• 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!

Help fixing upper bass nulls, rear test panel didn't change much

hutt132

Member
Joined
Jul 25, 2019
Messages
58
Likes
88
The sound system is 7.2.4 surround. All the speakers are from Ascend Acoustics and the subwoofers are Monoprice Monolith 15in 1000w.
  • Front LR - SIERRA-2EX V2
  • Center - ELX HORIZON
  • Surrounds/Heights - HTM-200SE
The room is 232 in. (19.33 ft.) wide x 203 in. (16.92 ft.) deep. Measured to the back wall—I can measure the entryway and hallway if needed.
Listening position is 60 in. from rear wall, so nearly at the 2/3 point in the room.

Canopy panels - 4" rockwool, which I mainly put up to absorb light bounce from the ceiling above the screen. Sadly I didn't take before and after freq measurements.
Blackout curtains - (heavy, double layered blackout) on the two right wall windows and the same curtains on the left wall to cover the entrance/wall to the kitchen.
Black rug - 3/8" thick felt underlayment under it. The red rug is thick.
I was planning to build a large rear wall panel with 4" Rockboard 60.

The green line REW measurement is of the front LR speakers (no subwoofers) from the center couch position at ear level.
The red line is with a 3" 80x60in gel foam mattress topper on the rear wall behind the shelf to test what I could expect from rear sound panel I've been planning.
(The topper only went up 60" from the floor, whereas the rear panel I'm planning goes to the top of the projector 108" up.)
The white line is the estimated in-room response of the speakers as published by Ascend Acoustics.

There's some gnarly nulls and peaks below 300 Hz, but above that it stays mostly compliant with the estimated speaker response.
I can't figure why the mattress topper made the 330 Hz and 530 Hz dips worse, while only slightly improving the nulls in the lower frequencies and not lowering the peaks at all.
I also added 3.5 in. x 15 in. x 47 in. rockwool on the wall directly behind the speakers between the screen and the shelves as a test for comb filtering and it didn't change the frequency response at all.

Any insight would be much appreciated to fix the nulls around 155-150Hz, 120Hz, 89Hz, and 72Hz.

L+R REW Measurements (no subwoofers)
room freq response2.png


IMG_6507.JPG

IMG_6508.jpg


Rendering of planned rear panel
back panel  semi wide black 0.5.png
 
Last edited:

Ron Texas

Master Contributor
Forum Donor
Joined
Jun 10, 2018
Messages
6,470
Likes
9,697
Try raising the crossover point for the subs to 100 Hz and see if that helps any. Also, experiment with moving the mains closer to the front wall with or without the ports plugged. The 155 Hz dropout looks like a quarter wave cancellation. Genelec recommends placement at either less than .5 M or more than 2 M from the front wall. You can also try side wall placement for the subs. Floor reflections are another possibility. You could temporarily reorient the carpet to cover more of the floor near the mains and see if that helps. I'm not super technical, but some of these things have worked for me. I have original LS50's and a pair of Rhythmic L12 subs.

The room has a nice vibe. Those subs look like they could go down to DC.
 
OP
hutt132

hutt132

Member
Joined
Jul 25, 2019
Messages
58
Likes
88
Try raising the crossover point for the subs to 100 Hz and see if that helps any. Also, experiment with moving the mains closer to the front wall with or without the ports plugged. The 155 Hz dropout looks like a quarter wave cancellation. Genelec recommends placement at either less than .5 M or more than 2 M from the front wall. You can also try side wall placement for the subs. Floor reflections are another possibility. You could temporarily reorient the carpet to cover more of the floor near the mains and see if that helps. I'm not super technical, but some of these things have worked for me. I have original LS50's and a pair of Rhythmic L12 subs.

The room has a nice vibe. Those subs look like they could go down to DC.
The measurements are with the subs off and speakers ran at full range. Figured I would try to fix the L+R before jumping down the subwoofer rabbit hole.
The front of the speakers are 31" (0.79m) from the wall.
I had some extra underlayment that I put on more of the floor next to the mains and it didn't change anything.

I could try moving the speakers back to 0.5m like Genelec suggests, but I wouldn't be able to keep the listening position in the equilateral triangle.
I don't know if the triangle is a real thing, but it seemed to help imaging when I put my speakers there (shrug).


1714793225577.png
 

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
2,877
Likes
6,627
Location
Melbourne, Australia
Gorgeous room you have there :)

Gnarly peaks and nulls is normal in any listening room. They may be caused by one of these things:

1. Room modes,
2. SBIR,
3. Phase issues causing improper summation at the XO point,
4. Inherent issues with the subwoofer or speaker, e.g. it might have a resonance, cancellation, etc.

All of them exist together and need different solutions.

Room modes change with microphone position. So your first step to confirm if it is a room mode or not is to repeat the measurement with the mic 1m forward, or use an RTA and observe the pattern of peaks and nulls. You should see the peaks and nulls move up and down in amplitude as the mic is moved. If you see any peak/null that is not moving, that is going to be either SBIR or a speaker/sub issue.

SBIR produces a constant peak or null which does not change with microphone position. It does change with speaker/sub position, determined by its distance to the closest reflecting surface. To diagnose SBIR, fix your mic at the MLP and move the speaker/sub back and forth, closer/further away from the wall. What you will see is the peak/null shifting up/down in frequency but not in amplitude. If there are phase issues causing improper summation at the XO point, moving the speakers/sub will cause the peak/null to shift in amplitude but not frequency.

By the end of this, if you still see peaks/nulls that do not move with mic position, and do not move with changing sub/speaker position, then by process of exclusion you have a speaker/sub which is not producing a flat frequency response. There are further steps you can take to confirm this (e.g. ground plane measurement) but it is not worth the trouble since it won't change the decision you have to make - either deal with it, or replace it.

Now, we come to what you need to do.

Room modes can be treated with acoustic treatments like you are proposing. Floyd Toole said that the only thing bass traps do is trap your money. I wouldn't go so far as to say they are completely useless, but they either excessively intrusive (velocity absorbers) or narrowband and need to be tuned (membranes). You could also use DSP.

SBIR can be improved by shifting your speakers/subs closer to the front wall. The closer distance means the reflection pathway is shorter, which means the wavelength is shorter, which shifts the problem frequency higher. Higher freq peaks/nulls are easier to deal with than lower freqs - for one, the thickness of the absorber required is much less. You could also consider rotating your subs 90 degrees so they fire sideways. It sometimes works. Or, if it is a peak you are suffering from ... you could use DSP.

Improper phase summation between the speaker/sub has only one solution. DSP.

I think you might see that there is a common solution no matter what your issue is, and that is DSP :)
 

sam_adams

Major Contributor
Joined
Dec 24, 2019
Messages
1,060
Likes
2,658
gel foam mattress topper on the rear wall

Not the best type of velocity-based absorber. Velocity-based absorbers—rockwool, fiberglass, cotton denim, melamine foam—rely on tortuosity and gas flow resistance to reduce reflections and standing waves by reducing the velocity of the particles of air in the pressure wave at the room boundaries. Gel foam does not have the necessary properties or thickness to accomplish this.
 

MAB

Major Contributor
Joined
Nov 15, 2021
Messages
2,234
Likes
5,077
Location
Portland, OR, USA
Nice room and setup.

I assume the speakers and the mic didn't change position one bit between the measurements.
The lower bass nulls are likely room modes.
The nulls and peaks above 200Hz are likely comb filtering, like illustrated below:
index.php


If so, the measurement with the mattress pad is different because for the changed effective speaker to wall distance. To test this, remeasure with the mic in the same position and move the speaker closer and further from the wall (I would do one speaker at a time to make things simpler). The room modes should stay roughly the same, and only these slightly higher frequency comb effects should vary. If so, the mattress pad is really doing nothing more at these frequencies than adjusting speaker to wall placement. The mattress pad is likely having effect at higher frequencies, and likely visible in early reflection and reverberation, but less change in the bass.
 
OP
hutt132

hutt132

Member
Joined
Jul 25, 2019
Messages
58
Likes
88
Nice room and setup.

I assume the speakers and the mic didn't change position one bit between the measurements.
The lower bass nulls are likely room modes.
The nulls and peaks above 200Hz are likely comb filtering, like illustrated below:
index.php


If so, the measurement with the mattress pad is different because for the changed effective speaker to wall distance. To test this, remeasure with the mic in the same position and move the speaker closer and further from the wall (I would do one speaker at a time to make things simpler). The room modes should stay roughly the same, and only these slightly higher frequency comb effects should vary. If so, the mattress pad is really doing nothing more at these frequencies than adjusting speaker to wall placement. The mattress pad is likely having effect at higher frequencies, and likely visible in early reflection and reverberation, but less change in the bass.
The front of the speakers are 31 in from the wall.
31 x 4 = 124 in wavelength = 109Hz
Which about lines up with the one null in my graph.

I've marked the f1 f2 f3 dips and their respective peaks on the chart to indicate that they're probably due to SBIR.
And the others are probably room modes. I'll have to move stuff around and do more testing.

1714800304689.png
 
  • Like
Reactions: MAB
OP
hutt132

hutt132

Member
Joined
Jul 25, 2019
Messages
58
Likes
88
Room modes change with microphone position. So your first step to confirm if it is a room mode or not is to repeat the measurement with the mic 1m forward, or use an RTA and observe the pattern of peaks and nulls. You should see the peaks and nulls move up and down in amplitude as the mic is moved. If you see any peak/null that is not moving, that is going to be either SBIR or a speaker/sub issue.

SBIR produces a constant peak or null which does not change with microphone position. It does change with speaker/sub position, determined by its distance to the closest reflecting surface. To diagnose SBIR, fix your mic at the MLP and move the speaker/sub back and forth, closer/further away from the wall. What you will see is the peak/null shifting up/down in frequency but not in amplitude. If there are phase issues causing improper summation at the XO point, moving the speakers/sub will cause the peak/null to shift in amplitude but not frequency.
Would moving the speakers but leaving the mic position fixed cause the room modes to shift frequencies as well?
 

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
2,877
Likes
6,627
Location
Melbourne, Australia
Would moving the speakers but leaving the mic position fixed cause the room modes to shift frequencies as well?

Yes, but the pattern is different. If you move your speakers by 1m, the room modal pattern will not change at all for low frequencies. As wavelengths get shorter, the amplitude starts changing slowly. Above the Schroder frequency, they change quickly.

Compare this to moving the microphone 1m. Then you will see the modal pattern changing very quickly in amplitude across low frequencies.

Have a play with REW's room simulator to see for yourself what happens.
 

DVDdoug

Major Contributor
Joined
May 27, 2021
Messages
3,189
Likes
4,205
I was planning to build a large rear wall panel with 4" Rockboard 60.
I'm NOT an acoustics expert but a wavelength of 4" is 3.3kHz (or about 1.5kHz "round trip") so it's not really working in the bass range.

(There are thin "tuned" or "resonant" bass traps.)
 

sam_adams

Major Contributor
Joined
Dec 24, 2019
Messages
1,060
Likes
2,658
Would moving the speakers but leaving the mic position fixed cause the room modes to shift frequencies as well?

The room modes are determined by the dimensions of the room itself. The placement of the speakers in the room will determine how the room's modes are excited.

The link below has an nice animated gif that demonstrates the effect of Speaker Boundary Interference Response (SBIR):

 
OP
hutt132

hutt132

Member
Joined
Jul 25, 2019
Messages
58
Likes
88
Thank you everyone for the feedback and help! I pushed the speakers back as close to the wall as possible, which is 13.25" from the face of the speakers now, and they have way less comb filtering going on.

The red line is before and the green line is with them pushed back. (it doesn't match up to my previous picture since I moved the mic slightly)
Now I need to measure the subwoofers and start pushing them around to see what spots have the least amount of dips. Though I'm very limited on where I can put them up front.
It's also a shame that Denon's Audyssey MultEq XT32 does NOT equalize two subwoofers separately, but it does control individual gain and distance.
"Audyssey Sub EQ HT™ makes the integration of dual subwoofers seamless by first compensating for any level and delay differences between the two subwoofers and then applying Audyssey MultEQ® XT32 to both subwoofers together."
freq response 2.png
 
OP
hutt132

hutt132

Member
Joined
Jul 25, 2019
Messages
58
Likes
88
These are the best results I could get for the subwoofers with moving them around up front. I tried turning them 90deg as well but that only improved the 70-100Hz range. I crossover at 80Hz anyway.

So looks like I have a huge null spot that's making a huge dip between 38-51Hz.

subs.png
 

middlemarch

Active Member
Forum Donor
Joined
Jan 4, 2023
Messages
149
Likes
172
Location
Seattle Area
Have you time aligned them using the time alignment tool in REW? It’s important to redo the time alignment for each combination of positions of the two subs to get an accurate picture of the situation at the mlp. Tedious, it is, but gets quicker pretty quickly. You’ll get there
i’m assuming you have some ability to affect the alignment, such as with a MiniDSP, etc. you don’t mention what your setup is other than the speakers, so my assumption may be incorrect.
 
Last edited:

kemmler3D

Major Contributor
Forum Donor
Joined
Aug 25, 2022
Messages
3,995
Likes
8,075
Location
San Francisco
I would try moving one of the subs to the back of the room and playing with phase... with some luck that will fill in some nulls and then you can shave the peaks with EQ.

Long story short nulls are mostly caused by the room itself and where the speakers are in the room. The only truly effective way to get rid of nulls is to move things around significantly, since the wavelengths we're talking about are 10-50 feet. Looks like you've made some progress already. You've got some good gear in there so I think your odds of taming most of them are good, though.
 
OP
hutt132

hutt132

Member
Joined
Jul 25, 2019
Messages
58
Likes
88
Have you time aligned them using the time alignment tool in REW? It’s important to redo the time alignment for each combination of positions of the two subs to get an accurate picture of the situation at the mlp. Tedious, it is, but gets quicker pretty quickly. You’ll get there
i’m assuming you have some ability to affect the alignment, such as with a MiniDSP, etc. you don’t mention what your setup is other than the speakers, so my assumption may be incorrect.
The red is sub 1 and orange is sub 2. Green is them both together. I time aligned them in REW but it was only 1.6ms delay which improved the null by about 3db.

I'm using a Denon X4700H receiver. "Audyssey Sub EQ HT™ makes the integration of dual subwoofers seamless by first compensating for any level and delay differences between the two subwoofers and then applying Audyssey MultEQ® XT32 to both subwoofers together."
So Audyssey will set the level and delay, but will EQ them as one unit afterwards.

I would try moving one of the subs to the back of the room and playing with phase... with some luck that will fill in some nulls and then you can shave the peaks with EQ.

Long story short nulls are mostly caused by the room itself and where the speakers are in the room. The only truly effective way to get rid of nulls is to move things around significantly, since the wavelengths we're talking about are 10-50 feet. Looks like you've made some progress already. You've got some good gear in there so I think your odds of taming most of them are good, though.
Sadly the only spot I could place it in the back is in the poster corner. There's no room behind the couch since that is a walkway, and it would block the front door if I placed it near there.
 

ozzy9832001

Senior Member
Joined
Feb 19, 2023
Messages
421
Likes
272
I'd have a sub between the 2 mains and then try another position for the 2nd, like in a corner -- front or back you'll get a different response. Next try playing with the crossover since they maybe different for each sub.
 

a2Thompson

Member
Joined
Apr 1, 2022
Messages
24
Likes
14
If you’re still going at it and have some bandwidth you can probably tame everything under that 60 hz to near flat using the multi-sub optimizer (MSO) and a minidsp using all phase filters since you have two subs…though you might need to move one to a different location. Or perhaps some of the advanced Dirac products which also balance two subs using frequency-dependent phase filters. These work well down low.
 
Top Bottom