Audio Processing

Microphone Input

Analog Devices MEMS Mic Digital w/ Built in Amplifier
Might need to up sampling frequency to the max on Arduino/XMega in order to get decent output on even the Sparkfun microphone. To test tomorrow: try to have sample audio coming in on microphone and sent right back out through the DAC to headphones (or logged at a really fast pace to computer).

A fun looking project if the need ever comes up…http://www.sparkfun.com/products/8482

dspquietcomfort3.pdf Great MIT Master's thesis by Nathan Hanagami on creating a DSP version of Bose's Quiet-Comfort 2 Headphones (which were previously all-analog!).
- Used a Blackfin processor, but did some feature comparison between different chips. Was limited to fixed-point math, and Blackfin had hardware support for only fractions < 1 or integers > 1…which is pretty inconvenient! Luckily it had a floating point emulation mode, but that made his code run a bit slower.
- Didn't take the time to implement noise cancellation, but rather add appropriate sound boosting?
- Talks a lot about the appropriate algorithm designs for filtering
- Is this implementable on an ARM processor? I doubt it…

Aquaticus ROV Sine Wave Hands down best tutorial on sine wave generation and RC low-pass filtering.
NerdKits Hardware PWM tutorial

Great example is PolyAVR blog post Site w/ source assembly code
- Only 4 voices max. Compute wave data for each tone if enabled.
- Utilize MIDI ADSR (Attack, Decay, Sustain, Release) settings for appropriate timings and voicings of tones so they sound more real. Purely volume-only, so no frequency-changes needed.
- Then it's simply an additive process after that to determine what the voltage output should be at a given time.
- ANY INTERRUPTS NEEDED?
Also ELM-Chan has a great perl txt2asm melody converter. Everything else is in assembly though :(
Making notes "slide" would be cool too, but his way of queuing up melodies is all wrong!