The Client’s Challenge
A recent client, a producer and educator, embraced the new wave of AI music tools, specifically Suno AI. His goal was to generate a piece of techno, a genre close to my own heart. He entered his request—’make me some techno’—and waited for the magic to happen. The result was, in his words, disappointing. It was technically techno, but it lacked the specific spirit he was imagining.
He tried to refine his request, describing what he wanted as ‘more on your shoulders at a festival type techno.’ This is a wonderfully evocative phrase. As a human collaborator, I know exactly what he means: the driving, immersive, slightly euphoric feeling of a peak-time festival track. For an AI, however, this is an impossible instruction. It’s an emotional cue, not a technical one.
The frustration was palpable. Here was a powerful creative engine, capable of generating entire songs in seconds, yet it couldn’t grasp the subtle, human feeling he was trying to capture. It’s a common and completely understandable hurdle when working with this new technology. You’re not doing anything wrong; you’re simply speaking a language the machine hasn’t learned yet.
Diagnosis: The Art of Prompt Engineering
The root of the issue isn’t user error or a limitation of the AI. It’s a Contextual Conflict—a mismatch between human creative language and the AI’s literal interpretation. Suno isn’t a mind-reader; it’s a highly skilled but very literal ‘session musician’. It needs a specific kind of direction, a discipline known as Prompt Engineering.
Think of it this way: you wouldn’t walk into a recording session and tell the drummer to ‘play something a bit more Tuesday afternoon.’ You’d give them a tempo, a style, a reference, and describe the dynamics. You’d say, ‘Give me a tight, four-on-the-floor kick at 128 BPM, with a driving, open hi-hat pattern, similar to early Richie Hawtin.’ One is a feeling; the other is a set of actionable instructions.
The Core Problem: Vague vs. Abstract
- Vague Prompt: ‘Techno.’ This is too broad. It’s like asking for ‘rock music’ and being surprised you didn’t get progressive metal.
- Abstract Prompt: ‘On your shoulders at a festival.’ This is too emotional. It describes the desired effect on the listener, not the musical elements required to achieve it.
The solution lies in translating the abstract feeling into a set of specific, concrete musical descriptors that the AI can understand and execute.
The Fix: Becoming an AI Music Director
We approached the problem not by trying to ‘fix’ the client’s input, but by building a new workflow for communicating his ideas. It’s a three-step process of translation.
-
1
Deconstruct the Vibe
First, we translated the feeling of ‘on your shoulders at a festival’ into musical ingredients. What does that actually mean in musical terms?
- Genre: Driving, hypnotic, minimal techno. Berlin-style.
- Tempo: Around 130-135 BPM.
- Instrumentation: Pounding 909-style kick drum, hypnotic bassline, sparse atmospheric pads, metallic off-beat hi-hats.
- Structure: Long, evolving build-up, adding elements every 16 bars, subtle filter sweeps.
-
2
Use an AI to Prompt an AI
This was the key insight. Instead of trying to perfect the prompt ourselves, we used a general-purpose Large Language Model (like ChatGPT or Gemini) as a ‘prompt writing assistant.’ We gave our human description to the LLM and asked it to create a prompt for a music generation AI.
Our Prompt to the LLM: “Write a detailed prompt for the music AI Suno. I want to create a track that feels like being on someone’s shoulders at a big music festival. It should be a hypnotic, Berlin-style minimal techno track around 132 BPM, with a heavy 909 kick, evolving synth pads, and a very driving feel.”
-
3
Execute the Refined Prompt
The LLM returned a beautifully structured prompt, rich with technical keywords that Suno could understand. It included details about the production style, the specific synth sounds, and the rhythmic structure. We copied this new, detailed prompt directly into Suno, and the results were instantly much closer to the client’s original vision. He could then iterate on this successful prompt, tweaking individual elements to get exactly the track he imagined.
Additional Reflections
The Creator as Creative Director
This case highlights a fundamental shift in the creative process. With generative AI tools, our role evolves from a hands-on craftsman to a creative director. Your most valuable skill is no longer just musical proficiency, but your ability to clearly and effectively articulate your vision.
The challenge is learning to give instructions that are specific enough to guide the AI, but not so rigid that they stifle its potential for ‘happy accidents’ and unexpected creativity. It’s a dialogue, and like any dialogue, it requires learning a new language. But as this case shows, once you become fluent, the creative possibilities are extraordinary.
If you are seeking professional help with AI music production and Suno AI prompting, one-on-one remote support and teaching services are available from Audio Support.
Not sure it’s the same problem? Ask a question first
