• 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!

Adding a third sub to cancel "Deep Mid-Bass Null"

1725908610945.png


Here you go. Pay attention to a few little changes I made:

- No change at all to your current speaker/sub layout. I am assuming it is that way for a reason.
- Adjusted delay and gain for sub 2.

I think this proves that with your subs in the existing position with a bit of DSP, you can get it to work. What you need to do is do proper time alignment for both the subs and cut that huge peak at about 28Hz.
 
View attachment 391197

Here you go. Pay attention to a few little changes I made:

- No change at all to your current speaker/sub layout. I am assuming it is that way for a reason.
- Adjusted delay and gain for sub 2.

I think this proves that with your subs in the existing position with a bit of DSP, you can get it to work. What you need to do is do proper time alignment for both the subs and cut that huge peak at about 28Hz.

You are right @Keith_W
I was also playing around with it and came to this conclusion:

1725909975190.png


I don't think there is a need for a 3rd sub here
 
35hz is sub bass not midbass. :)

It's also worth noting that if you've already tried all possible locations and they all excite the null, then adding a third sub to one of those locations will still excite the null, and not fix it.
 
View attachment 391197

Here you go. Pay attention to a few little changes I made:

- No change at all to your current speaker/sub layout. I am assuming it is that way for a reason.
- Adjusted delay and gain for sub 2.

I think this proves that with your subs in the existing position with a bit of DSP, you can get it to work. What you need to do is do proper time alignment for both the subs and cut that huge peak at about 28Hz.
I wish I could move the main speakers. The left side has pocket doors, which puts that speaker in the center of the right door. The right side is the bathroom door, right behind me is a window and in the middle of the right wall are the patio doors.

This screenshot is after Arc Genesis correction. I would have to buy a DSP for the 28hz.

Thank you and ppataki very much for taking the time to look at the room sim and responding :)
 
35hz is sub bass not midbass. :)

It's also worth noting that if you've already tried all possible locations and they all excite the null, then adding a third sub to one of those locations will still excite the null, and not fix it.
That isn't good news. This is the results I got when I move the sub from the front right to the right back. When both were placed in the front, I got the 70hz null. The only other place might be somewhere along the mid-wall, left. With a rectangle room 20' X 13' I don't understand why the "Nulls"

Thank you for the response.
 
That isn't good news. This is the results I got when I move the sub from the front right to the right back. When both were placed in the front, I got the 70hz null. The only other place might be somewhere along the mid-wall, left.

Thank you for the response.
That's why I included "appropriately placed" in my response, and as @sigbergaudio notes that is often the problem, no good place to physically and aesthetically place the sub to counter the null.

There are many online calculators; this one shows a 70 Hz mode in your room: http://www.mcsquared.com/modecalc.htm

1725918530642.png
 
I didn't look closely, just ran the calculator. Looking further, it's a height null if you've an 8' ceiling, but the real answer is the actual nulls depend upon what else is in the room, the actual shape and any openings, your listening position, and so forth. Measurements always trump calculations (assuming correct measurements and calculations). Since you know the null, I wouldn't worry too much about what calculators show at this point.

Here's the one I use most often (from the Harman site):
1725922548846.png
 
This screenshot is after Arc Genesis correction. I would have to buy a DSP for the 28hz.

Thank you and ppataki very much for taking the time to look at the room sim and responding :)

You're welcome. You don't need to buy a DSP unit. If you are using a PC as a source and using something like Roon, JRiver, or Foobar, all you need is to figure out how to get REW to make correction filters for your subs. You also need your subs to be on separate DAC channels, so you need 4 DAC channels in total. If you happen to have a spare 2ch DAC and the ability to take measurements, then this could be a zero cost upgrade for you.

Of course, a DSP unit is nice. You can decide if you want to spend money or not :)
 
That isn't good news. This is the results I got when I move the sub from the front right to the right back. When both were placed in the front, I got the 70hz null. The only other place might be somewhere along the mid-wall, left. With a rectangle room 20' X 13' I don't understand why the "Nulls"

Thank you for the response.

70hz = 4.9m, half wave will be 2.45m. Floor to ceiling?
 
A true room-mode null is due to signal cancellation from the sound interacting with room boundaries.

Sorry Don, but I don't think that's correct. Room modes are frequencies at which standing waves can occur in a room. They are solely dependent on the room geometry. Whether or not they become a problem also depends on speaker placement (speaker near a velocity anti-node tends to excite the room mode more) and where the listener is located (no sound pressure at a velocity anti-node, for example).

The sound bounces off a boundary and recombines with the direct sound but out of phase so the sound cancels.

Yes, but this destructive interference, known as SBIR (speaker-boundary interference response) is unrelated to room modes and affects a frequency at which the speaker is located at 1/4 wavelength distance from the boundary.

Assuming Rich measured the room correctly, REW shows that there are no room modes near the problem range. On the other hand, the room height (2.39m) suggests an SBIR frequency in the 35-40Hz range. Assuming the subs are near the floor, this would affect them all. I'd suggest placing one subwoofer on a chair or something for a test, and checking how that null changes.
 
Assuming the subs are near the floor, this would affect them all. I'd suggest placing one subwoofer on a chair or something for a test, and checking how that null changes.
8, 10 and 12+ lids all have common characteristics for their heights. Raising where a driver is located (in relation) to the lid has always mediated problems
I've had, with 50, 80 and 120hz and the peaks on either side of those frequencies. I used mechanical (Helmholtz) adjustable tune resonators for the peaks
at 40&60&90htz, treated the ceiling in front of and over the seated position, and EQed the rest after considerable sub-moving, in and out from the
mid-wall and staying away from the corners. I raised both subs on the long walls to 3.5 feet off the floor and used subs firing from two sides with passive
radiators and an active pointing in front of the seated position.

The type of sub you use has a lot to do with how it performs in a room. Columns changed EVERYTHING especially if they are a 2 or 3 sided design.
I ran 4 servo columns for 12 years. I tried every imaginable combination and then placed them mid-wall. It didn't look as good but it worked PERFECT
and we didn't have the software available then we do now to measure with.

I then went to 12" double GRs OB servo and only needed 3 for a 300sf room with 8ft lids. Again the type of cabinet design and height makes all the difference in the world. The top driver is almost 40" off the ground and there are 3 12" drivers firing in both directions mid-wall.

I'm a mechanic I think of the physical requirements before I EVER apply DSP. I'm having to set up a new room myself but I've learned through the years
"know your room" and I learned from my 45 years of trials in the same types of rooms. 3-400sf rectangles with 8, 10, 12+ft lids. Corners are for resonators
just like either side of subs. 3-4 feet. 300sf with 3 dual subs takes 6-8 resonators.

I didn't start with a problem, I didn't have an answer for, Mechanically first. I try to stay out of trouble, not get out of trouble.

You'll get it figured out but it will take physical location not DSP software and if you do use it it will be VERY minor peaks, unless you have some weird
mains like I use. LS hybrid using small Planars. They require a quick cone response in the sub/bass 250hz < to blend and sound right for my ears. Cone
speaker are pretty easy to blend in comparison.

Regards
 
I suppose most people with multiple subs have them sitting on the floor (I do). How much of an issue with regard to ceiling SBIR this becomes, depends on the ceiling. In my case, it's plasterboard, with a roof cavity above it. I'm lucky, because this sort of ceiling is fairly transparent to deep bass. My 40Hz ceiling SBIR suck-out is mild and easily dealt with by bringing down the subwoofer output above and below this region a little. For someone with a concrete ceiling this would be a harder problem so solve. In this case, installing subs at different heights may be good strategy.
 
Sorry Don, but I don't think that's correct. Room modes are frequencies at which standing waves can occur in a room. They are solely dependent on the room geometry. Whether or not they become a problem also depends on speaker placement (speaker near a velocity anti-node tends to excite the room mode more) and where the listener is located (no sound pressure at a velocity anti-node, for example).
Room geometry = boundaries of the room. Modes occur because the sound wave hits a wall (floor, ceiling, boundary) and transitions (reflects) back leading to constructive and destructive interaction with the direct wave. Yes, the dimensions determine the frequencies, but I am curious what you think causes the peaks and valleys if not sound waves reflecting from room boundaries? I suspect there's a fundamental difference in how we define standing waves, room modes, and reflections/resonances in a room (or any system). It is also possible the definitions have changed since my ancient courses in acoustics.

Note that I distinguish room modes, due to sound interacting with room boundaries no matter the speaker or listener position, from SBIR (speaker boundary interference response) which is caused by speaker placement relative to a boundary. A source anywhere in the room should excite standing waves to at least some degree; SBIR is generally treated separately (by me, anyway) than room modes. The listener has no impact on room modes; where the listener sits relative to the modal peak or valley is important, of course, but generally in the bass a listener is too small a volume to significantly affect the modal response. Moving the listener away from the modal null is one solution. Another solution is as you said, that I and many others do, add a subwoofer at the null position to "override" the null, or phased to do the same.

Unfortunately, right or wrong, this is the way I've been defining them for 40+ years, so it may take me a little time to correct.

Yes, but this destructive interference, known as SBIR (speaker-boundary interference response) is unrelated to room modes and affects a frequency at which the speaker is located at 1/4 wavelength distance from the boundary.
See above. My definition implies interference is responsible for both room modes and SBIR.

Assuming Rich measured the room correctly, REW shows that there are no room modes near the problem range. On the other hand, the room height (2.39m) suggests an SBIR frequency in the 35-40Hz range. Assuming the subs are near the floor, this would affect them all. I'd suggest placing one subwoofer on a chair or something for a test, and checking how that null changes.
I have found over the years that room mode calculators tend to be a little off due to inexact room measurements and other things in the room changing the effective dimensions and such. I mentioned the height but did not know his room's height; I just used 8 feet as that is a common ceiling height in the U.S.A. Placing a sub on a chair is a good idea.

My room is asymmetric so normal calculators for rectangular rooms are a little off the mark in determining modal frequencies. I have used amcoustics' amroc calculator to good effect: https://amcoustics.com/tools/amroc
 
Sorry Don, but I don't think that's correct. Room modes are frequencies at which standing waves can occur in a room. They are solely dependent on the room geometry. Whether or not they become a problem also depends on speaker placement (speaker near a velocity anti-node tends to excite the room mode more) and where the listener is located (no sound pressure at a velocity anti-node, for example).



Yes, but this destructive interference, known as SBIR (speaker-boundary interference response) is unrelated to room modes and affects a frequency at which the speaker is located at 1/4 wavelength distance from the boundary.

Assuming Rich measured the room correctly, REW shows that there are no room modes near the problem range. On the other hand, the room height (2.39m) suggests an SBIR frequency in the 35-40Hz range. Assuming the subs are near the floor, this would affect them all. I'd suggest placing one subwoofer on a chair or something for a test, and checking how that null changes.
The subs are on the floor and bringing up the ceiling, I have to say that the ceiling isn't flat all the way across; there is a duct chase that runs the length of the ceiling on the left side. The chase is 12" high X 24" deep, see attached. Could this be my problem, since Room Mode calculators cannot account for the chase?
 

Attachments

  • 20240910_125459.jpg
    20240910_125459.jpg
    85.7 KB · Views: 30
Room geometry = boundaries of the room. Modes occur because the sound wave hits a wall (floor, ceiling, boundary) and transitions (reflects) back leading to constructive and destructive interaction with the direct wave. Yes, the dimensions determine the frequencies, but I am curious what you think causes the peaks and valleys if not sound waves reflecting from room boundaries? I suspect there's a fundamental difference in how we define standing waves, room modes, and reflections/resonances in a room (or any system). It is also possible the definitions have changed since my ancient courses in acoustics.

Note that I distinguish room modes, due to sound interacting with room boundaries no matter the speaker or listener position, from SBIR (speaker boundary interference response) which is caused by speaker placement relative to a boundary. A source anywhere in the room should excite standing waves to at least some degree; SBIR is generally treated separately (by me, anyway) than room modes. The listener has no impact on room modes; where the listener sits relative to the modal peak or valley is important, of course, but generally in the bass a listener is too small a volume to significantly affect the modal response. Moving the listener away from the modal null is one solution. Another solution is as you said, that I and many others do, add a subwoofer at the null position to "override" the null, or phased to do the same.

Unfortunately, right or wrong, this is the way I've been defining them for 40+ years, so it may take me a little time to correct.


See above. My definition implies interference is responsible for both room modes and SBIR.


I have found over the years that room mode calculators tend to be a little off due to inexact room measurements and other things in the room changing the effective dimensions and such. I mentioned the height but did not know his room's height; I just used 8 feet as that is a common ceiling height in the U.S.A. Placing a sub on a chair is a good idea.

My room is asymmetric so normal calculators for rectangular rooms are a little off the mark in determining modal frequencies. I have used amcoustics' amroc calculator to good effect: https://amcoustics.com/tools/amroc
The ceiling isn't flat all the way across; there is a duct chase that runs the length of the ceiling on the left side. The chase is 12" high X 24" deep, see attached. Could this be my problem, since Room Mode calculators cannot account for the chase?
I suppose most people with multiple subs have them sitting on the floor (I do). How much of an issue with regard to ceiling SBIR this becomes, depends on the ceiling. In my case, it's plasterboard, with a roof cavity above it. I'm lucky, because this sort of ceiling is fairly transparent to deep bass. My 40Hz ceiling SBIR suck-out is mild and easily dealt with by bringing down the subwoofer output above and below this region a little. For someone with a concrete ceiling this would be a harder problem so solve. In this case, installing subs at different heights may be good strategy.
 

Attachments

  • 20240910_125459.jpg
    20240910_125459.jpg
    85.7 KB · Views: 18
I really appreciate all the help, everyone has offered to correct the problem and taking the time to respond to my post.... Thank You :)
 
You're welcome. You don't need to buy a DSP unit. If you are using a PC as a source and using something like Roon, JRiver, or Foobar, all you need is to figure out how to get REW to make correction filters for your subs. You also need your subs to be on separate DAC channels, so you need 4 DAC channels in total. If you happen to have a spare 2ch DAC and the ability to take measurements, then this could be a zero cost upgrade for you.

Of course, a DSP unit is nice. You can decide if you want to spend money or not :)
I do have an extra DAC, but I don't understand your configuration. The subs are powered by the Anthem MRX 1140 it was two outputs, but they are not stereo.
 
The configuration would go like this:

PC --> DAC 1 --> L, R speakers
--> DAC 2 --> Subwoofers 1 and 2

On the PC, download VB Audio Matrix. Enable the virtual ASIO channel. Set it up to output channels 1, 2 to DAC 2 (i.e. to the subwoofers 1 and 2), and channels 3, 4 to DAC 1 (i.e. to L, R speakers). Make sure the microphone is recognized by VB Matrix.

Open REW and set it up in ASIO mode to output to VB Matrix. Select the appropriate microphone channel.

Then you need to take some timing measurements. Set REW to use a timing chirp and set this as your left speaker (i.e. channel 3). Now do sweeps of the subwoofers at the listening position. REW will automatically tell you what the delay is with respect to the timing chirp, but you can check by overlaying the impulse responses (aligning all the impulses with the tweeter chirp) and measuring the delay manually that way.

Then you need to figure out how to make a filter that will delay the subwoofer appropriately. I use a different tool for this, so I can't talk you through the REW procedure.

Load the filters into the playback software of your choice. It will need a built-in convolver.
 
Back
Top Bottom