Hahaha, good point.

Let's throw in an if statement:
If sound travelled at the speed of light, we wouldn't have to worry about TA or phase issues due to speaker location. Too bad that isn't true...and I wonder if that would alter the propagation of the sound that it acts like light...I think that would be kinda horrid.

Adam, I wasn't going to be silly on this one. , but I'd have to suggest that you probably want to mount those mids more solidly, and not just the panel. That's just not kosher. [img]smile.gif[/img]

 

Absolutely Foof!

Seriously, I'm looking at making a set of molds to build the kick enclosures out of a kind of cement.

Really, I'm serious.

Adam

 

Adam,

You should have bought a truck, not a BMW.

 

It is getting pretty heavy...

 

I disagree and I agree.
Isn’t that the point of time alignment to delay the sound coming from the driver closest to the driver (the left channel is right thinking English speaking countries) so the sounds from the L-R channels arrive at the same time or at least closer to their 'time of conception'? I have been hitting the web and I am having a tough time finding the amount of time the commercial or car audio TA processors can actually delay the signal. A delay of 1 millisecond will essentially move the driver or group of drivers away from you 1.08 feet.
Currently we balance the center stage by equal wire lengths and careful placement of the drivers, we then adjust the balance to even out the SPL between left and right channels, and with TA, theoretically we can move the driver another 1 foot per 1 ms of time delay. This means moving the sweet-spot not the balance ****. But I do not know how much 'time swing ' is available to the processor. Also this TA stuff would work well with either a single channel driving the entire frequency spectrum (the woofs and tweets would have to be close together for max effect), OR an active channel (mids or highs or lows). Since all my drivers are at least 1 foot away from each other this could be a real soundstage saver IF it doesn’t add any sonic artifacts of its own.

All the ad copy on digital TA basically disses analog EQ and passive X-overs (BS IMO) but on paper TA does look like a real nice solution. I don’t know what real value the product has, since I have never heard one or seen the before and after results. In other words I will keep an open mind on this, though I think their Ad copy is nothing more than a tease to get you to chump out the big bucks because digit MUST be better than analog.


Pause for affect [img]graemlins/boring.gif[/img]

 

To complete this evenings sermon

I have been repeatedly referring to the use of TA to correct phase of a driver and that is and isn’t true I think. The TA will delay a complex signal a given period of time, a complex signal would be a multiple frequency (music) signal. By ‘moving the driver closer or farther away’ you affect the amplitude of the sine wave that hits your ears. By doing this I have been obsessing over phase correction. BUT music is a complex wave form not a single frequency, as you move one freq into phase you can be moving another out of phase. The problem would be due to the fact that a 20 Hz waveform has a wave length of ~ 16.6 meters and a 20K Hz signal has a wavelength of .0166m (.65 inches) so if you were to move the sound with time alignment one foot you would have marginal affect on the phase of the bass wave but almost 20 full cycles of the 20k Hz signal would have been affected. So as far as moving into or out of phase goes I was, technically speaking, pissing up a rope.

I hope this either helps

 

I agree.
I think the idea is to get the speakers in phase at the xover point, so they sum correctly.

 

What frequencies in a car are most sensitive to Phase ?


Anything between 20hz and 20 khz.

 

Few things........

you will never get the complete freq range in phase in the entire car. If you can do this you are a magician. Simply moving your ears a few inches can and will change the response charateristics of what you hear from your system.

Have you ever looked at a TEF reading. It will blow your mind with how certain freq's are delayed, out of phase ect ect. They if you could fix all those issue's....move the mic 3 inches in any direction and all those fixes means nothing because it looks just as bad as it did when you started.

Any type of processing has it's issues. What it can fix it will mess up somthing else. Every digital processor I have seen has preset values for TA steeps that can't be changed. so you compromise on which setting sounds as close as possible.


I have heard cars with processing coming out the **** that have sounded increadible. I have also heard cars like Earl Zuasmers BMW540 with no eqing, no processing other the x-over sound even better.

And again I heard the worst of both


Rule of thumb. Do everything you can in the design process to eliminate the need for any processing (even crossover active or passive) Then when you need processing you'll only need a little

Very hard to do

 

JohnVroom,

I was thinking along the same lines regarding time alignment and phase alignment. It is possible to have 2 drivers time aligned but not perfectly phase aligned. With regards to frequency response, yes you can align at the x.o. point, but then it gradually shifts as frequency changes.

Also, if anyone has ever seen an impulse response graph of a speaker, you'll understand that the waveforms reproduced by a speaker an not symetrical...that is that the positive signal does not get mirrored by the negative signal. So flipping the polarity of a driver and then time aligning another driver to match isn't going to give you totally coherent waveforms. It will get you close though.

The arrival of 2 waveforms could be exactly the same using time alignment, but deppending on the content of the 2 waveforms, it is possible for some cancelation to still occur.

In my mind, it's a compromise, and you arrive at it by testing, not by theorizing.

Of course, my line of thinking could be totally wrong. I just thought I'd ramble on some more...

Adam