MixScope (Alpha)

Mixscope is a tool built to answer one simple question: Does my mix sound good?

It achieves this through a two-step process. First, it extracts local audio data using FFmpeg. When you drop an audio file into the app, it reads the data and calculates scores for core mix characteristics like Loudness, Spectral Masking, Stereo Width, and Dynamic Punch.

However, raw numbers rarely tell the whole story. Standard audio tools measure dynamic range, frequency balance, or loudness against rigid statistical baselines, but music is fundamentally contextual and creative. A track might score poorly on standard mix metrics—showing excessive low-end saturation, reduced dynamic punch, or unconventional stereo phase alignment—simply because the producers deliberately pushed those boundaries to achieve a specific vibe, energy, or aesthetic character. By integrating Gemini, Mixscope moves beyond simple numerical output to evaluate the intent behind the metrics, providing context-aware mix feedback that respects artistic choices rather than blindly penalizing them.

I understand that many creators are skeptical of AI in music production, but this integration serves a completely different purpose. The AI isn't mixing or mastering your track—it simply interprets the raw DSP data and explains why certain numbers might deviate from typical baselines.

I designed this tool to work seamlessly with a free Gemini API plan. The application only sends the extracted JSON metrics to the API, and all audio processing happens locally on your machine—your audio files never leave your PC.

You can download the alpha version to test it out. Please share your feedback and thoughts once you've had a chance to try it!

How to use:
Drop your audio and in the textbox below write your description of the track and question for Gemini. For example "This is a fast paced orchestral music for a boss fight. Check the accurancy of the scores and update them to fit the track's genre and usecase"

You can ask whatever you want and describe your track in any way you see fit. Next if you plan to use Gemini, you have to paste your API key. Then press analyze audio to get the scores. And then click on Ask Gemini to get the answer from Gemini. I will post a video tutorial asap.