Loudspeaker impulse response optimization

[thadman]

I'm interested in what factors affect the impulse response of a loudspeaker driver, specifically the relationship between the factors.

If we are interested in optimizing the impulse response of the loudspeaker system, could we simplify the problem by parameterizing the system and separating the rear chamber optimization from the driver design optimization or are the parts interdependent?

ie could the best rear chamber design be achieved without regard to the electro-mechanical parameters of the driver (assume driver is optimized)? could the best driver design (impulse response) be achieved without regard to the modal behavior of the enclosure (assume enclosure is optimized)?

[soongsc]

I am interested in what will be said on this subject.

[tinitus]

To me that would be the total phase behaviour of the complete speaker and crossover

[gedlee]

Since FR and impulse response are the same thing, optimize one and you have optimized the other. But impulse response is not everything, in a real room the impulse response has the room effects attached and these are just as important if not more important.

[soongsc]

Impulse response is only one way of looking at data. Normally the assumption is the system is minimum phase, and the output is single output. Due to the fact that sound radiation source varies with frequency, this makes things more complicated. So, for driver improvement, it is necessary to look at impulse variation with distance starting from standard near field measurements out to about 5 times diameter of the driver. After identifying cause of impulse response, then the solution can be tried.

[Aoxomox]

Hi thadman,

I think it is no easy to answer in general to your question.

The impulseresponse contains all information to create the frequency- and phaseresponse. Just optimizing the frequency response (e.g. absolutely flat spectrum) could end in a terribly locking phase response. So the idea of optimizing impulseresponse means optimizing both.

A typical bass / midrange driver can be modelled by the Thiel Small parameters. This assumes concentrated elements like a spring, the magnet, coil, moving mass and so on. This way it is possible to model the speaker as a fourpole transforming electrical voltage/current into sound preasure/velocity. By adding a cabinet arround the speaker you add a mechanical spring (compressible air) and for the bass-reflex system the air mass.
This model will work as long as your speaker cabinet is quite compact. An extreme case were the model is not that easy would be a long tube were the rear sound has quite a time to travel before it will be reflected by the next wall.

To answer your question, the enclosure will influence your driver and thus the impulse response of the system.
I am not quite sure what kind of optimizations of the enclosure you are thinking about.
As long as the cabinet is small according to the wavelength then reflections from inside can be omitted. For the higher frequency range this can be a problem without proper dampening. The driver generates soundpresure that will travel from the rear side of the conus to the backside wall, were it is reflected and leaves the cabinet through the conus. In rear traveling soundwave is fully reflected it will interfere with the sound just generated. But this effect will also be identified in the frequency spectrum and not only in the impulse response (comb filter effect).

[Ron E]

Originally posted by Aoxomox
I think it is no easy to answer in general to your question.

The impulseresponse contains all information to create the frequency- and phaseresponse. Just optimizing the frequency response (e.g. absolutely flat spectrum) could end in a terribly locking phase response. So the idea of optimizing impulseresponse means optimizing both.

Think of frequency, phase and impulse as the shadows of an object when lit by the Sun, moon and venus. It is all the same object. If you correct any one, you will correct all three. With a minimum phase filter, you cannot correct one leaving the others unchanged.

[Dr. Ram]

Originally posted by thadman

could the best rear chamber design be achieved without regard to the electro-mechanical parameters of the driver (assume driver is optimized)? could the best driver design (impulse response) be achieved without regard to the modal behavior of the enclosure (assume enclosure is optimized)?

If I understand what you're asking, my simplest answer would be No - for example, if to you "optimization" of impulse response (which could indeed mean different things to different people) meant reducing transients and therefore would most likely mean a "Bessel alignment" for the system, we would have to speak of the alignment of the system=driver+enclosure and not just driver or enclosure.
-Ram