Audacity spectrograms of MZ2SYNTH outputs

[frankenbeans]

For Show & Tell today we will look at spectrograms of output files created by means of MZ2SYNTH.

MZ2SYNTH was used to generate three output files using the following for command:

for Z in 22050 44100 48000; do time ../SOURCE/mz2 -verbose -overwrite -x -advance 4 -sampling-rate $Z -output-file output_$Z.au -channel-select LMMM -r 0.05 example-02.ppm ; done ;

These files were then postprocessed with sox to add chorus and reverb effects:

for K in *.au; do OFN=$(echo $K | sed s/.au/.wav/) ; sox $K $OFN loudness chorus 0.7 0.9 55 0.4 0.25 2 -t reverb 75 75 100 lowpass -1 8000 dither ; done ;

The resulting WAV files were loaded into Audacity and viewed in spectrogram-mode to see what they look like. The 22050 and 44100 c.p.s. clips are shown below - the 48000 c.p.s. clip is very similar to the latter. Audacity operating in this way makes a very nice visualization tool for learning one's way around MZ2SYNTH - I can see that I have merely scratched the surface :-)

(1) Audacity Spectrograms ('Mel' Y scale)

(2) Fortran code that calculations the frequency in Hz of each "row" n

ELEMENTAL FUNCTION OscFreq(n) RESULT(f)
IMPLICIT NONE
REAL(KIND=RKIND) :: f
INTEGER,INTENT(IN) :: n
! --- VARIABLES ---
REAL(KIND=RKIND),PARAMETER :: rdenom=1.0_RKIND/REAL(N_OSC_PER_OCT,RKIND)
REAL(KIND=RKIND) :: x
! --- EXE CODE ---
x=REAL(n-REFERENCE_SMT_NUM*N_OSC_PER_SMT,RKIND)*rdenom
f=REAL(FRQ_REFERENCE_OSC,RKIND)*2.0_RKIND**x
! --- END CODE ---
END FUNCTION OscFreq