<aside>
<img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/e08fe2bc-0270-4c07-9712-020ac6e37e9c/344d0d49-169c-456a-aad9-0015ece6b9ff/Group_2865.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/e08fe2bc-0270-4c07-9712-020ac6e37e9c/344d0d49-169c-456a-aad9-0015ece6b9ff/Group_2865.png" width="40px" /> Home
Getting Started
Showcase
Use Cases
Best Practices
Roadmap
Beta Program
Feedback
</aside>
This guide outlines best practices for creating effective prompts in our Al-powered audio effects tool. By following these guidelines, you can achieve more accurate and desirable results when applying effects to your tracks.
Best Practices for Creating Effective Prompts in AI-Powered Audio Effects Tool
-
Be Specific
- The more detailed your description, the better the AI can interpret your desired effect.
-
Use Descriptive Language
- Employ vivid adjectives and comparisons to paint a clear picture of the sound you're after.
-
Reference Familiar Concepts
- Mention well-known environments, eras, or styles to give context to your desired effect.
-
Combine Multiple Elements
- Don't hesitate to layer different ideas for a more complex effect.
-
Quantity When Possible
- Use terms like "subtle," "moderate," or "heavy" to indicate the intensity of an effect.
Structure of an Effective Prompt
A good prompt typically includes the following elements:
- The Instrument or Sound Source
- The Desired Effect or Transformation
- The Context or Environment
- Any Additional Details or Nuances
Example:
"Make the [INSTRUMENT] sound like [EFFECT/TRANSFORMATION] in [CONTEXT/ENVIRONMENT] with [ADDITIONAL DETAILS]."
Examples of Effective Prompts
- "Give the drums a huge, stadium rock sound with plenty of room ambiance and a touch of compression for punch."
- "Transform the vocals to sound like they're coming through an old-school telephone, with a thin and slightly distorted quality."
- "Create a dreamy, shoegaze effect on the guitar with lots of reverb and a subtle chorus."
- "Add a slap-back echo to the vocals, reminiscent of 1950s rock and roll recordings."
- "Make the synth sound like it's underwater, with a warbly, muffled quality."
- "Give the bass guitar a punchy, tight sound with mild distortion suitable for funk or rock."
- "Add a lo-fi effect to the piano, making it sound like it's coming through an old vinyl record player."
- "Transform the electric piano to sound like a vintage Rhodes with some tremolo and a touch of drive."
- "Make the strings sound like they're playing in a lush scoring stage, with rich reverb and a wide stereo image."
- "Create an 80s-style gated reverb effect on the snare drum for that classic pop sound."
Special Effects
Dynamics
- "Apply heavy compression to the drums to give them a punchy, in-your-face sound."
- "Use subtle limiting on the master bus to increase overall loudness without obvious distortion."
EQ
- "Apply a low-cut filter to the vocals to reduce muddiness."
- "Boost the high-mids of the guitar to make it cut through the mix better."
Modulation
- "Use a slow, subtle chorus effect on the clean guitar to add shimmer and width."
- "Apply a fast, intense flanger to the synth for a sci-fi effect."
Time-Based Effects
- "Apply a rhythmic delay to the lead guitar to create a sense of movement."
- “Add a long, spacious reverb to the synth pad to create a sense of depth.”
Pitch
- "Create a subtle doubling effect on the vocals by adding a slightly detuned copy."
- "Apply octaver effects to the guitar solo to extend its range."
Distortion
- "Add a warm, tube-like overdrive to the bass for extra growl."
- “Apply heavy fuzz distortion t o the lead guitar for a garage rock tone.”
Common Pitfalls to Avoid
Being Too Vague
- "Make it sound better" doesn't give the AI enough information to work with.
Conflicting Instructions
- Avoid contradictory terms like "make it loud but quiet.
Unrealistic Expectations
- Remember that the AI works with the existing audio. It can't add new instruments or completely change the genre.
Overcomplicating
- While detail is good, try not to overwhelm the AI with too many simultaneous requests.

Built in Flotion.