Three basic types of results were recorded: results with no oscillation, indicating that a resonance mode was not excited, periodic oscillations, indicating that a resonance mode was excited, and strange aperiodic oscillations or oscillations with unexpected tone, that indicate either strange parameter combinations or issues with the model.
The results of the first set of mode excitation tests are summarized in Table 2. Each set of parameters is used to generate a three-second simulation. When oscillation is produced, the value of ξequily for which the amplitude is maximized is indicated, as well as observations of the audio produced.
flip (Hz) | pmax (Pa) | ξequily for maximum amplitude (m) | Observations of sound produced | Audio example |
---|---|---|---|---|
60 | 2000 | 0.002 | Sounds like a pedal oscillation | |
100 | 2000 | No oscillation | N/A | N/A |
100 | 2500 | 0.002 | A bit lower than the first resonant mode of the trumpet. | |
150 | 2500 | No oscillation | N/A | N/A |
200 | 2500 | No oscillation | N/A | N/A |
250 | 2500 | -0.0001 | Sounds like a middle C | |
300 | 2500 | 0.002 | Sounds like a D4 | |
300 | 3000 | 0.002 | Sounds like a D4 | |
350 | 3000 | 0.002 | Sounds like an F, the third resonance mode of the B flat trumpet | |
400 | 3000 | 0.002 | Really weird high-frequency component to the sound | |
400 | 3000 | 0.0012 | Steady oscillation around a G4 | |
450 | 3500 | No oscillation | N/A | N/A |
500 | 3500 | No oscillation | N/A | N/A |
500 | 4000 | No oscillation | N/A | N/A |
550 | 4000 | No oscillation | N/A | N/A |
600 | 4000 | No oscillation | N/A | N/A |
600 | 5000 | No oscillation | N/A | N/A |
650 | 5000 | No oscillation | N/A | N/A |
700 | 5000 | No oscillation | N/A | N/A |
As shown in Table 2, most of the lip frequency and blowing pressure did not excite an oscillation for any of the ξequily values tested. Only three of the six resonance modes that Adachi and Sato report having excited were excited in these trials, and some tones were excited that do not correspond to resonance peaks. There were also some tones produced that had unexpected timbres.
To attempt to excite the specific resonance modes, the peaks in the impedance were selected using a peak detection algorithm, and used directly as the flip values, along with the blowing pressure values given in Adachi and Sato (1996). The results of this method are in Table 3 below.
Resonance mode | Impedance peak frequency/flip (Hz) | pmax (Pa) | ξequily for maximum amplitude (m) | Sounding frequency from FFT (Hz) | Observations of sound produced | Audio example |
---|---|---|---|---|---|---|
I | 86 | 2000 | 0.002 | 87.667 | No oscillation for the first two seconds; oscillation of a pedal tone begins around the last second. | |
II | 234.6 | 2500 | -0.0001 | 237.33 | Sounds around a B flat3, the second resonance mode, but the tone decays almost like a struck instrument instead of sustaining until the end of the simulation. | |
III | 352 | 3000 | 0.002 | 350 | Periodic oscillation around an F4, the third resonance mode. | |
IV | 471.2 | 3500 | No periodic oscillation. | N/A | N/A | N/A |
V | 593.6 | 4000 | No periodic oscillation. | N/A | N/A | N/A |
VI | 714.4 | 5000 | No periodic oscillation. | N/A | N/A | N/A |
Once again, the higher resonance modes are not excited at all, so that only the first three of six resonance modes produce oscillations.