Ever feel like AI music generators are just a slot machine? You cross your fingers, hit ‘generate,’ and just hope for the best. I know I have. The lack of real, granular control has been the biggest hurdle. But I just watched a video from an AI professional that blew me away, because it looks like that’s finally changing.
This innovator showcased a new feature in the AI music tool Udio called Sessions, and it’s a total game-changer. It gives you a DAW-style (Digital Audio Workstation) editing space where you can actually shape your song with intention instead of just rolling the dice.
⚙️ Two Ways to Craft a Song
The video’s creator breaks down two awesome approaches for using this new tool.
1. The Full Song Approach (Easy Mode):
This is the fastest way to get started. First, the expert dialed in a perfect style prompt. He wanted a blues song from the perspective of a Pac-Man ghost (genius, right?), so he crafted a prompt like:
“Modern electric blues with raspy male vocals, gritty tremolo guitar, deep bass, and minimal drums. Swampy, dark, and soulful.”
Then, he generated the full song. Afterward, he brought it into Sessions to make small fixes. In the video, a line of lyrics came out mumbled. No problem! He just highlighted that section, hit “replace,” and Udio generated two new takes for him to choose from. So simple!
2. The Section-by-Section Approach (Pro Control):
For maximum control, the YouTuber showed how you can build a song piece by piece.
📌 He started with a 30-second clip he liked.
📌 He brought it into Sessions and used the “Extend” button.
📌 He added the next set of lyrics for the chorus and generated that section.
📌 He kept repeating this process: verse, chorus, bridge, and even a proper outro, shaping each part exactly how he wanted before moving on.
I think this method is incredible because you can make sure the whole song flows perfectly as you build it.
🚀 Pushing It to the Absolute Limit
This is where my jaw hit the floor. The creator decided to stress-test the platform by creating a song that changes genre every 10 seconds. Seriously.
He did this using the Style Reference feature inside Sessions. Here’s how this industry pro pulled it off:
✅ He started with a soft, ambient intro.
✅ He used “Extend,” but this time, he uploaded a different audio file as a style reference, a glitchy electronic track.
✅ He then jumped from that to country, then to metal, then to rap, and back again, all in the same song!
He even showed how you can adjust the “style strength” to either blend genres or make a hard switch. The transitions were shockingly smooth. Going from full-on country to heavy metal without a hitch? Insane.
The final song turned out so good that the mind behind it even created a full AI music video for it using Midjourney and Kling.
This is a massive step forward for AI music. We’re finally getting the creative control we’ve been waiting for. For the full deep-dive and to actually hear these wild songs, make sure to watch the original video from the creator!