If SBIR Is At The Same Frecuency As Your Mains what do you end up hearing?

[Trdat]

I know its not often that your sub will play the same frequency response as your mains, but I get good results from overlapping lower frequencies according to where nulls are. Recently, I gave the flanking subs technique a try which on paper is meant to help with SBIR by using a similar concept of multiple subs but at a higher frecuency range.

I added a flanking subs to my mains which roll off like a closed boxed and the flanking subs with a low pass around 170hz, so both the mains and the flanking sub play around that 170hz but the sub which was next to the wall added SBIR between 70 and 80 hz which wasn't in my original measurements with my main speakers.

Irrespective of the SBIR, I am curious if the mains were playing 70 to 80 hz as per normal prior to adding the sub and the flanking sub introduced a major null highly likey related to the sub being not close enough to the wall and the measurement displays the null, are we hearing a null or the mains 70 to 80hz?

Essentially, is the null from the newly added sub woofer cancelling the flat reponse from the main speakers?

 

[ernestcarl]

Essentially, is the null from the newly added sub woofer cancelling the flat reponse from the main speakers?

Possibly. You can observe the phase and magnitude interaction of the sub and front mains set in the REW’s alignment tool. All you need is the IR of the sub only and the left+right only. Move the delay sliders around if there is no time reference and/or align the phases somewhere between the xo overlap.

 

[abdo123]

I think the idea to deal with SBIR with subwoofers is a correct one, but the execution is very poor.

basically what you want to do is that the first SBIR null (at 1/4 wavelength) of the mains you want the subwoofer to play instead, so that's dealt with.

the second SBIR peak (at 1/2 wavelength) you should bring down with EQ.

The 3rd SBIR null (at 3/4 wavelength) you should absorb.

For example, the speakers are 1 meter away from the wall / boundary, the first null will be at 86Hz, can be comfortably played by subs.

the second null will be at 258Hz, can be comfortably absorbed with 8 inch panels .

This is how you should go about things in my opinion. Any decently sized room can support speakers where the second null at 3/4 wavelength is absorb-able

 

[Trdat]

I think the idea to deal with SBIR with subwoofers is a correct one, but the execution is very poor.

For example, the speakers are 1 meter away from the wall / boundary, the first null will be at 86Hz, can be comfortably played by subs.

the second null will be at 258Hz, can be comfortably absorbed with 8 inch panels .

Your system to tackle SBIR is great, but just to confirm my understanding of SBIR I just need to understand it better to execute the solution.

When you talk about 1/4 then 1/2 then followed by 3/4 wavelnght are we talking about succesive SBIR phenomenans that occur as the frequency increases? And as the frequency increases we use a different solution to tackle that particular SBIR? Hence you have menioned the first null at 86hz and second null at 258hz do these correspond to the wavelenghts you mention?

basically what you want to do is that the first SBIR null (at 1/4 wavelength) of the mains you want the subwoofer to play instead, so that's dealt with.

Yes, but in my case its the subwoofer is what caused the SBIR not the mains. I am presuming that sub woofers close to the wall in essence should not introduce SBIR hence the idea that the subwoofer should play this frequency instead of the mains which usually cause the SBIR? This is why I am so curious what we hear, in my particular case the mains or the SBIR from the sub?

the second SBIR peak (at 1/2 wavelength) you should bring down with EQ.

To futher my understanding of this complicated wavelenght rules, is there a typical frequency range when you talk about the 1/2 wavelenght in a small room or I have I misunderstood this totally? I think from my understnading yes, each wavelenght corresponds to a typical frequency in a typical sized room? Like you mentioned above just want to confirm i have understood this correctly?

The 3rd SBIR null (at 3/4 wavelength) you should absorb.

Okay, I think I am getting this as the frequency increases absorbers will work, I have 2 times 15 cm thick 80cm by 60 cm absorbers that I can place nearly anywhere in the room, so is the only way of finding out where the SBIR is by a SBIR calculator? And I place the absorbtion at that point right?

 

[Trdat]

Possibly. You can observe the phase and magnitude interaction of the sub and front mains set in the REW’s alignment tool. All you need is the IR of the sub only and the left+right only. Move the delay sliders around if there is no time reference and/or align the phases somewhere between the xo overlap.

I do have a measurement of the sub which produces the SBIR left and right seperately.

So your saying I have to align the phases in the aligment tool? I will do some catching up and get back to this post.... Alignment tool is new to me.

But will this tell me what I am hearing is this the point of the alignment tool?

 

[abdo123]

Okay I'm going a step back, lets start with the basics. there are two things effecting your in room frequency response the room modes and boundary interaction.

you have the room modes, which are basically a quality of the room's physical dimensions. this is a quality that you cannot change unless you damp the room (fake walls filled with INSANE amounts of fibrous material) to the point where it doesn't matter what the dimensions are. each room has at least 3 room modes for each axis, called axial room modes. the first 3 axial room modes (the first for each axis) contribute 50% of the destructive energy in a room, so dealing with them alone brings massive improvements.

usually the first axial room mode is 1/2 wavelength of the axis's dimension. so a ~5 meter wide room would have a first-order room mode at 34Hz (10.1 meter wavelength). the second order (weaker influence, but still an influence) mode would be at 68Hz, and the third order mode would be at 102Hz.

It's important to keep in mind that not all room modes are equal, the axis of the room with concrete (high rigidity) walls on its boundary will be the biggest offender as concrete walls do not pass sound. so try to figure out (via regular in-room measurements) which room mode is the most destructive in your own particular room and listening position and try to position your subwoofers to combat it.

Which brings us to our next question, how do you position the subwoofers to combat axial room mode? well you need two subwoofers, and you need to place each subwoofer at opposite polarities of the mode (check above diagram and below). So theoretically, with two subwoofers, you can eliminate ALL the room modes, but they need to be raised off the floor / hanged off the ceiling for the height axis of the room.

With that covered, we move to the next issue, boundary interaction.

99% of speakers omit sound omni-directionally till a certain frequency (for most speakers it's between 400-1000Hz, depending on the baffle dimension). as the sound hit the boundary it reflects back to the listener (again concrete walls are the most offensive here and will bounce back more sound).

this effect can be enchancing (peak) or diminishing (nulls) at a particular frequency. this is determined by the distance of the speaker to the boundary (again all 6 boundaries, not just the back wall). there is usually a null at 1/4 wavelength of the distance between the speaker and the boundary, and a peak at 1/2 wavelength of the distance between the speaker and the boundary.

if the wall is infinitely rigid (concrete), the 1/4 null is usually infinitely deep, you cannot EQ it in anyway or form. so it's best that the speaker either does not play that frequency at all, or that you absorb it with panels.

However just like room modes we have harmonics / orders higher SBIR effects. (a null at 3/4, a peak at 4/4 wavelength .etc with progressively diminishing effects). I hope that makes my previous response a bit more clear, you need to figure out whether your nulls are room modes or boundary effects and deal with them accordingly.

 

[abdo123]

This is why I am so curious what we hear

you hear the driver that is not canceled via the phenomenon. it's straight forward.

but again do you really 'hear' drivers below ~120Hz?

 

[ernestcarl]

I do have a measurement of the sub which produces the SBIR left and right seperately.

So your saying I have to align the phases in the aligment tool? I will do some catching up and get back to this post.... Alignment tool is new to me.

But will this tell me what I am hearing is this the point of the alignment tool?

You certainly don't need to use the tool. But it makes it easy to visualize the interaction of both mains and subwoofer when summed at the listening position. It will tell you if there is going to be a negative interference between the two.

With regards to hearing the dip... well, that depends on how deep and narrow/wide it is.

 

[abdo123]

With regards to hearing the dip; well, that depends on how deep it is.

i think you mean how wide it is? (Q).

 

[Trdat]

you hear the driver that is not canceled via the phenomenon. it's straight forward.

but again do you really 'hear' drivers below ~120Hz?

Yes, point taken. But now I have to see if the flanking subs are actually providing me with any improvement other than the placebo of having another driver playing.

Regarding your great explanation on SBIR and room modes, I understand the basics but there are a few questions that will cement my understanding but I have been studying your post since you posted and my brain is fried. I will get back to once I am convinced I can't figure it out myself.

 

[Trdat]

You certainly don't need to use the tool. But it makes it easy to visualize the interaction of both mains and subwoofer when summed at the listening position. It will tell you if there is going to be a negative interference between the two.

With regards to hearing the dip... well, that depends on how deep and narrow/wide it is.

Great, I found a youtube video on the tool I will look into it.

And yes I suppose it has a lot to do with how deep it is and how wide it is.

 

[dasdoing]

I don't know why the easiest and most effective way to deal with subwoofer SBIR is so unpopular: just put it on the wall, and turn it around so the woofer is as close to the wall as possible. problem solved

 

[dasdoing]

Now I have thought of not high passing my mains which speak down to about 34Hz. the problem is that even aligning the mains with the sub, the group delays wont totaly overlap. phase issues are programed

 

[ernestcarl]

I don't know why the easiest and most effective way to deal with subwoofer SBIR is so unpopular: just put it on the wall, and turn it around so the woofer is as close to the wall as possible. problem solved

My own front wall resonates audibly and the FR is a tad bit more uneven that way.

 

[ernestcarl]

Now I have thought of not high passing my mains which speak down to about 34Hz. the problem is that even aligning the mains with the sub, the group delays wont totaly overlap. phase issues are programed

Why not just adjust the slopes to achieve the best fit by mixing and matching different HPF/LPF settings for either. If that doesn't work, manipulate them a bit with FIR phase EQ.

*Closing the ports of my S8 monitors helped as well.

 

[dasdoing]

My own front wall resonates audibly and the FR is a tad bit more uneven that way.

wow. now that is a solid argument against this aprouch. I would say that most walls in the world are not resonant like this though.

Why not just adjust the slopes to achieve the best fit by mixing and matching different HPF/LPF settings for either. If that doesn't work, manipulate them a bit with FIR phase EQ.

*Closing the ports of my S8 monitors helped as well.

you managed to overlap group delay?

I also have another reason to not go that path. I somehow can localize some <80Hz frequencies in the corners where my mains are. So I prefer that all bass comes from my centered sub

 

[abdo123]

You fellows are much smarter than I am, so perhaps you can comment on a question I have that is related to this. In small rooms, don't room reflections fill in the dips caused by cancellations?
I would think this happens more in higher frequencies, but could it contribute audibly in the bass?
I apologize if the question is stupid. Jim

‘Room reflections’ is usually an umbrella term used for late reflections. Here we’re talking about the very first reflections, usually 1-10 ms after the direct sound. These are very destructive and the brain cannot differentiate them from the direct sound.

 

[thorvat]

Possibly. You can observe the phase and magnitude interaction of the sub and front mains set in the REW’s alignment tool. All you need is the IR of the sub only and the left+right only. Move the delay sliders around if there is no time reference and/or align the phases somewhere between the xo overlap.

Exactly. What you see as predicted summed response in REW can be verified with MMM measurement with sub+left+right playing together.

 

[ernestcarl]

you managed to overlap group delay?

It effectively "overlaps" when summed or vector averaged...

Achieving absolutely identical group delay between channels is unnecessary. I'd rather work to have a good enough phase match/coherency and time alignment between all channels. Haven't bothered trying to achieve a perfect-looking step response like what you see in Acourate and Audiolense corrected IRs...



Currently I use 80Hz 12dB/oct BW HPF for the fronts (ports closed), 60Hz 48dB/oct BW HPF for the surrounds (ports closed), and the sub's own internally fixed 120Hz 12dB/oct LPF (sealed sub). The sub is midpoint the LR channels so no worry of any localization issue.

Not using any FIR for the sub (other than 256 taps each for the most basic linear phase EQ & phase linearization done for the surrounds and mains) which also helps me achieve minimal overall system delay in my MCH setups.

 

[thorvat]

Achieving absolutely identical group delay between channels is unnecessary. I'd rather work to have a good enough phase match/coherency and time alignment between all channels. Haven't bothered trying to achieve a perfect-looking step response like what you see in Acourate and Audiolense corrected IRs...

Correct. Getting the amplitude levels aligned together with phase/time alignement between sub(s) and mains is crucial for good sub integration, so the same thing as when designing any other XO between various drivers.

Btw, once you get that to work the only thing you need to do in order to have "perfect-looking" step response is to implement phase correction of the XO between sub(s) and mains. And that is the only "magic" what Acourate and Audiolense do when integrating subs..