TOPIC: Delay time of the Xover/PEQ block

Hi, This is Yuichi

My question regarding the 2x4 HD Xover and PEQ function is a s follows.
What is the delay time of those function block. The delay means the time delay form signal input to output, not the phase delay. Does the delay time depend upon the setting contents, may be. What is the magnitude? 1-2 msec or more?

My experience is this. I measured the 2 way system with traditional LC network. I found about 1 msec difference of time arrival form two speakers. Because the tweeter is the horn type.
Then, I implemented 2x4 HD utilizing Xover and PEQ. Then, the time arrival matches. This is before I implement FIR filter.

So, I thought that there are about 1 msec time difference between two channels, I mean Xover and/or PEQ has delay?.

Anyone have answer to me?

Yuichi

I am asking a different thread that the sound is different between the PEQ (minimum phase) by FIR and GUI's PEQ.
Since this is a case where PEQ is completely replaced with FIR's one, if there is a delay in GUI's PEQ, there is a possibility that the sound has changed due to this delay which makes time alignment shift.

I'm having a bit of trouble understanding your post; I think you're asking what the time delay of the PEQ and Crossover functions are.

The PEQ and Crossover functions are implemented using Biquad IIR filters, which have no delay associated with them. There is a pretty significant amount of reading on this site about biquads; if they're implemented correctly (as I'm sure they are) there shouldn't be a measurable delay. However, if you decide to use the FIR filtering, there will absolutely be a delay as the design software will typically shift the impulse by several ms (10.667 ms in my case) to achieve the desired response. Make sure you align your non-FIR channels to your FIR filtered channels.

Ummm...

I recently measured the delay IIR Xover On/OFF.

I input only 1 wave of 1 kHz sine wave to a certain channel then I turned ON / OFF 5 kHz LPF which do not effects to 1kHz sine wave.
when Xover was enabled, a delay of about 3 msec occurred. Of course the shape of the waveform is exactly the same.
This is thought delay is changed depending on whether processed by the Xover block or bypassed. How is it?

Yes, the miniDSP 2x4HD unit has an inherent 3mS latency.
When you program with any (IIR) crossover/PEQ blocks, they will have exactly the same delay characteristics as analog equivalents.
So, you will have the 3mS inherent delay, plus any delay associated with your particular xover/EQ settings.

When you program with FIR filters, it's a different ball game. Depending upon how you program, those might have much more inherent delay.

Hope that makes sense.

Dave.

Dave,

Are you saying that there is a 3ms delay if any biquads are enabled period? Does this delay increase as more filters are added or is it capped at 3ms?

There's a 3mS minimum delay no matter what. (Even with no biquads enabled.)

As you add filters the delay will increase further than the 3mS....depending upon the filter you activate.
IIR filters have a frequency-dependent delay.

Dave.