case. Experiment, have some fun with a variety of prompts and instructions before trying to make a masterpiece.
An AI’s output is only as good as the human input. You have to feed the AI content that is both high quality and reflective of your unique vision for your project. So what is your vision? Ask yourself these questions:
Once you can answer these questions, you can more easily guide your AI tools to an acceptable output. Make sure you include the answers to these questions in the prompt you give the AI.
ChatGPT may be able to generate simple songs, but you’re better off using an AI software specifically built for music creation. Suno AI is one such software, allowing users to create music by entering a prompt. These prompts generate two different variations of a song, at which point you can “Like” a song to add it to your library or “Dislike” a song to remove it from view.
Generated songs cut off at the 2-minute mark. “Liked” songs can be worked on further by using the “Extend” feature. This lets you choose a starting point from the original snippet and add a new 2-minute segment to it. You can repeat this process on and on until you have your perfect song.
Want more control over the songs you create? Suno also offers a “Custom Mode” that allows you to input your own lyrics and specify the “Style of Music,” meaning a mood or genre you’re trying to appeal to.
Once you’ve created a track, you can share it with the Suno community by making it public on their homepage. The homepage features a myriad of songs made by people using Suno. Hear something you like? You can use the “Extend” feature on any public song. Or try the “Reuse Prompt” feature to easily generate a song with the same parameters and build something off that base.
Another great feature for finding inspiration on Suno is the Explore section. It’s a menagerie of unique music styles, some created by AI, that you can then go on to include in your “Style of Music” prompts.
Suno’s Basic Plan is completely free, allowing you 50 credits a day to spend on song generation (each generated song costs 5 credits). If you want to expand your song generation potential, you can always upgrade to the Pro Plan (2500 credits, $10/month) or the Premier Plan (10000 credits, $30/month).
Udio is a competitor to Suno. It offers much of the same features, such as extending tracks and picking where in a track to start an extension, while following a different methodology. You still input a prompt and instructions, and you still get two outputs. But instead of two-minute songs, Udio gives you 30 second snippets. You can then extend these snippets 30 seconds at a time, up to the 15-minute song duration limit, crafting an amazing track piece by piece. This method of generation is closer to the traditional experience of making music, giving you more control over the process.
If you ever create a track that’s mostly fine but has a few moments that need some work, Udio has an audio inpainting feature that could be of use. It lets you target specific parts of a song and make minor changes, such as fixing glitches or correcting lyrics.
Like Suno, Udio has a free plan that allows for a modest amount of song generations a day. You can pay for the Standard ($10/month) or Pro ($30/month) plan to generate more songs and access extra features.
Artificial intelligence is very much what you make of it when it comes to music production. It won’t make you an expert composer or musician, but it can certainly aid in your creative process. Hopefully these tips and tools will help you integrate AI into your process in an effective way.
Let us know down below if any of these tools help you. And tell us about any other AI tools you’ve added to your repertoire.
