Thinking about FM distortion and speaker polarity

Pauln

Well-Known Member
This is one of those "get a cup of coffee" and put on your thinking cap kind of questions. I want to check my thinking on something and perhaps initiate a discussion about it. I'm going to index my ideas and thoughts so as to make analyzing this a little easier to consider and discuss.

1] Describing the recording polarity
Here I just want to start off by distinguishing some ways to record. I think some ways of recording are going to be more subject to the effect I have in mind. Some recordings, old ones, classical, jazz, blues, will be of the kind where all the instruments were played together and recorded at the same time, perhaps with one or just a few microphones, all channels having the same number of gain stages, so that the resulting signals were all the same polarity. Other recordings, more modern, processed, etc. will comprised channels of mixed polarity. I'm thinking that the effect I have in mind is going to be present with the recordings where all channels share the same polarity.

2] FM distortion
Frequency modulation distortion is the idea that a lower frequency modulates a higher frequency - that is, the geometric source location of the higher frequency is being accelerated back and forward by the lower frequency when both frequencies are played through the same driver... causing an irritating Doppler shift in pitch of the higher frequency.

3] FM distortion using independent signals
If one generates two independent electronic signals, one high and one low, and sums them, there is no FM distortion in the electronic waveform. But, when one plays them through a driver, there will be. The reason is because the signals are transverse waves, voltages, they sum linearly. The waves produced by the driver are longitudinal waves, acoustic, and there sum is occurring at the moving driver surface.
Another way to grasp this distinction is that the voltage signals' variation, being transverse, is orthogonal to the time axis, whereas the movement of the driver surface, being longitudinal, is in parallel alignment with "time" as represented by the waveform. This will make sense in a minute when I describe the microphone behavior.

4] FM distortion in microphones with independent speakers
Now let's generate the two independent electronic signals and play them independently each through its own amp and speaker (to represent two acoustic instruments). The microphone element is detecting sound as a longitudinal wave. This means that when the lower frequency sound presses the microphone element backwards, it is at that position getting an earlier and older part of the higher frequency wave-train (because it has had to travel further to get there), and likewise when due to the lower sound the microphone element is in its most forward displacement the higher sound detected at that point is a later and newer part of that wave-train.
This is just the reverse of the driver FM distortion polarity.

5] Assuming all is right so far, the implication is that for recordings that are "iso-polar" where all the music is recorded in the same polarity, one polarity of hooking the speakers is going to exaggerate FM distortion, and the other way is going to act to correct it.
To see why,
- microphone element moves back during low frequency
- high frequency captured is earlier older part of wave-train
- recording preserves this
- driver polarity matches original microphone polarity
- at corresponding point playback driver plays low frequency with position "back"
- high frequency reproduced is slightly older earlier
- high frequency has longer to travel to listener
- FM distortion is decreased by the original time misalignment of the microphone being reversed by the driver
(so now, if one switches the driver polarity the part of the high frequency that needs to be delayed and sent from when the driver is displaced back to actually emit from the driver when displaced forward - the wrong part of the high frequency wave train is being aligned to the low frequency wave).

Thoughts...

a] Back when recordings were mono, recording engineers were more likely to preserve polarity either by deliberate practice or by accident or simpler recording equipment?

b] With recordings that almost preserve polarity but for one gain stage reversal, one would like to have a "speaker polarity" switch in order to get the right configuration - a switch I have never seen nor heard of on any playback equipment.

c] Some people believe that the nature of instrument sound waves being asymmetrical suggests that they be played back in "absolute polarity", meaning that the part of the acoustical wave that is forward longitudinally is presented by the driver moving forward, and vise versa. In the case of thought "b]" above, this presents the choice of switching to favor absolute polarity for the sake of the correct asymmetry of the acoustic waveform or the other way to achieve the minimal FM distortion polarity.

d] Even for "modern" recordings that due to processing and engineering present the individual original channels all mixed up in polarity, might the "polarity switch" from thought "b]" result in hearing one way sound overall better than the other?

e] Specifications for FM distortion are done by injecting two electronic signals as described in "3]" above, never by acoustically capturing two independent signals as in "4]" above, yet it is the later that is how we hear it... have measurements of FM distortion been understated or incorrect?

f] Any technical folks have any ideas for a polarity switch design that takes into account that solid state amps tolerate open circuits but not shorts, whereas tube amps tolerate shorts but not open circuits?

Anyway, looking forward to thoughts of others... :)
 
What are you using for an antenna? As someone wrotesaid recently, a tuner is only as good as it's antenna. That's the starting point.
 
What are you using for an antenna? As someone wrotesaid recently, a tuner is only as good as it's antenna. That's the starting point.

So, you read it (or didn't read it) with or without the coffee, your thinking cap, or what?
 
Well, I am an Amateur Extra Class, but this thread is neither about radio nor antennae...
 
I think changing thread titles is problematic, but if it's possible you might consider changing this one to say "acoustic FM distortion". That might help to prevent over-excitement of FM tuner junkies like myself.

Too late!

I'm not sure I'm equipped with sufficient coffee or cap, and I've probably have given insufficient contemplation, but... Is one of the major ideas here that FM distortion due to microphonic pickup can be offset by the right choice of speaker playback polarity? If so, I think it needs to be taken into account that microphone diaphragm displacements & velocities are much smaller than for speakers, no?

Cheers,

chazix
 
I'm not sure I'm equipped with sufficient coffee or cap, and I've probably have given insufficient contemplation, but... Is one of the major ideas here that FM distortion due to microphonic pickup can be offset by the right choice of speaker playback polarity? If so, I think it needs to be taken into account that microphone diaphragm displacements & velocities are much smaller than for speakers, no?

Yes, basically. The relative proportion would be about the same but the absolute magnitude in the speaker would be much greater... so direct radiators would be overstating the correction, drivers for horns less so.
 
Back
Top Bottom