TextToSample is a free VST3 plug-in by Samplab that lets you generate AI music samples in a DAW. You don’t need to have access to the internet once it’s installed, nor do you need to have any programming knowledge to make use of it.
TextToSample is one of the first plug-ins to streamline the process, letting you generate and use AI samples directly in your DAW. For anyone who wants to experiment with AI music, this is an easy, free, and convenient option to explore.
What Is TextToSample by Samplab?
TextToSample by Samplab is a VST3 plug-in that can generate music samples using AI. Similar to manyfree AI art generators, it works by entering a short text prompt that describes what you want to hear. In turn, it generates an audio sample that reflects your description as best as possible.
AI models are being improved all the time, with each new release producing higher-quality audio that aligns better with the given text prompt. For now, you’re able to try ideas like “relaxed folk guitar” or “soft synth pad” and see what you think of the results.
If you’re unsure about the potential use of generating music samples in an instant, there are plenty of other creativeways to use AI in music production.
But for anyone who has been curious about AI generation, TextToSample gives you a free and easy way to test it out in your production workflow.
If you don’t have a DAW, you may still generate AI samples using the standalone version of TextToSample.
How to Generate AI Music Samples With TextToSample
To run TextToSample, you will need to have Windows 10 or higher, or MacOS 12 or higher.
Keep in mind, when you first use TextToSample, it will take time to download the AI model from the internet. This only happens once—after that, the plug-in will run locally off your computer.
Installing the TextToSample Plug-In
Here’s how you can install the plug-in:
Depending on what DAW you use, installing third-party plug-ins may or may not work automatically. If TextToSample doesn’t appear in one DAW, try testing it in different software.
Using TextToSample in Your DAW
The interface is relatively minimal with just a few core settings you can play with. Here’s how to create a sample using TextToSample:
That’s all there is to it. The results aren’t guaranteed to be great, but the good thing about this plug-in is it gives you a quick way to experiment with AI music generation inside a DAW.
TextToSample: Models, Training Data, and Licensing
Powering TextToSample is the free, open-source AI model previously released by Meta, MusicGen. Samplab wanted to build off this research to:
“…enable the producer community to experiment with state-of-the-art generative AI models. We want to show what’s currently possible with these models, especially when running them locally on your computer. TextToSample’s purpose is certainly not to replace human musicians, which it’s not capable of doing anyway.”
Whenmaking music using Meta’s MusicGen, you have several options for model weights. In other words, you can try out faster and more powerful AI models if you have the right hardware.
It’s worth bearing in mind that, at the time of writing, TextToSample uses only the “small” model weights from Meta’s lineup. In case you encounter limitations, this is likely one of the reasons why.
In terms of training data, Meta trained the MusicGen model on 20,000 hours of music, using a mix of licensed music from its own dataset, alongside data from Shutterstock and Pond5.
Experiment With AI-Generated Samples
Most of the fun in producing music comes from taking the time to play with different sounds and experimenting with arrangements. AI skips all of that to produce audio with just a few words.
Whether AI generation becomes a useful tool in their workflow or remains just a bit of fun, will be up to the people who use it. you may get a quick idea of whether this technology will benefit you by trying the TextToSample plug-in today.
