Hi All,

I wanted to test the psychoacoustic parameters of threshold in quiet of ISO AAC and for that I plotted the qsthr value and compared it with Terhardt's formula. I observed that the Threshold in Quiet(qsthr) values in ISO AAC for long and short window have a very strange curve as compared to Terhardt's formula(w/ 70 partitions + 44KHz+1024FFT, etc...).

I noticed that the the
- ISO AAC qsthr and Terhardt's curve start at 40dB SPL, but Terhardt's goes way below upto -4dB where as ISO AAC qsthr stays at 25dB
- ISO AAC qsthr curve is like a staircase, whereas is Terhardt's curve is smooth
- Why ISO AAC curve does'nt fall below 25dB?

Is there anyone who knows why there is such a major difference in the two curves?.

This is what I plotted ISO qsthr against.

Terhardt_qsthr=[
136.6 37.686 23.542 17.536 14.143 11.937 10.376 9.2072 8.2949 7.5605 6.9544 6.4443 6.0075 5.6282 5.2947 4.9979 4.7313 4.4893 4.2153 3.9195 3.6515 3.4044 3.1728 2.9521 2.7386 2.5292 2.3211 2.0767 1.7916 1.4962 1.1876 0.86335 0.47802 0.023718 -0.45674 -1.0107 -1.6354 -2.3176 -3.0293 -3.7239 -4.3351 -4.7817 -4.9796 -4.8616 -4.3696 -3.4997 -2.3499 -1.1059 0.02666 0.91228 1.5375 1.9895 2.3999 2.8729 3.4855 4.2971 5.3674 6.7966 8.6921 11.234 14.68 19.326 25.645 34.3 46.367 63.45 87.945 123.73 177.19 223.9]


Regards,
Nabeel

p.s. EAQUAL analysis reports that the Terhardt's curve gives bad quality. Subjective listening is not done as yet.