Flipped Polarity Speaker Design and Harmonics-No Hope?

[Deleted member 17820]

I wish I had time to read all the acoustic and speaker design books I have (and have access to) so I could learn about in-polarity designs vs in-phase-at-xover designs.

I come from a live sound background, and it seems this is the equivalent, please correct me if I am wrong:

1. Take the sound from a source like a piano
2. Place a crossover point around 2200hz
3. Flip the top part of the crossed over signal polarity by 180degress
4. Use and all pass filter to phase align top and bottom signal in the x-over range

This would result in a smooth frequency response but would REALLY mess with things in the time domain. ANY sound that is "dipping" in the Xover range would be time-distorted, resulting/contributing, I believe to what I hear often as bad timing in speakers.

As a live sound engineer I know how important phase, off axis coloration and timing are to having a great sound.
A good example is mic'ing a drum kit. Using just a few mics in some well known configurations gives you the most "natural" sound partly because true phase of the source is best maintained. Start putting 6,11, and more mics on each drum kit piece and without proper mics, placement level, gating, you will always get a bad sound.

Bad sound is:

1. Timing, which includes attack and sustain
2. washy/splashyness
3. muddiness
4. comb filtering (not so relevant here)
The point is that proper phase relation of the source is vital to a natural sound, so why would speaker makes bake in such a thing whole band polarity flip?]

Maybe some would say the benefits outweigh the negatives?

 

[AnalogSteph]

The point is that proper phase relation of the source is vital to a natural sound, so why would speaker makes bake in such a thing whole band polarity flip?

Because they know that human hearing pretty much doesn't care about absolute phase outside of the bass range, as research has shown? And I mean, why would it - it's basically a frequency analyzer.

This is being done to line up relative driver phase in the crossover region; it's typically necessary for odd-order crossovers. Otherwise driver output would be 180° out of phase and you'd get one huge dip in the frequency response.

As I like to say, you hear what you hear but it may not be what you think it is.

Here's a good article by Rod Elliott:
Phase, Time and Distortion in Loudspeakers

Your drum kit example pretty much does not apply because in that case we are talking a bunch of microphones all covering the same frequency range. The problems arise because sound travel time from various parts of the kit to the mic varies depending on mic position. Once you have to mix together two or more mics in one channel, you might get one particular sound source perfectly lined up in both mics but another may be wildly out of phase and suffer comb filtering as a result.

Maybe it kind of does apply, but only in the crossover region where two drivers operate. Speaker designers are well aware of crossover dips as a problem area, and they are using various techniques to alleviate the problem (close placement of drivers or even coaxials, waveguides of the exact right depth, or DSP delay). You can easily see the results in horizontal / vertical spin data.

I would suggest getting familiar with the whole scope of presently established speaker performance measurements, they are pretty comprehensive. No need to reinvent certain round things.

 

[FeddyLost]

Maybe some would say the benefits outweigh the negatives?

For sure. Recently it was just impossible or very expensive to do proper time and impulse, especially in passive speakers.
So, it's all just price.
Also, I'm not sure that people really interested so much in proper impulse behaviour. It's very cool, but how often do you have "good impulse" in big orchestra or live band with sound reinforcement?