commands

to generate a signal and save it to a file
    ./ofdm filenameNoExtension
generates a file called filenameNoExtension.wav with the audio that plays

ofdmDecoder takes samples directly from stdin, so arecord can pull samples from the microphone like this
    ./ofdm filename; arecord -c 1 -r 44100 -f S32_LE | ./ofdmDecoder

This helps cut out the first second of audio while alsa actually prepares the mic amplifier
    arecord -r 44100 -c 1 -f S32_LE -d 1 > /dev/null; ./ofdm test & arecord -c 1 -r 44100 -f S32_LE | ./ofdmDecoder
    arecord -r 44100 -c 1 -f S32_LE -d 1 > /dev/null; ../build/ofdm test & arecord -c 1 -r 44100 -f S32_LE | ../build/ofdmDecoder

or, you can pipe samples directly stdout | stdin if you pass '-' as the filename to ofdm
    ./ofdm - | ./ofdmDecoder


You could playback a wav file generated by ofdm, shoving the samples directly to ofdmDecoder using ffmpeg to read the file
    ffmpeg -i 16QAM.wav -c copy -f s32le pipe:1 | ../build/ofdmDecoder
you can adjust the gain for raw files so it doesn't overwhelm the reciever
    ffmpeg -i test.wav -filter:a "volume=0.25" -f s32le pipe:1 | ../build/ofdmDecoder

or record a sound from speaker -> microphone for later playback using the above command
    ./ofdm bpsk_window32 & arecord -c 1 -r 44100 -f S32_LE bpsk_window32_recorded.wav

compare the synthesized signal to the one picked up on the microphone
    arecord -c 1 -r 44100 -f S32_LE -d 1 /dev/null; ./ofdm synthesizedSignal & arecord -c 1 -r 44100 -f S32_LE -d 2 recorededSignal.wav

record both the original synthesized waveform and the recorded resulting waveform along with running the decoder on it
    arecord -r 44100 -c 1 -f S32_LE -d 1 > /dev/null; ../build/ofdm synthesizedSignal_multiModulation_60s & arecord -c 1 -r 44100 -f S32_LE -d 60 recordedSignal_multiModulation_60s.wav;


audacity is a great program to sanity check the waveforms and recordings.


Big problem I just figured out was that the gain on my microphone was way too high, and it's not adjusted automatically, I have to manually adjust it with
    wpctl set-volume ## 0.25
to something that matches the environment. Not sure how to automate that yet
But the performance is much better with the mic gain lower. The waveforms don't clip nearly as much.


debugging with gdb

make a fifo
    mkfifo testFifo
    ffmpeg -i test.wav -f s32le pipe:1 > testFifo
then in gdb in another terminal (or with & at end of las command), run with input from the fifo
    (gdb) r < testFifo

it allows you to make a nice pipeline

this command uses ffmpeg to resample the result output of ofdm and saves a couple files before passing the resampled version off to ofdmDecoder
    ../build/ofdm - | ffmpeg -f s32le -r 44100 -i pipe:0 test.wav -filter:a aresample=osr=192000 test2.wav -y -filter:a aresample=osr=192000 -f s32le pipe:1 | ../build/ofdmDecoder
for testing the resampler


Generally how I do a transmission and record, decoding offline
    arecord -r 44100 -c 1 -f S32_LE -d 1 > /dev/null; ../build/ofdm - | ffmpeg -f s32le -r 44100 -i pipe:0 -y test.wav -f s32le pipe:1 | aplay -f S32_LE -c1 -r 44100 & arecord -c 1 -r 44100 -f S32_LE -d 120 test2.wav
then
    ffmpeg -i test2.wav -filter aresample=osr=192000 -f s32le pipe:1 | ../build/ofdmDecoder

or from file instead of gen
    arecord -r 44100 -c 1 -f S32_LE -d 1 > /dev/null; aplay synthesizedSignal_pilotPitch_100_seed_1.wav & arecord -c 1 -r 44100 -f S32_LE -d 120 recordedSignal_pilotPitch_100_seed_1_fading.wav