Mechanisms Determining Crispness and Its Retention in Foods with a Dry Crust - Food Colloids: Self-Assembly and Material Science - page 494

distance between the successive signals. 40 We have investigated if this effect also
plays a role in determining the low frequency part of the sound signals emitted
on fracture of a crispy product. 7 To do this, we have manipulated the measured
sound signal recorded during the cutting of a toasted rusk roll at 0.2 mm s 1 .
The signal contained 30 separated sound events in a time interval of 4.5 s.
Random noise was removed using Sound Quality software. Next, using Sound
Forge software, the silent periods between the sound events were shortened
without affecting the properties of these sound events. It resulted in a sound
signal of 100 ms containing the same 30 separated sound signals. The FFT
analysis was done on both the 4.5 s and the 100 ms sound signals. The results in
Figure 4 show that the overall shape of the frequency spectrum at higher
frequencies (roughly above 1 kHz) was not affected, but the shortening of
the interval time clearly caused an increase in peaks at lower frequencies (below
1 kHz) and the appearance of peaks at the harmonic frequency of the occur-
rence of the sound events (300 Hz). This shows that the calculated frequency
spectrum below 1 kHz depends primarily on the number of separate sound
events and much less on the properties of the single sound events. This implies
that, for cellular solid crispy foods, the low frequency part is primarily deter-
mined by the morphology and architecture of the product, and only to a lesser
extent by the solid matrix properties.
Figure 4 Frequency spectrum of sound emitted during slow cutting at 0.2 mm s 1 of a fresh
toasted rusk roll: (A) sequence of 30 fracture events in 4.5 s; (B) same sound with
reduced silent periods in between the sound events (30 events in 100 ms)