I am looking for a clever person with experience with Java and Android and, preferably, experience with FFT or Fast Fourier Transforms (Frequency Detection of Recorded Audio), or enough knowledge , as for to be able to improve the Real-Time Frequency Analysis an Android App does.
This App has a public license, and I want to improve it.
You can download the .zip of the Android Studio Project of the App and run it on your cellphone so you get acquainted to it.
The problem to solve, is, that the Real-Time Frequency Analysis does not offer correct results mainly for I would say 70% of the times.
This means that, if, for example, I play a B note on the Guitar, 70% of the time it detects different notes that I am not playing, even with a volume that should make the analysis to give a good decent result.
The App has its own XX Times-a-Second Refreshing Text View that tells you both the Detected Main Frequency, and also, the Note (C, B, D, or whatever it would be) that belongs to that Frequency, alongside with a graph that shows visually all the audio analysis.
You can see a picture of the App working in the Attached Files, detecting a G3 Note.
So, I need the Real-Time Audio Analysis to give a correct result for at least 70% of the Time it is detecting Any Frequency (on a single instrument, one note at a time, from 20 to 4,000 Hz, and with a low ambient sound).
So, who thinks can improve the Real-Time Audio Analysis?.
I expect your applications.
Është zgjedhur fitues:
Use an adaptive filter to enhance the signal collected by the microphone before doing frequency domain analysis. Relevant Skills and Experience Adaptive filter, digital signal processing, Java, kotlin.