AVS Forum banner

1 - 20 of 34 Posts

·
Registered
Joined
·
1,156 Posts
Discussion Starter #1 (Edited)
I am looking for some advice on my current sub to main integration, after running M.S.O & the Dirac Live I decided to check with REW how the response was looking around the crossover point & got this response.




The null at 78Hz to me looks like poor sub to main integration so I decided to measure the subs & then the centre separately to see what I am working with.




To my untrained eye it looks like the centre channel is the cause of the null at 78Hz due to it's poor response?

I decided to try to usual trick of adjusting the subwoofer distance in my AVR to see if I could improve things, I measured from 0ms to 5 ms delay.


Whilst I was able to significantly reduce the null at 78Hz I seem to be negatively effecting the response in the 90Hz t0 110Hz range the more I increase the delay in the AVR.

I tried crossovers of 100Hz & 120Hz with delays up to 5ms to see if any offered a better response.




It seems the same negative on the response is caused no matter the crossover point, I look through all the responses I had collected I decided that a crossover of 120Hz at 0ms delay offers the best response.



My concern is that a crossover of 120Hz is to high, leading to bass being able to be located in the room.

So I have tried everything I can thing of other than moving the seating position which I don't have the luxury of doing or moving the mains but I am not sure that is going resolve the issue & will compromise their positioning in the room in relation to the seating.

My best guess is that if I could improve the response of my centre channel I would resolve the issue?

Sorry for the long post but I thought it would be best to explain what I have tried up to this point.

Adam
 

·
Registered
Joined
·
12,225 Posts
So I have tried everything I can thing of other than moving the seating position which I don't have the luxury of doing or moving the mains but I am not sure that is going resolve the issue & will compromise their positioning in the room in relation to the seating.

My best guess is that if I could improve the response of my centre channel I would resolve the issue?
Few things.

1) Many of us use with powered subs that have onboard DSP/EQ tend to need delay in the main channels to better match the delay in the sub. Which AV processor are you using? If it talks about speaker position in terms of distance, you might have to add meters of distance to the subwoofer, like 3 or 5 or more, whatever it takes. Adding distance to the sub adds delay everywhere else. Try that and see if you find a point where the nulls both are minimized.

2) The center response is indeed a problem, as all by itself it has deep dips around 130 Hz. Is it in a cabinet? Floor bounce?

3) Does your system have any kind of auto EQ?

4) Where is the sub (or subs) relative to the rest of the speakers?
 
  • Like
Reactions: AdamAttewell

·
Registered
Joined
·
1,156 Posts
Discussion Starter #3 (Edited)
Few things.

1) Many of us use with powered subs that have onboard DSP/EQ tend to need delay in the main channels to better match the delay in the sub. Which AV processor are you using? If it talks about speaker position in terms of distance, you might have to add meters of distance to the subwoofer, like 3 or 5 or more, whatever it takes. Adding distance to the sub adds delay everywhere else. Try that and see if you find a point where the nulls both are minimized.

2) The center response is indeed a problem, as all by itself it has deep dips around 130 Hz. Is it in a cabinet? Floor bounce?

3) Does your system have any kind of auto EQ?

4) Where is the sub (or subs) relative to the rest of the speakers?
Many thanks for you feedback & advice, it's greatly appreciated:)

1) My receiver is an Arcam AVR850, once I have ran Dirac live the speaker distances are expressed in milliseconds instead of feet. I guess this is because the distances are in the time domain now rather then physical distance? 1ms is around 1ft I believe.

I was adding distance to the subs in the AVR but it is expressed in milliseconds, it just seems that no mater how much delay I add I can reduce the null at 78Hz but then I negatively effect the response in another area. I did only try up to 5ms because the response was getting worse the more delay I added. Do you suggest trying even more delay? Or do you mean add delay to the front left, right & centre channels?

2) It's not in a cabinet but on a stand slightly tilted upward, I have included a couple of picture of my room. I guess floor bounce could be the issue? I could place an acoustic panel on the floor in front of the speaker to see if it makes any difference?

3) The Arcam has Dirac Live built in, I am also using a MiniDSP 2X4 HD to allow me to use M.S.O for my two subs

4) The front sub is next to the centre channel firing into the front wall as can been seen in the image. The rear sub is behind the rear row which you cannot see but this is firing into the rear wall.
 

Attachments

·
Registered
Joined
·
1,156 Posts
Discussion Starter #4 (Edited)
I am not sure what to try next so I did some testing in regard to the floor bounce you mentioned.

I took a measurement of the centre channel only (set to large) & just to be sure I switch off the subs, I then placed an acoustic absorbing panel on the floor in front of the centre channel & took another measurement, so the only difference between the two was the acoustic panel being in place.


These are the two responses.


Whilst it does not solve the problem it does seem to slightly reduce nulls at 129Hz & 202Hz. Problem is its not really practical to have an acoustic panel on the floor:(

I left the panel in place to see what effect it has when the centre & subs are summed together, I first tried 80Hz with 0ms delay.


Still a nasty null at 75Hz, so I started increasing the delay up to 5ms like I did last time & got the same result with the response from 90Hz to 120Hz being negatively effected by the increased delay.


I tried a 100Hz crossover up to 3ms but the same issue occurs. Here is a 100Hz crossover with 0ms & 3ms delay responses.



I did try a crossover of 120Hz at 0ms & whilst is greatly improves the null at around 75Hz the slight null at 126Hz is made worse.



From here I am unsure how to proceed, the only other avenue I am yet to go down that I know of is adding delay to the mains but having never done it before I don't know if that is a good or bad idea & what the consequences are.
 

·
Registered
Joined
·
12,225 Posts
Are you sure the subwoofer vs mains are in same polarity? What happens if you flip the subs?

I realize the measurement are from REW, but what does Dirac show for the C response before/after correction?
 

·
Registered
Joined
·
1,156 Posts
Discussion Starter #6 (Edited)
Are you sure the subwoofer vs mains are in same polarity? What happens if you flip the subs?

I realize the measurement are from REW, but what does Dirac show for the C response before/after correction?
Using the SVS app I switched both subs from positive to negative polarity. Here is both responses, positive & negative.



Here is the measured response in Dirac Live, I use the wide imaging seating profile which includes 17 measurements of my two rows of seating.


The testing I did with the panel in place I had Dirac turned off as the nulls were present in the response with or with Dirac enabled.
 

·
Registered
Joined
·
12,225 Posts
Using the SVS app I switched both subs from positive to negative polarity. Here is both responses, positive & negative.
Ok, the blue line makes it pretty clear you had that right. :)

Here is the measured response in Dirac Live, I use the wide imaging seating profile which includes 17 measurements of my two rows of seating.
Since the Dirac result is so different from the REW result, it suggests we should not consider the REW 1-point curve represents the same reality as the 17-point area sampling.

Can Dirac measure mains and subs together? Or is just each speaker individually (until their new bass management software is released)?
 
  • Like
Reactions: AdamAttewell

·
Registered
Joined
·
1,156 Posts
Discussion Starter #8
Ok, the blue line makes it pretty clear you had that right. :)

Since the Dirac result is so different from the REW result, it suggests we should not consider the REW 1-point curve represents the same reality as the 17-point area sampling.

Can Dirac measure mains and subs together? Or is just each speaker individually (until their new bass management software is released)?
I believe its just measures each speaker individually.

Putting Dirac aside for a moment, I guess there is nothing I can do about the poor response of the center channel above the crossover point other than moving the location of the speaker or the seating position?
 

·
Registered
Joined
·
12,225 Posts
Putting Dirac aside for a moment, I guess there is nothing I can do about the poor response of the center channel above the crossover point other than moving the location of the speaker or the seating position?
Well, if you believe Dirac, there's nothing to fix. If you don't believe Dirac, then turn off the Dirac EQ.

Earlier you said you added distance to the sub but it was expressed in ms. So just want to confirm, if Dirac is adding 5 ms to the sub, that might still be an issue. I'd want to try adding delay in the Center and see how that goes.

As an experiment I built an LR-4 crossover at 80 Hz in AudioMulch, then summed the outputs. Ran sweeps with REW.
With equal delays in both paths, the result is flat (blue).
With 5 ms delay in LP, there's a big dip at 86 Hz (red).
With 10 ms delay in LP, there are two dips, 63 and 140 Hz.

 

Attachments

  • Like
Reactions: AdamAttewell

·
Registered
Joined
·
1,156 Posts
Discussion Starter #10
Well, if you believe Dirac, there's nothing to fix. If you don't believe Dirac, then turn off the Dirac EQ.

Earlier you said you added distance to the sub but it was expressed in ms. So just want to confirm, if Dirac is adding 5 ms to the sub, that might still be an issue. I'd want to try adding delay in the Center and see how that goes.

As an experiment I built an LR-4 crossover at 80 Hz in AudioMulch, then summed the outputs. Ran sweeps with REW.
With equal delays in both paths, the result is flat (blue).
With 5 ms delay in LP, there's a big dip at 86 Hz (red).
With 10 ms delay in LP, there are two dips, 63 and 140 Hz.

I will have to run the sweeps tomorrow due to it being 1:00am here in the UK now.

So to clarify after I finish running Dirac its sets the sub delay at 0.95ms which still results in a null in the crossover region.

Do you want me to set the sub delay at 05.00ms & the centre delay to 05.00ms & run a sweep?
 

·
Registered
Joined
·
12,225 Posts
So to clarify after I finish running Dirac its sets the sub delay at 0.95ms which still results in a null in the crossover region.

Do you want me to set the sub delay at 05.00ms & the centre delay to 05.00ms & run a sweep?
I would start adding delay to the center, 1 ms at a time, while watching the SPL at 80 Hz. If adding delay starts reducing the null, keep going until it maximizes. Then run the REW sweep.
 
  • Like
Reactions: AdamAttewell

·
Registered
Joined
·
12,225 Posts
Leave the sub delay as it is.
 
  • Like
Reactions: AdamAttewell

·
Registered
Joined
·
1,156 Posts
Discussion Starter #15
So I played the 80Hz sine wave with the SPL meter running & 1ms at a time I increased the center channel delay.

At 1ms the SPL dropped a little but I carried on increasing the delay & the SPL increased.

I carried on until I got to 7ms & increasing it further results in no further gain in SPL.

I then ran a sweep at 7ms delay, I have included the response before adding the delay to the center channel for comparison. This is with the crossover set at 80Hz.

 

·
Registered
Joined
·
12,225 Posts
So I played the 80Hz sine wave with the SPL meter running & 1ms at a time I increased the center channel delay.

At 1ms the SPL dropped a little but I carried on increasing the delay & the SPL increased.

I carried on until I got to 7ms & increasing it further results in no further gain in SPL.

I then ran a sweep at 7ms delay, I have included the response before adding the delay to the center channel for comparison. This is with the crossover set at 80Hz.

One thing I see is that the output from 70-80 Hz is higher than any of the results when delaying the sub from 0ms to 5 ms.


That suggests to me the added delay in C is helping. But not if it must come at the expense of a dip from 50 to 70 Hz. Does that look any better at some lower delay? Or some higher delay?

Back to the center speaker, and still thinking about the dip at 68 Hz in the "center alone" plot. Does not appear to be floor bounce. But could it be SBIR from the reflection off the subwoofer? I think it will be important to get a better looking center curve before the combined center+sub curve will look better.

Can you temporarily move the sub so it's even with or behind the center speaker drivers to see if that affects the center-alone response?
 
  • Like
Reactions: AdamAttewell

·
Registered
Joined
·
1,156 Posts
Discussion Starter #17 (Edited)
One thing I see is that the output from 70-80 Hz is higher than any of the results when delaying the sub from 0ms to 5 ms.


That suggests to me the added delay in C is helping. But not if it must come at the expense of a dip from 50 to 70 Hz. Does that look any better at some lower delay? Or some higher delay?

Back to the center speaker, and still thinking about the dip at 68 Hz in the "center alone" plot. Does not appear to be floor bounce. But could it be SBIR from the reflection off the subwoofer? I think it will be important to get a better looking center curve before the combined center+sub curve will look better.

Can you temporarily remove the sub so it's even with or behind the center speaker drivers to see if that affects the center-alone response?
What I can do is remove the sub completely & then run a sweep with the centre channel set to large. I that a good idea?

Moving all the base trapping behind it is a big job so I would prefer not to do that.
 

·
Registered
Joined
·
1,156 Posts
Discussion Starter #18
So I completely removed the subwoofer from where it was located beside the centre channel & took a sweep with the centre channel set to large & Dirac enabled.

I then reinstalled the subwoofer & took another sweep making no changes to any settings.
 

·
Registered
Joined
·
12,225 Posts
So I completely removed the subwoofer from where it was located beside the centre channel & took a sweep with the centre channel set to large & Dirac enabled.

I then reinstalled the subwoofer & took another sweep making no changes to any settings.
Well, those cancellations aren't from the sub cabinet either.

BTW, is this with the center's crossover HP filter active? The sub and center alone plots ought to use the crossovers if they didn't already, at least when considering how they will blend together.

I think I'm out of ideas. Very sorry to not have come to any meaningful resolution.
 
  • Like
Reactions: AdamAttewell

·
Registered
Joined
·
1,156 Posts
Discussion Starter #20
Well, those cancellations aren't from the sub cabinet either.

BTW, is this with the center's crossover HP filter active? The sub and center alone plots ought to use the crossovers if they didn't already, at least when considering how they will blend together.

I think I'm out of ideas. Very sorry to not have come to any meaningful resolution.
For those measurements I had the centre channel set to large so no crossover is active in the AVR & the speaker gets the full range signal.

I get the feeling I have screwed up somewhere for some reason. I have taken quite a few measurements & they have been with Dirac on & off and am concerned that this has screwed with things

I would like to start over just to confirm I have not messed up, where would you suggest I start?

Do you think I should just disable Dirac & start over?

Also I don't know if I mentioned it before but I am running a MiniDSP 2x4 HD & am using M.S.O to create filters to EQ for my four seating positions. Could this be a possible cause to my issues?
 
1 - 20 of 34 Posts
Top