Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[John Roberts {JR}]

In what application can you swap a simple broadband delay in place of an APF and get sufficient results?

Way back decades ago, when digital delay was too expensive to be practical, some premium loudspeaker crossovers used APF to generate small delays to be used for driver time-alignment. These days, especially when already in the digital domain, simple delay is almost free.

JR

[Frank Koenig]

In what application can you swap a simple broadband delay in place of an APF and get sufficient results?

I was thinking of phase aligning loudspeaker passbands in the crossover region. In a bi-amped system most folks just dial in the delay they need and call it a day. When using passive crossovers we don't have that luxury and analog LC all-pass networks can be used for this purpose. But I think it's rare, and correct me if I'm wrong, that all-passes are used for crossover alignment when delay is available.

There are situations when an all-pass is required, as when attempting to correct excess phase, as I alluded to. I fooled around (in simulation) with using an all-pass in place of delay for a phased subwoofer array but I never tried it in real life. In fact, for my little gigs, I'm back to omni subs since I don't need the directionality and they sound better.   -F

[Hayden J. Nebus]

I was thinking of phase aligning loudspeaker passbands in the crossover region. In a bi-amped system most folks just dial in the delay they need and call it a day. When using passive crossovers we don't have that luxury and analog LC all-pass networks can be used for this purpose. But I think it's rare, and correct me if I'm wrong, that all-passes are used for crossover alignment when delay is available.

There are situations when an all-pass is required, as when attempting to correct excess phase, as I alluded to. I fooled around (in simulation) with using an all-pass in place of delay for a phased subwoofer array but I never tried it in real life. In fact, for my little gigs, I'm back to omni subs since I don't need the directionality and they sound better.   -F

Makes sense, thanks for the illumination.

[Tom Danley]

In what application can you swap a simple broadband delay in place of an APF and get sufficient results?

I have not found them to be very useful actually as "normal crossover" above 6dB/oct already have an all pass phase response (where the LF is delayed relative to the HF) . What would be more useful (if it were possible with passive filters etc) would be an inverse all pass response which could cancel the phase / GD in the named variety of xovers.

More to the point, a number of the articles i have seen which implement them appear to have an improvement because they are not actually looking at the loudspeaker systems acoustic phase but instead assign a new time reference.  Acoustic phase as described by Heyser is the systems phase response once all the fixed and propagation delays are removed.
Best,
Tom

[Hayden J. Nebus]

Way back decades ago, when digital delay was too expensive to be practical, some premium loudspeaker crossovers used APF to generate small delays to be used for driver time-alignment. These days, especially when already in the digital domain, simple delay is almost free.

JR

It's easy to forget there was a time not too long ago when a few hundred kb of memory cost something.

[Ivan Beaver]

There are situations when an all-pass is required, as when attempting to correct excess phase, as I alluded to.

A number of years ago some of us stuck around after a Synaud con class while a demo of using all pass filters to "correct for excess phase".

When it was set up-the two different alignments (one with all pass and one without).

If you listened carefully you could hear "something different" but nobody would say if they liked one or the other or that one was better than the other or which one was more accurate.

[John Roberts {JR}]

It's easy to forget there was a time not too long ago when a few hundred kb of memory cost something.

Back in the '70s I paid $1,000+ for 32K of memory.

JR

[John L Nobile]

Back in the '70s I paid $1,000+ for 32K of memory.

JR

I now feel better about paying 450 for 64k ram for an apple 2e. But that was the 80s.

[Tom Danley]

Back in the '70s I paid $1,000+ for 32K of memory.

JR

Those were the days (i don't miss).
The first computers they bought for us at Intersonics in the 80's had a 386 processor & 387 math co-processor, they ponied up for  4mb of ram which was an entire full length card full of chips and cost a bunch.
I don't recall the cost but i think each computer was around 5 grand and ran "windows for work groups".
Now one can get 64gb on a usb thumb drive, how times have changed!
Best,
Tom

[John Roberts {JR}]

Those were the days (i don't miss).
The first computers they bought for us at Intersonics in the 80's had a 386 processor & 387 math co-processor, they ponied up for  4mb of ram which was an entire full length card full of chips and cost a bunch.

Yup my 32k memory was a full sized PCB for the DEC card frame.  This was before the PC, the 386 was 2nd or 3rd generation (after 186 and 286). IIRC my first PC was a 286.

I don't recall the cost but i think each computer was around 5 grand and ran "windows for work groups".

My computer cost me about $6,500 and there was assembly require (Heath kit version of DEC mini computer). Software back then was slim pickings. I had Basic programming language on it and wrote my own mail order business, inventory, and even a cute filter design program to help plot out actual response of cascaded multiple pole active filters, with real R and C values. I even used the tab function of my dot matrix printer to plot out crude frequency response curves. 

Now one can get 64gb on a usb thumb drive, how times have changed!
Best,
Tom

I put a DSP chip in a prototype tuner that has more computing power (and memory) than my first real computer, all in an IC smaller than my pinky fingernail and costing about $5.   

Moore's law rocks...

JR

PS: So much of modern audio gear is starting to resemble computers with specialized I/O... But they won't be able to miniaturize the speakers (too much).

[ProSoundWeb Community]