Optimising Bass Microphone Servo control

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
…it may be right in some sense, but quite wrong, misleading
Sure, how collections of data are presented can be manipulated to serve different purposes or even mislead.
Reminds me of the wise words of Homer Simpson, "Facts are meaningless. You could use facts to prove anything that’s even remotely true."

Straw man aside, we are talking about presentation of single measurements in the time domain or frequency domain. Suppose you are looking at an output signal on your oscilloscope and you see 2Vrms 100Hz sine wave that is shifted to the right by 2.5mS relative to your 1Vrms 100Hz input signal. This same scale and shift information could be presented in the frequency domain as +6dB point at 100 Hz with -90deg phase. Both representations correctly and unambiguously define what was measured. Since we are dealing with just one frequency, it is easy to move between time and frequency domain.

A dirac input pulse contains all frequencies at uniform magnitude and zero phase. A digitally captured response output to a dirac or delta pulse input will measure the time domain impulse response of the device you are testing. But how to get to the frequency domain since we are no longer dealing with just one frequency? The digital capture sampling rate and length determine a unique set of sine waves that when individually scaled and shifted will sum to create the measured impulse response. Plotting the magnitude and shift of these sine waves provides the frequency domain magnitude and phase response. There are very simple (but extremely repetitive) calculations to determine the scaling/shifting. They provide the desired path between the time domain and frequency domain representations of the digitally captured impulse response. The free DSP book mentioned in the other MFB thread walks through examples of how it is done.
Commercial motional feedback woofer available sort of

The same technique for converting impulse response to frequency response can be used for a digitally captured sine wave. The added benefit is that any harmonic distortion content will show up in the frequency domain plot even if it is too small to see on the oscilloscope.

…For sure, it needs lots of boundary constraints to be the true…
The constraints were provided…"a speaker with or without MFB" which covers your concern. You could add amplification and EQ/filtering in there and it would still be true. You might look at measurements in previous post; compare impulse and frequency response for C and D relative to B.


****
Just noticed your response
The attached impulse plots are hard to read on my gear.
The plots are large standard format uncompressed *.png image files…not sure I can do better.
Attached is a combined *.pdf file…lower resolution but maybe your gear will be happier.
If still no joy, contact me off list and we will figure something out.

Alternatively, you might try importing the *.wav files into REW for viewing.
 

Attachments

  • MFB_Compare_Plots.pdf
    145.5 KB · Views: 60
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.