BFD DSP1124P FIlter Chains

[t35tB0t]

Has anyone thought about daisy-chaining the Left and Right channel filter sets on the 1124P to setup a 24 filter EQ?

I know SNR will suffer with the double A/D conversion, but will it make a real difference? Note that the delay for a single channel is reported as a mere 1msec. So, the two channels daisy-chained will add a 2 msec delay. Not too bad?

Those extra equalizers might help with more complex room environments - ones with some notch-dropouts that need some narrow filter boosting. Finally, there is a lot of focus on getting our sub's responses smooth along a target line. What about the effect of all the equalization filters on phase?

When I built my satellites, I was extremely careful to setup both very flat response thru the cross-over as well as very very well match phase. It was quite a bit of work to get the first one right but it really made a difference. The human ear is very sensitive to phase - more so than level in some situations. My concern is that a wacky set of filters on the 1124P might generate a beautiful response curve insert horrible phase delay curve. Does this matter much below 100Hz?

[Ayreonaut]

t35tB0t wrote:

Those extra equalizers might help with more complex room environments - ones with some notch-dropouts that need some narrow filter boosting. Finally, there is a lot of focus on getting our sub's responses smooth along a target line. What about the effect of all the equalization filters on phase?

1. Those narrow notches are best left alone.
2. You should really take measurements at all of the seating locations and design filters to correct the average response.
3. Don't worry about phase, worry about FR in a room full of standing waves.

Once you look at the average response, and forget about boosting at troughs you'll find that half a dozen filters is all that it really takes.

[Otto]

Hi t35tB0t,

Yeah, chaining channel 1 to channel 2 on the BFD has been discussed and implemented by some. As Ayreonaut above indicates, it's usually not necessary to use that many filters, and you probably don't need to worry too much about narrow-band dips or humps.

As for the delay, the REW/BFD manual suggests adding an extra foot of distance to your pre/pro/receiver setup for distance to your sub. This will help account for the added delay incurred by the BFD's processing. I'd imagine that if you do any auto-setup, it will take that into account already since the BFD will be in place.

Of course, all the digital filtering will change the phase response of the signal. There is a medium-sized discussion about that around here somewhere. However, I'm not sure how much you will notice it. Most of us have focused on getting FR to where we like it, with a total disregard for phase. I not suggesting that disregarding phase is a bad thing, it's just the way it works out. I've considered measuring phase about the crossover points of my sub, but I have about 4,862 things that I have to do before I get to it.

Good luck.

[Ayreonaut]

Otto wrote:

I've considered measuring phase about the crossover points of my sub, but I have about 4,862 things that I have to do before I get to it.

You should adjust the distance or delay setting for your sub to minimize cancellation (maximize output) around the crossover frequency. The "phase" at that frequency is important.

[Otto]

Ayreonaut wrote:

You should adjust the distance or delay setting for your sub to minimize cancellation (maximize output) around the crossover frequency. The "phase" at that frequency is important.

Correct, I think that's a good empirical method to get it right. But if I had the time and energy, I'd like to see what's going on there as well.