• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

4 sub integration methods using Acourate

dallasjustice

Major Contributor
Joined
Feb 28, 2016
Messages
1,270
Likes
907
Location
Dallas, Texas
I have a stereo playback setup with four subs. My room is rectangular. Each of the four sub subs are placed at each midwall. IOW, a Welti four sub setup. Of course, each sub is mono and the crossver is a 2nd order NT.

The challenge is to find a method to best time align each sub to the R/L speakers. The following video demonstrates a few different methods to accomplish time alignment. The first method used was to simply line the impulses up by applying delay to various speakers.

The second method used is probably best called the "sine wave method". The goal of this method is to line the subs up with a theoretically perfect sinewave at the crossover. In theory, this method offers the greatest precision. As you will see, this method didn't workout that well. In the end, this method assumes that if a sub is precisely time aligned at the crossover point (which they were), the same sub will be time aligned overall throughout the low pass. As anyone can see from the group delay plots, that is not a valid assumption when it comes to very low frequencies. As Uli said, this was a good learning experience. It separates the theory from actual practice in room.

Finally, Uli was able to achieve excellent time alignment by playing each sub with the R/L speaker. He then rotated the delay for each sub based on the step response. In my setup, all subs play before the R/L. Without going into too much detail, my DAC over AES introduces enough latency that all subs need delay applied to be time aligned to R/L (including frontwall sub which is about 12' further away from MLP than R/L).

Uli and I tried this 4 sub setup a few weeks ago. It didn't workout that great. The reason was that I didn't take care to balance all subs and level match all four together to R/L.

All methods demonstrated herein require, in the end, all four subs be measured along with the R/L speakers. So it's important that all four subs play equally and combine to the same level as the R/L speakers. Before the video, I used REW preference tab to send low frequency pink noise to each sub and balanced them all with the volume knobs on each JL audio fathom sub. I then matched level for all four subs played together to the R/L. This preparation made a huge difference.

Overall, the bass is unassailable. I feel like I can listen to anything and the system won't break a sweat or show a flaw. Thanks Uli!
 
Last edited:
OP
dallasjustice

dallasjustice

Major Contributor
Joined
Feb 28, 2016
Messages
1,270
Likes
907
Location
Dallas, Texas
Couldn't get the video to play. Is it really 3.5hrs long??
It plays. Try going to this link.

It has 18 plays since I posted it on Acourate forum, so I'm sure it's working.
 
Last edited:

AJ Soundfield

Major Contributor
Joined
Mar 17, 2016
Messages
1,001
Likes
68
Location
Tampa FL
It plays.
It was my browser. I have Firefox set to stop all video playback without permission, something I find intolerable with many sites.
Worked fine with MS new "E"...but it's 3+ hrs!! Too nice a day here, will have to watch later! I did peek at your coke test Michael.:D. Interesting.
I'm guessing Uli is the Acourate guy? Wouldn't you have to delay the JBLs to match the front subs, then maybe delay the rear/side subs relative to the JBLs?

cheers,

AJ
 
OP
dallasjustice

dallasjustice

Major Contributor
Joined
Feb 28, 2016
Messages
1,270
Likes
907
Location
Dallas, Texas
It was my browser. I have Firefox set to stop all video playback without permission, something I find intolerable with many sites.
Worked fine with MS new "E"...but it's 3+ hrs!! Too nice a day here, will have to watch later! I did peek at your coke test Michael.:D. Interesting.
I'm guessing Uli is the Acourate guy? Wouldn't you have to delay the JBLs to match the front subs, then maybe delay the rear/side subs relative to the JBLs?

cheers,

AJ
Normally, one would expect R/L speaker delay. As I said in the OP, my external DAC over AES introduces latency.

My Lynx Hilo routes everything in the mixer. All subs go out the Hilo line and monitor outs. The Mola Mola DAC gets the Hilo AES out.

Yes, his name is Dr. Ulirich Breuggemann. I call him Uli. He's a great guy.
Acourate is very much geared to customization and precision. It's not designed for convenience. However, the results are always better than any other software I've used. It does take a while to use it though.
 

AJ Soundfield

Major Contributor
Joined
Mar 17, 2016
Messages
1,001
Likes
68
Location
Tampa FL
Normally, one would expect R/L )speaker delay.
What I meant was the subwoofer low pass filter itself creates a delay. With one behind(?) the JBLs on the front wall, this would create further delay, so it would have to be the JBLs delayed relative to the front sub rather than vice versa, as one would have with side/rear subs, which I assume are closer to the LP.

Yes, his name is Dr. Ulirich Breuggemann. I call him Uli. He's a great guy.
Acourate is very much geared to customization and precision. It's not designed for convenience. However, the results are always better than any other software I've used. It does take a while to use it though.
Looks like a very cool program, have heard good things, will have to check out some time.

It's a really nice day out...must get away from computer now....

cheers,

AJ
 
OP
dallasjustice

dallasjustice

Major Contributor
Joined
Feb 28, 2016
Messages
1,270
Likes
907
Location
Dallas, Texas
All filters are FIR. I don't use any filters inside my subs. There's no relative delays from filters.

Whenever you have time, you can see the step plots for each sub compared to R/L speaker. Their arrival time is obvious and can easily be corrected using the rotation function. The subwoofer plot is a different color than the R/L speakers. I'd recommend skipping ahead the the last 30 minutes. You'll see the method which proved most successful. It also happens to be the simplest method.


What I meant was the subwoofer low pass filter itself creates a delay. With one behind(?) the JBLs on the front wall, this would create further delay, so it would have to be the JBLs delayed relative to the front sub rather than vice versa, as one would have with side/rear subs, which I assume are closer to the LP.


Looks like a very cool program, have heard good things, will have to check out some time.

It's a really nice day out...must get away from computer now....

cheers,

AJ
 
OP
dallasjustice

dallasjustice

Major Contributor
Joined
Feb 28, 2016
Messages
1,270
Likes
907
Location
Dallas, Texas
One interesting factoid about the JBLs is that you can see that the tweeter is in negative polarity and the woofer in positive polarity. I'm sure there's a good engineering reason for this. It probably makes them better in the time domain. Since it's easier to integrate subs with tweeters in the same polarity as subs, I wired my JBLs backward so that the tweeter is now in positive polarity. That doesn't change the performance of the speaker but it does help with integration.
 

iridium

Addicted to Fun and Learning
Joined
Feb 28, 2016
Messages
525
Likes
114
One interesting factoid about the JBLs is that you can see that the tweeter is in negative polarity and the woofer in positive polarity. I'm sure there's a good engineering reason for this. It probably makes them better in the time domain. Since it's easier to integrate subs with tweeters in the same polarity as subs, I wired my JBLs backward so that the tweeter is now in positive polarity. That doesn't change the performance of the speaker but it does help with integration.

Please keep us updated on how this continues to progress!
Thank you,
iriidium.
 

Purité Audio

Master Contributor
Industry Insider
Barrowmaster
Forum Donor
Joined
Feb 29, 2016
Messages
9,114
Likes
12,303
Location
London
Fantastic, always grey here so I will have a chance to look at it later.So it that it or do you have plans ?
BW Keith
 
OP
dallasjustice

dallasjustice

Major Contributor
Joined
Feb 28, 2016
Messages
1,270
Likes
907
Location
Dallas, Texas
Get Lowe!
 

NorthSky

Major Contributor
Joined
Feb 28, 2016
Messages
4,998
Likes
942
Location
Canada West Coast/Vancouver Island/Victoria area
Lol :cool:
 
OP
dallasjustice

dallasjustice

Major Contributor
Joined
Feb 28, 2016
Messages
1,270
Likes
907
Location
Dallas, Texas
here is a sweep with Acourate filters in place. The FR chart is unsmoothed. The important part as far as subs are concerned is really the group delay plot. That will give the best information on how well they are integrated. The IACC is a artificial calculation of channel coherence. I doubt whether it's a very useful tool. I don't use it for anything.
test log sweep.PNG
 

AJ Soundfield

Major Contributor
Joined
Mar 17, 2016
Messages
1,001
Likes
68
Location
Tampa FL
I'd recommend skipping ahead the the last 30 minutes.
I did. It was a bit tough to follow, with 240db (!!) scaling and I thought I saw him say the same as me, the subs should be "arriving late" or some such, but the bottom line is that you are getting relatively smooth amplitude at the LP, so there's obviously some sort of time alignment at the mic.
If it sounds good, your work is done.:)

cheers,

AJ
 

nefilim

Member
Joined
Apr 18, 2018
Messages
52
Likes
13
Location
San Francisco, CA
hi @dallasjustice - my struggle is starting all over again having moved to a new house (open floor plans :rolleyes::rolleyes::rolleyes::rolleyes:).

i have 3 JL subs, i only have 2 channels available for them unfortunately, so one is cascaded from the other on the first channel and the third is on the second channel, using audiolense to generate convolution filters.

i'm curious about this part:

" Before the video, I used REW preference tab to send low frequency pink noise to each sub and balanced them all with the volume knobs on each JL audio fathom sub. I then matched level for all four subs played together to the R/L. This preparation made a huge difference."

doing the individual subs with their level controls, did you do a near field measurement or a measuring from the MLP? ie, taking room nodes into account or not? i'm guessing even with the Sub Pink Noise in REW room nodes will still have an effect on the level?
then for matching all the with with L/R - do you mean you adjusted the channel levels on your Hilo until all 6 matched ?
 

JohnPM

Senior Member
Technical Expert
Joined
Apr 9, 2018
Messages
344
Likes
919
Location
UK
As an aside, for symmetric sub placements in a rectangular room to be effective in cancelling room modes per the old Welti layouts they need to have the same output volume (at each sub), the same frequency responses and to be synchronous, no relative timing differences between the subs. The modal cancellations rely on the subs exciting the various resonances in antiphase, any differences in their outputs degrades the cancellation. Wall midpoint positioning is more sensitive to mismatches than corner placement.
 
Top Bottom