Transient perfect background help request.

[BobEllis]

This weekend I was playing around with my system, creating an active 3 way + sub system using http://www.thuneau.com/ Frequency Allocator crossover program. I managed to get a nice configuration then I made a "mistake." I switched on the Arbitrator, which uses some form of magic to make the crossovers transient perfect. Now I have to have transient perfect crossovers. The problem is that Allocator/Arbitrator require processing time and the delay messes up lip synch in my multi purpose music/HT system. If I fix the mains the surrounds are early, not a good effect.

While I have an engineering degree and had a fair amount of calculus, it was a long time ago and has not been used in decades. I can still handle basic calculus, but that's about it. Can anyone suggest some reading/self learning to get me up to speed on the math needed to understand and design transient perfect crossovers? I will go active, having plenty of quality amplifiers. In the past I just plugged the equations Linkwitz provided into a spreadsheet. Once you get into imaginary numbers my ability to create a spreadsheet formula gets left behind.

My main problem is I don't know where to start learning. Thanks in advance for your help.

[cjd]

Please don't take this as gospel as I may be quite wrong here.

I think the math is the same, the target is the key.

Without physical time alignment, it requires induced delay to keep a zero degree phase (or so) across the board. It should be trivial to do with a digital crossover, but it means everything is delayed to the "slowest" signal.

[BobEllis]

Thanks Chris. I don't really have a good handle on the math - I just crossed my eyes and read through Linkwitz' theory discussion pages to get to the handy dandy ready to use formula for the filter I want. Poles and zeros? Zero comprehension.

I was thinking to get the computer out of the equation, since it's noisy. I can make it quieter, though. Allocator can run multiple instances, as long as I have enough ASIO outputs to handle all channels. That may be the way to go. Let's see three way LCR, sub, two way surrounds, and I'll go passive for the rear surrounds, so I only need 16 outputs. Will have to check how many slots I have left.

[cjd]

I know MiniDSP will let you do this too. If they're noisy, it's a different kind.

Up till recently my media server was completely fanless - had to upgrade to be able to do 1080p streaming cleanly, not quite as quiet any more but I haven't tinkered with it to try to get it to shut up again. I think it can.

FWIW, I stopped trying to understand the math so I could explain it - I could have once, years ago, which is enough for me to get it at enough of a level to do what I've done so far... it may still be limiting me but hey - it's a hobby, so I don't care.

[BOBinGA]

Here is the best crossover primer I know of:

http://www.rane.com/note160.html

Not a lot of math, just plain English.


John K. used to have a paper on transient perfect 2nd order crossovers, but I can't find it anymore on his site.
He has a paper explaining transient response of different crossovers and there is a section on transient perfect 2nd order filters:

http://www.musicanddesign.com/Speaker_Transient_Analysis.html

A little more math, but you should be able to handle it.

-Bob

[BobEllis]

Thanks Bob