How to Use Your Own Voice with Suno AI: Professor Bear Tutorial
This tutorial shows how to upload your own voice recording into Suno AI, choose a style, and use meta tags to shape it into a finished, structured song.
Most people think of Suno AI as a text-to-song tool: type a prompt, get a track. This tutorial shows a different starting point, feeding Suno your own spoken voice recording and shaping it into a finished, structured song through style choices and meta tags.
Starting from a real recording, not a text prompt
The workflow begins with an existing spoken-word recording, in this case, audio stripped directly from an earlier video. Rather than typing a description of a song, the process starts with an actual voice sample, uploaded as-is, and lets Suno work from that raw material. This matters because it's a different creative starting point than a blank prompt: the voice, pacing, and cadence in the original recording carry through into what Suno produces.
Choosing a style and uploading the audio
Inside Suno, the first real decision is style. For a spoken recording like this one, a spoken word style is the natural fit, since it matches how the source audio was actually delivered rather than forcing it into a musical style it wasn't recorded for. From there, the workflow is to add an audio upload, in wave format, save it, and select the option to create a cover from that upload. Hitting create runs the generation, and the result can then be reviewed.
What a raw upload actually produces
An important distinction shows up here: when you upload a recording and simply create a cover from it, Suno transcribes what it hears in the audio and turns that transcription into lyrics, but the output isn't really a structured song. It's closer to the original spoken content wrapped in musical backing. Getting something that actually sounds like a composed song, with a real verse-and-chorus structure, takes a further step.
Enriching the output with meta tags
That further step is meta tags. Adding tags to the generation gives Suno explicit instructions about structure and arrangement that a plain audio upload doesn't communicate on its own. One useful example is a duet tag, which automatically introduces a second, complementary voice, in this case a female voice, alongside the uploaded voice, enriching the overall arrangement considerably. Meta tags act as a layer of direction on top of the raw audio, telling Suno how to shape the material rather than just transcribe it.
Iterating toward a finished song
Getting a polished result isn't a single generation. The described process is playing a version, adjusting the approach, playing again, trying different styles and different tags, and repeating that cycle until the output actually holds together as a song. That iterative loop, generate, listen, adjust, regenerate, is presented as simply how working with Suno AI goes: there isn't a single correct prompt that produces a finished track on the first try.
Key takeaways
- Suno AI can start from an uploaded voice recording, not just a text prompt, letting you turn your own spoken audio into a song base.
- Matching the style selection to how the source audio was actually delivered, such as spoken word for a spoken recording, produces better starting results.
- A plain cover generated from an audio upload transcribes what it hears but doesn't produce a fully structured song on its own.
- Meta tags, including tags like duet, give Suno explicit structural direction and can add complementary voices to enrich the arrangement.
- Producing a polished result takes iteration: generating, listening, adjusting the style or tags, and regenerating repeatedly.
Who this is for
This tutorial is one of a planned series of short skills videos made for Humanitarians AI Fellows, covering practical tools that fellows need when they're just getting started on creative and technical projects. It's a useful starting point for anyone curious about using their own voice as raw material for AI-generated music rather than starting from a blank text prompt.
Full transcript(auto-generated, with timestamps)
[0:00]Right here. Uh this is one of a few videos we'll be making for fellows, the humanitarian fellows. We'll make real little short videos for sort of skills that you need when you start. Uh this will be one of many of those videos. Okay, it's Professor Bear here. I'm going to take one of the recordings from my one of my videos, upload it to Sunno, and see if we can convert it into a song. So, here's the recording of uh I just stripped the audio from a video. Hey, it's Professor Bear here again. I'm going to talk about the projects we have at Humanitarian AI. So, what we do here is we add a style.
[0:41]So, I'm just going to pick a spoken word style because I sort of speak words. We are going to add an audio. Upload. Here's my audio recording for that in a wave format. Hit save. We want to do a cover. We hit continue. And we just create create right. And then we listen to some. Hey, it's Professor Bear here again. I'm going to talk about the projects we have at Humanitarian AI. This is Microoft. Like the other projects, this is an agenic AI project. So, if you're going to be a Microoft fell feller, uh you're going to start by building little components. So components for um Microoft are
[1:50]Financial. You might write a component which looks >> and here's the uh uploaded voice with different lyrics. So when I uploaded the audio, it just took whatever it heard in the audio and transcribed it and made that into a quote unquote song, but it wasn't really a song. Here's more something structured as a song. Welcome friends. Gather around in here. A little secret for your AI here. You want your sunot tracks to truly shine. Welcome friends. Gather around in here. A little secret for your AI here. You want your sudo tracks to truly
[3:00]Shine. Welcome friends. Gather around and hear a little secret for your AI. Okay. So, I'll enrich that a bit by adding metatags. Metatags will allow me to do something like create a duet. So, if I add a duet tag, it will automatically put a female voice in there with my enhanced voice and then enrich the whole experience quite a bit. So, we'll see what that sounds like. Welcome friends. Gather around and hear a little secret for your AI ear. And finally after playing, iterating, playing iterating, try different styles, different tags, different this, different that. Uh AI is about playing iterating, playing iterating, playing iterating. This is what we get.
[4:02]>> Welcome friends. Gather around and hear a little secret for your AI ear. You want your sunot tracks to truly shine. Master meta tags, they're your gold mine. These clever codes, these magic words will structure songs and shape the verbs. Want to say it proud. Need a intensity. Shout it loud. They guide the flow. Control the feel. Transform your track with pro appeal. Tags will send your music reeling. So grab your notes, don't lag behind. The metatag song will blow your mind. >> Metatags, they do it all.
[5:13]From start to start to hem my sue, you just add the tag and watch it do. Yeah. I was sitting at my laptop late one night trying to get my suno song just right. Type the line, hit go, it came out flat till I learned the secret tags. Imagine that. Oh, meditate that little magic flakes. Oh, meditate. They'll jazz up all your drag. I said and two. Now I'm breaking through with a full strong groove. My song began to really move.
[6:26]Want a solo? No need to beg. Just typing all at a left to set the tone or triumph and feel at home. Oh, meditate. The little magic flex. Oh, meditate. They'll jazz up all your drag. >> Need just add the tag. >> If you want to sound kind of swag, you're a song will steal the show. Throwing rain, >> thunder sound when he went around with this tag. We'll make it
[7:36]Metag to set the scene sound and clean. Medit. Oh, meditate the little magic flame. Oh, meditate. Now watch the crowd go mad. So next time you craft an AI tune, don't just type a lyric too soon. Use your metatags big and small and you'll make the finest song of all.
More videos
2:08Bridging the Pixel Gap in Browser Automation.
2:23How One Narrow Safety Rule Can Make an AI Less Safe Everywhere Else.
2:04Why splitting a chunk from its document makes it retrieve for the wrong question
4:20Three You Can Take Back. One You Can't.
2:21Why a 50-turn agent pays for the same screenshot 35 times unless it caches the pixels
1:53