TOPIC:

The input to the DDRC-24 is TOSLink.
Dirac 1 is routed to Output1 (Left Front Speaker) and Output3 (Subwoofer)
Dirac 2 is routed to Output2 (Right Front Speaker) and Output3 (Subwoofer)

The front speakers are mounted to the wall on speaker stands (front of speakers ~13 inches from the wall and 5 1/2 ft above the floor) and are eight feet apart.
The listening position is about eight feet from the front wall. 
The subwoofer is about 2 ft and 3 ft from the left back corner of the room and about 2 ft and 3 ft from the listening position.
The back of the room has a six-foot wide doorway.

I used REW to observe the response of the speakers and decide how to set up the Xover filters.

The PEQ filters were cleared before running Dirac Live.

The Impulse measurements in Dirac Live show a 8 mSec delay. I am assuming this is the delay for subwoofer.
There are only Dirac 1 and Dirac 2, so the subwoofer does not have its own channel.

Should I manually set an 8 msec delay for Output3 (Subwoofer) or is Dirac Live compensating automatically?

ArneEOlsen wrote:

The Impulse measurements in Dirac Live show a 8 mSec delay. 

Could you clarify exactly what you are referring to here? A screenshot may help. This will not be a subwoofer delay. 

(I have a guess of what you're referring to but want to be sure before answering...)

This is from the Impulse tab after the room measurements have been made.

ArneEOlsen wrote:

This is from the Impulse tab after the room measurements have been made.

Thanks, that is what I was guessing. That difference reflects the delay introduced by the applied filter I think. It's not something you need to do anything with.

We should maybe go back a step though. When you were sorting out your main/sub crossover filters in REW, did this include optimising the time delays?

I tried to time-align the front speakers and subwoofer using align phase slope at cursor, but the times seemed much too large (26 ms), so I selected Xover points and left the delay at zero.
When Dirac Live measurements were done, the delay in miniDSP was set to zero.

ArneEOlsen wrote:

I tried to time-align the front speakers and subwoofer using align phase slope at cursor, but the times seemed much too large (26 ms), so I selected Xover points and left the delay at zero.
When Dirac Live measurements were done, the delay in miniDSP was set to zero.

If you upload your REW data I'll have a look for you if you like? I would look at this using a combination of manually looking at impulse responses and phase plots myself, rather than using any automatic feature. It's much easier to talk through a real example.

The .mdat file is contained in the zip file, to match the permitted file extensions.

Each measurement will have the parameters of the Xover.
The filter types and slope are constant, and the cutoff frequencies for:
the front speakers varied from 60 to 100 Hz 
the subwoofer varied from 80 to 130 Hz.
These measurements had PEQ filters applied.
The subwoofer Xover: HP 30 Hz BW 48 dB/oct and LP (80-130) Hz LR48 dB/oct
The Front speakers:  HP (60-100) Hz BW 24 dB/oct
These are Audioengine HD-6 front speakers 50 Hz to 22 kHz and S-8 subwoofer 27 Hz to 180 Hz

Subwoofer scanned from 30 Hz to 250 Hz
Front speakers scanned from 30 Hz to 20,000 Hz

Thank you in advance.

File size too large, subset of two Xover files
 

I can't see an attachment I'm afraid?

I took a subset of two files, which is attached to the previous post and to this post the average of the two front speakers and also the sub (no PEQ filters).

ArneEOlsen wrote:

I took a subset of two files, which is attached to the previous post and to this post the average of the two front speakers and also the sub (no PEQ filters).

I see the files now - I'll try to have a look over the weekend. The crossover filters don't tend to affect what the best delay is in my experience so a subset should be fine.

Just to check, were the measurements made with the same acoustic timing reference for all measurements, and with not Dirac time delays present?

Yes, measurements were made before using Dirac Live and zero delay set in miniDSP. 

Right, I've had a quick look but I'm afraid what you've measured there isn't useful. It is essential that the acoustic timing reference that I asked about above is used. This is a REW feature where a short high frequency sound is emitted from a chosen speaker to set the time reference for each of the actual measurements. You want to pick say the left main speaker and keep it at this while you perform three measurement in turn: the left main speaker on its own; the right main speaker on its own; and your subwoofer on its own. Do this with no crossover filters applied at all actually as the results will also be helpful in term of choosing what crossover frequency to choose.

Okay, I will make another set of measurements with the timing reference. It may be a couple of days, however. Thanks again.

ArneEOlsen wrote:

Okay, I will make another set of measurements with the timing reference. It may be a couple of days, however. Thanks again.

No problem. I myself have been trying to get to some measurements I need to take since before Christmas!