The Carver Ribbons tested were not broken in. These were also the first tests we ran that day, so the curves show a little of the learning curve we had to go through on the IMP/MLS system. Most measurements were performed over 8 averages with an 4096 point FFT. Microphone calibration was used throughout all testing. These ribbons were tested without any notch filter installed The graphs show clearly the cavity resonances, typical for this construction.
Testing was performed in two sessions. One with a sample rate of 61.5 kHz or the medium and high frequency response. In order to obtain a more accurate display in the low frequency range, a second test was performed using a sample frequency of 1.92 kHz.. One third octave smoothing was applied in the following graphs because this allowed me to make direct comparisons with graphs that have recently been circulated over the list. Check here for a educational list of graphs depicting raw frequency date versus 1/12th octave smoothing (John's favorite) versus 1/3rd octave smoothing (my favorite).
The above graph shows the frequency response of the two Carver 60" ribbons that we tested. Testing was done in this case with the microphone at 1 meter, no notch filter installed for the 6.5 to 7 kHz resonance. The second graph below shows the low frequency response on a different scale. To obtain this response a new measurement was made at a lower sample rate. In both case were 8 measurements averaged.
The colors between the two graphs above do not correspond. The white and the yellow curve are from the same drive.