Suggestion for subwoofer orientation (not placement)

[gorman]

I am adding a third subwoofer to my setup.

The floorplan is more or less this. Only thing changed is the center speaker is now on top of the furniture piece with the AVR and sources.

Following pictures will show current subs placement and the location I have available for the new one. I don't think I have any other choice for placement. Below the desk (which would have space enough... the space is occupied by a drawer cabinet).
The sofa has been moved a bit back, seating now begins at 120cm from the furniture rack, not 95. But more or less this is it.

This is from the sofa. The front speakers and center speaker will be substituted with KEF R7 Meta and KEF R6 Meta in a matter of days.
The two subwoofers are SVS SB1000. There's acoustic absorbing panels on the walls and ceiling, three bass traps (only one shows here).
I use Dirac DLBC to equalize and I have satisfying results. This after investing in acoustic treatment, before it was ok but after acoustic treatment and after adding a second subwoofer... sound became best I've heard in my life (at home). Which is what prompted me to change the front and center speakers and add a third sub, hoping to be able to further improve the experience (current speakers are 20+ years old Dynaudio Audience).

Now, on to the subject of my question. Third sub placement. I can place it where I point with red arrow in the drawing above, underneath a small table I have (an Ikea piece that's probably older than me...).

Here I show it without the table:

Radiators are not in use (I use air inverters nowadays), so don't worry about temperature.

My question is: what would be the best new orientation for the subwoofer (a SVS SB2000)? Does it matter?
I know, in the end, one could say, try and measure. But I'm interested in avoiding common knowledge mistakes.
Maybe it's well known not to orient toward an absorbing mass (the sofa), maybe it's known that it would be better to have it face the other subs in the front, maybe have it face opposite, towards the back wall. Maybe having it underneath the small table worsens considerably its performance...
Any suggestion or info is welcome. As stated, unfortunately, I don't have much freedom to move it elsewhere. What it could bring it has to be brought from there...

Thanks!

 

[DVDdoug]

I makes very little difference because of the long wavelengths at bass frequencies 100Hz has a wavelength of about 10 feet, 50Hz about 20 feet, etc.

 

[gorman]

I makes very little difference because of the long wavelengths at bass frequencies 100Hz has a wavelength of about 10 feet, 50Hz about 20 feet, etc.

So neither the sofa nor the table underneath which I'd place it would influence its performance? Not in a significant way, at least?

 

[No_hair_left]

So neither the sofa nor the table underneath which I'd place it would influence its performance? Not in a significant way, at least?

Mine did when I measured pointing straight forward and then facing right in the exact same position, both variations had different peaks and nulls.

 

[DWPress]

I have a SB1000 and had it under a small table like you propose. No issues, I had a towel on the surface too for dampening but be aware whatever you have on the table can potentially cause resonances. Of course the most convenient place for a sub may not be the best place for it, measurements will tell all....

 

[gorman]

both variations had different peaks and nulls

Ok, this is interesting. Meaning that, if at my first "version" of the placement I don't get good results I could try alternative facings. Thank you!

I have a SB1000 and had it under a small table like you propose. No issues, I had a towel on the surface too for dampening but be aware whatever you have on the table can potentially cause resonances. Of course the most convenient place for a sub may not be the best place for it, measurements will tell all....

Ok, good to know too. Yeah, I'm aware that the most convenient place isn't necessarily the best (Murphy's law would suggest the opposite ;-) ).
Generally speaking I don't intend to keep a whole lot of stuff on that table, more often than not I'll keep it empty.

 

[NTK]

Mine did when I measured pointing straight forward and then facing right in the exact same position, both variations had different peaks and nulls.

The acoustic center of a subwoofer can be many inches to more than 1 ft in front of the cone front surface, and is dependent on the driver diameter (and frequency too). Therefore, when the subwoofer is reoriented, even if the assumption that it is an omnidirectional point source is completely valid, it can still have the net effect of moving the subwoofer a few feet (e.g. when it is flipped around 180°).
From:

Desensitize Yourself

'When a loudspeaker is reproducing sound with a wavelength considerably larger than the size of its enclosure, we all know that it should be essentially omnidirectional and send out sound waves spherically.' John Vanderkooy (BEng, PhD – Distinguished Professor Emeritus) And yet, when we walk...

www.merlijnvanveen.nl

 

[gorman]

The acoustic center of a subwoofer can be many inches to more than 1 ft in front of the cone front surface

In my case, if you refer to the picture I've posted, this means the acoustic centre of the sub could be somewhere *inside* the sofa?

 

[NTK]

In my case, if you refer to the picture I've posted, this means the acoustic centre of the sub could be somewhere *inside* the sofa?

Let's run a couple of simplified (2-d cylindrical model, FEM) simulations to find out.

Acoustic center is clearly in front of the front surface of the speaker.

Place a barrier in front, and it moves the acoustic center back (actually slightly behind the speaker front surface now).

 

[gorman]

Place a barrier in front, and it moves the acoustic center back.

So the sofa could contribute to put the center back "inside" the sub. Interesting. Thanks a lot for working on this and for the graphs. Much of this flies over my head but, so far at least, I'm glad that nobody has screamed at me "FOOL!! NEVER DO THAT!".

 

[Keith_W]

Let's run a couple of simplified (2-d cylindrical model, FEM) simulations to find out.

Acoustic center is clearly in front of the front surface of the speaker.

Is the acoustic center the same as the zero delay plane? How can it be in front of the speaker? That would imply negative group delay.

 

[NTK]

Is the acoustic center the same as the zero delay plane? How can it be in front of the speaker? That would imply negative group delay.

I think the acoustic center is more for the steady state when we need to figure out the phase/distance relationship between different sources (as we need the phase to compute the vector sum).

 

[Keith_W]

Hmm, I am sure you are right (you do after all know vastly more than me) but I can't get my small brain to compute. I'll have to go read up some more.

 

[NTK]

Hmm, I am sure you are right (you do after all know vastly more than me) but I can't get my small brain to compute. I'll have to go read up some more.

May be I have made it sound more complicated than it is. Say, when we fire up REW to run a room sim. It asks you for the location of each loudspeaker, which is a point in space. The acoustic center would be, I believe, the best location for that point.

 

[mj30250]

The acoustic center of a subwoofer can be many inches to more than 1 ft in front of the cone front surface, and is dependent on the driver diameter (and frequency too). Therefore, when the subwoofer is reoriented, even if the assumption that it is an omnidirectional point source is completely valid, it can still have the net effect of moving the subwoofer a few feet (e.g. when it is flipped around 180°).
From:

Desensitize Yourself

'When a loudspeaker is reproducing sound with a wavelength considerably larger than the size of its enclosure, we all know that it should be essentially omnidirectional and send out sound waves spherically.' John Vanderkooy (BEng, PhD – Distinguished Professor Emeritus) And yet, when we walk...

www.merlijnvanveen.nl

Yup. I was surprised when the difference between a ~60Hz-90Hz null and a reasonably flat response in my room was to leave the sub in the same spot and simply flip it to face the opposite direction.

As for the op, this is the correct answer regarding the ideal orientation:

I know, in the end, one could say, try and measure.

At minimum, try it and listen.

 

[gorman]

Yup. I was surprised when the difference between a ~60Hz-90Hz null and a reasonably flat response in my room was to leave the sub in the same spot and simply flip it to face the opposite direction.

Quite the gold nugget you left here. Very interesting to know. I'll keep it in mind (it's similar to what No_Hair_Left said).

At minimum, try it and listen.

I'll listen but I'll measure too. Dirac's measurements are usually quite honest (it's the post corrections simulation that needs to be double checked the most, IME).
I posted just to avoid glaring errors, since I know I have so much to learn (which is oddly comforting, to be honest).