Real-time audio processing in Android

there is a sensing framework from MIT media labs called funf: http://code.google.com/p/funf-open-sensing-framework/
They already created classes for audio input and some analysis (FFT and the like), also saving to files or uploading is implemented as far as I've seen, and they handle most of the sensors available on the phone. You can also get inspired from the code they wrote, which I think is pretty good.

Using AudioRecord is overkill. Just check MediaRecorder.getMaxAmplitude() every 1000 milliseconds for loud noises versus silence.

If you really need to analyze the waveform, then yes you need AudioRecord. Get the raw data and calculate something like the root mean squared of the part of the raw bytes you are concerned with to get a sense of the volume.

But, why do all that when MediaRecorder.getMaxAmplitude() is so much easier to use.

see my code from this answer: this question

If you use MediaRecorder (the example, above) it will save compressed audio to a file.

If you use AudioRecord, you can get audio samples directly.

Yes, what you want to do should be possible.