Artificial Intelligence is here to stay. The first platforms that allowed access to them were ChatGPT for and Midjourney to generate images. Since then, there have been many alternatives to these that have reached the market. But, in addition, we have also seen how the use of AI is much broader and we can also use it to create compositions without having any musical knowledge.
We are talking about Riffusion. Riffusion is an Artificial Intelligence platform that uses the basis of Stable Diffusion (hence its name) to generate a spectogram based on a text, a text that must include the type of composition that we want to create and that this AI will later analyze.
How Riffusion works
The first thing to keep in mind when starting to use Riffusion is that it is a platform that is still under development, so it has a lot of room for improvement. This platform is available through the following .
To start creating our own melodies, the first thing we must do is write the text that describes the sound what we want to create The simplest thing is to use the name of the instrument that we want to use or the type of music that we want to generate.
Next, we must specify duration that we want it to have the rhythm to create (by default it is set to 8 seconds, but we can extend it up to a maximum of 10 and a minimum of 5 seconds) and in Negative prompt, we can add what we want to avoid, that is, the parameters that we don’t want the AI to take this into account when generating the composition.
Next, we click on Get a new spectrogram and expect both the audio and the spectrogram to be generated. Once we have found the composition we are looking for or if we simply like it and want to use it in a project or keep it, we can download the created audio.
To do so, we just have to click on the three vertical points located to the right of the playback control. We can also save the generated spectrogram image using the right mouse button and selecting the Save As option. The file is not downloaded in .MP3 format, but rather uses the Windows .WAV format, a format that takes up much more space than traditional MP3s.
As it is content generated by an AI, at the moment, it is free of copyright, so we can use it without any problems. As we have commented above, this AI is still in beta phase and the results, depending on what type of composition we want to create, may have a metallic tone, a tone that, using applications to reduce or practically eliminate using filters.
Facebook Comments
Disqus