Injected some broad-band noise in DHARD pitch, with a shape tailored to match closely the error signal above few Hz. Analysis will follow tomorrow.

Noise shape:
zpk([0;0;-6.2832-i*25.1327;-6.2832+i*25.1327;-6.2832-i*25.1327;-6.2832+i*25.1327;37.6991-i*62.8319;
    37.6991+i*62.8319;37.6991-i*62.8319;37.6991+i*62.8319;-157.0796],
    [-3.1416-i*5.0265;-3.1416+i*5.0265;-6.2832;-4.3982-i*18.8496;-4.3982+i*18.8496;-4.3982-i*18.8496;
    -4.3982+i*18.8496;-6.2832-i*18.8496;-6.2832+i*18.8496;-6.2832-i*18.8496;-6.2832+i*18.8496;
    -251.3274],1)
gain(0.0115462)cheby1("LowPass",4,1,100)

ampl 300
     1137997278 Wed Jan 27 22:21:01 PST 2016
     1137997409 Wed Jan 27 22:23:12 PST 2016
ampl 600
     1137997432 Wed Jan 27 22:23:35 PST 2016
     1137997556 Wed Jan 27 22:25:39 PST 2016
ampl 900
     1137997581 Wed Jan 27 22:26:04 PST 2016
     1137997706 Wed Jan 27 22:28:09 PST 2016
ampl 2000
     1137997738 Wed Jan 27 22:28:41 PST 2016
     1137997884 Wed Jan 27 22:31:07 PST 2016