A fork of Audio spectrum Analyzer for Android (See README.old for its original readme)
This software shows the frequency components' magnitude distribution (called spectrum) of the sound heard by your cell phone. Can be used to help tuning musical instrument or tone in singing, (tentative) measure environmental noise and sound revent education or experiments.
You can install this app through Google Play Store: Audio Spectrum Analyzer. Comments are welcome.
This software, Audio Spectrum Analyzer for Android, is released under the Apache License, Version 2.0.
- Show spectrum or spectrogram in real-time, with decent axis labels.
- Linear, Logarithm and (Musical) Note frequency axis support.
- You can put a cursor in the plot, for measurement or as a marker.
- Easy gestures to fine exam the spectrum: i.e. pinch for scaling and swipe for view move.
- Show peak frequency in a moderate accuracy (FFT + interpolation).
- Show dB or A-weighting dB (dBA), although not suitable for serious application.
- Possible to take averages of several spectrum then plot, make the spectrum smoother.
- You may record the sound (while analyzing!) to a WAV file (PCM format). Then you can deal with it with your favorite tool.
- Support all recorder sources except those need root privilege (see list in Android reference: MediaRecorder.AudioSource)
- Support all possible sampling rates that your phone is capable. e.g. useful to find out the native (or best) sampling format for you phone.
- Android version (API level)
- >= Android 2.2 (API level 8), up-to app version v1.6.8.
- >= Android 2.3 (API level 9), master branch, up-to year 2017.
- >= Android 4.0 (API level 14), since year 2018.
- These minimum version requirement is due to the "targetSdkVersion" and corresponding supported version range for library "com.android.support:support-v4".
Permissions:
- Microphone, of course.
- External storage (e.g MicroSD card), if you want to record the sound.
git clone then open it use Android Studio. Install the SDK platform if requested (e.g. rev 116 needs API level 20), or tune the compileSdkVersion to the value that fits your needs.
(Notes on project import for old revision (rev <= 115))
The whole program structure is roughly follows the MVC model:
AnalyzerActivity.java is the controler, as the main activity, it receives user inputs and system events, then sent corresponding commands to views or sampling and analyzing procedures.
AnalyzerViews.java is the view in MVC. It is used to manage (initialization, display, refresh) UI texts, buttons, dialogs and graphics. AnalyzerGraphic.java is a main sub-view which manage display of spectrum(SpectrumPlot.java) and spectrogram(SpectrogramPlot.java).
SamplingLoop.java is more or less the "model" part. It performs the sampling and FFT analysis, and inform the graphics update.
The data process loop is located in run() in SamplingLoop.java (after commit c9e430b (Feb 06, 2017), but basicly this process didn't change since the initial commit), as well as the trigger of graphics refresh.
In every loop of while (isRunning), it reads a chunk of audio samples by
record.read(audioSamples, 0, readChunkSize);
, then "stream" it to STFT.java by
stft.feedData(audioSamples, numOfReadShort);
which calculates RMS and FFT whenever enough data is collected. The view is then informed through
activity.analyzerViews.update(spectrumDBcopy);
which ultimately calls invalidate() of the graphic view to request an update, then the AnalyzerGraphic.onDraw(Canvas c) will be called automatically.
The code Audio spectrum Analyzer for Android gives me a good starting point, for learning Java and write this software (that I desired long ago).
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent