Square room horror

[Deleted member 48726]

It doesn't do much as the peak is there even when you don't play anything (in environmental noise) and neither will acoustic absorbers when it's that low. I have one at 43~46 Hz. Experimented with absorbers, closed enclosure speakers and subs which did help make it less prominent but it didn't disappear nor it ever will. When you combine all bit by bit you menage to get it better and ideal doesn't exist. At least that's my experience.

Well I'm sorry you haven't had luck getting rid of yours. I'm well aware of my room modes and I have integrated and tamed my subwoofer just fine using Dirac and before that REW constructed EQ.

And if I don't activate the mode with any sound it's not there (It's not a problem). When EQ'ing the 31 Hz peak down some 15 dB or so (can't remember) then it cuts -15 dB at any given volume to match the rest of the response, hence flattening it, effectively eliminating the room mode. Some nulls are different beasts though. I don't boost very much in my EQ.

 

[audiofooled]

Hi guys

Thanks for all the input - it's very much appreciated.

1. Sorry, will post pictures of the room when I can. Don't have any to hand. There's a stone fireplace to the left of the room and a sort of stone alcove/shelf extending into the room near to the left speaker in the picture.

2. A second sub is a possibility once my new AVR arrives, as it supports dual subs. AVR is a Yamaha RX-V6A. I know YPAO is not always very highly thought of, but I also have a UMIK-1 and I believe the AVR allows for tinkering with the PEQ manually to some extent.

3. The speakers are Focal Aria 906, front-ported. I ended up with the placement so close to the wall based partially on the results of REW sim, but also the suggestions (link below) which you may have already seen around the web. I'm not averse to trying them in a different position, though! The sub is a wallet-friendly Polk PSW10.

The #1 Speaker Placement Tip That Speaker Manuals Get Completely Wrong — SonicScoop

Speaker placement tip #1: it’s not enough to simply set up your speakers in an equilateral triangle and be done with it.

sonicscoop.com

Here you can see an interactive view of your room behavior (not so different from your REW simulation) https://amcoustics.com/tools/amroc?l=400&w=400&h=210&r60=0.6

Further more, from what I can see in Amir's review of Focal 906 https://www.audiosciencereview.com/forum/index.php?threads/focal-aria-906-speaker-review.14085/

and from response of Polk PSW10 that Erin had measured in his 10 inch subwoofer showdown, I can see that both speakers and a sub are peaking around 60 Hz. And all of them are front ported. This too has to be taken into account for good integration and to avoid phase cancelation issues. You can try moving the speakers away from the wall a bit to avoid SBIR problems and to less excite first order mode of 43 Hz. Also, try placing the speakers equilaterally at 25% distance from the left and right wall to avoid further exciting the 60 Hz mode which is peaking in the room corners and could produce unwanted ringing and muddying of the bass. In your case, the vertical centers of each speaker should be about 1m away from the left/right wall, which sets them about 2m apart. Then play with toe in with respect to your listening position.

As for the sub, you can keep it in the middle between the speakers but fairly close in line with them because of the ports and modal issues. Then you can play around with plugging the ports of the speakers or preferably that of a sub and crossover/DSP to get a more even response (82 Hz peak is excited more due to the sub being close to the floor, so EQ may help bringing this down). The listening position should not be too close to the rear wall.

I also always recommend this thread https://www.audiosciencereview.com/forum/index.php?threads/analytical-analysis-room-gain.23211/ about differences between room interaction of sealed and ported subs.

And this example from Dr. Toole's book (option d for your speakers). I hope you find this helpful.

 

[krabapple]

It doesn't hurt being repeated

"Alternately, put two subs (fed the same content) on front wall at 1/4" width positions, and sit ~2/3-3/4 of the room length from the front wall, centered between the left and right walls....[as Floyd Toole noted]"

And who would have guessed that the dastardly corporate shill Floyd Toole would promote such useful advice?

 

[Descartes]

It doesn't hurt being repeated

"Alternately, put two subs (fed the same content) on front wall at 1/4" width positions, and sit ~2/3-3/4 of the room length from the front wall, centered between the left and right walls....[as Floyd Toole noted]"

And who would have guessed that the dastardly corporate shill Floyd Toole would promote such useful advice?

What is the page number!
Just bought the newest version of his book it is on sale at Amazon! Had the previous version on the bookshelves already.

 

[Descartes]

My second system is in a square room of barely larger dimensions than your room and I have enjoyed it for years. However, the room has a fair amount of acoustical treatment, the system always employs DSP-based roomEQ, there are multiple subs and it is a multichannel system (either 5.1 or 5.1.4). All of those contribute to the taming of the acoustics.

How many subs?

 

[Kal Rubinson]

How many subs?

3

 

[krabapple]

What is the page number!
Just bought the newest version of his book it is on sale at Amazon! Had the previous version on the bookshelves already.

The mode cancellation image I linked to is not from his book, but from a white paper. But audiofooled's post #43 above shows a clearer image of the same thing and it shows as being from Fig 13.11 , page 221 of the 1st or 2nd ed. of Sound Reproduction. In the 3d edition e-book , it is figure 8.13 (page 237-8 of 490). I don't have my 3rd ed. hard copy (yes I have both book and e-book!) handy so I can't provide a physical page reference right now, sorry.

 

[Descartes]

So I am confused the conclusion on the slide shows 2 subs front and back?

But you show two subs in the front so which one is better?

 

[krabapple]

So I am confused the conclusion on the slide shows 2 subs front and back?

But you show two subs in the front so which one is better?

View attachment 259201

The Welti work in your image has a different focus. Its goal is to minimized variation in perceived bass response over a range of seats.
The method I showed you is for cancelling the three primary width modes at one seat , located in the midline of a rectangular room.

You should dig into Toole's book where it discusses these subjects. Welti's paper is also freely available IIRC, or at least, i seem to recall someone posting a PDF link to it on this forum.

 

[Digby]

Or at least a combination. Treat as much as feasible given WAF etc.

If you treat the W, AF may increase (this is never guaranteed!), making it possible to add more treatment to the room too.

 

[richardm]

Hi, I'm Johnny Knoxville; and this is square room subwoofer alignment!

I'm going to either take up drinking or burn the house down and start over. I've been chipping at this for dozens of hours and I've learned heaps about delays and all-pass filters in EqAPO. This office room is 11'4" x 11'11" with a 9ft ceiling (363x345x274cm) and a tile floor.

The information in this thread has been helpful and it's influenced my thinking but still no solution yet.

In the room-sim screenshot attached I've circled the acoustic tomfoolery that's making me mad as a hatter. The left/right/front/behind positions are twelve lousy inches in any direction. It's a 40db variation just from leaning forward or back in my chair. Room-sim is moderately accurate for me -- the main discrepancy is the frequency where trouble occurs with any given layout (it tends to be 5-10Hz higher in real life)

I've got nulls swirling around my head and looking to the left or right grants one earful of bass. A tone sweep supplies bass to one ear or the other across multiple ranges of frequencies. I've tried:

• Combinations of one, two, and three subs
• I've reversed phase on one or two subs
• I've applied delays -- modestly at first -- then so heavily as to "wrap-around" a sub to the next cycle.
• I've gently cut-in low-Q AP (all-pass) filters centered far away from the target frequency and gradually brought them closer to target -- both from below and above target Fs.
• I've surgically hammered nulls and "phasing swirls" with obscenely high-Q APs
• I've tried dissimilar xover filters (eg. LR4 on the main sub and LR6 on the secondaries) including xovers that phase invert.
• I've put subs up on a shelf above the room centerline.
• I've stacked multiple AP filters in a single channel.
• I've tried bandpass filters on the aux. subs (in place of low-pass) so as to surgically target problematic frequencies.
• I've just figured out how to parallel bandpass filters so I'll fiddle with this today (normally multiple BPs on one channel operate in series so very little gets through).
• Stuffing every port with every pair of socks in the house.

What's not causing problems:

• Sub/mains alignment. This is child's play -- takes five minutes after moving subs around.
• Deep bass. All problems disappear under 37Hz.

I've also pondered how much easier this would be with larger mains that dig down to 45-50Hz. I sit only 4ft away so direct radiation would likely brute force a lot of this nonsense into oblivion. What's a good cheap near-field passive with an 8" woofer?

 

[richardm]

An update to my post from Sunday...

I've had a breakthrough that I don't fully understand. Set mains to 19ms delay as my baseline. Then I advanced every sub by one full wavelength at 80Hz (12.5ms). From there I muted all but one sub and the mains. Playing an 80Hz sine wave I incrementally advanced further the sub-under-test until max volume was realized (i,e, that sub was maximally "helping" the mains). Repeated this for each sub+main combination -- I have 4x subs right now.

When done I unmuted everything, went to 40Hz (where the room's worst null exists) and muted one sub at a time to ensure all were contributing and none were subtracting. Also tried every combination of sub pairs together. Found one of the four in "conflict" so I reversed its phase, went back to 80Hz and repeated the mains agreement exercise on this guy. With this new timing I went back to 40Hz, made sure everyone was in agreement.

From here I tested 30Hz since it's at the very low-end of what my "auxiliary" subs can reproduce -- just to make sure none are fighting against the "big" sub.

Whatever I've stumbled upon IT WORKS. The 58-72 phase flip bullshit swirling around my head: GONE. Sharp 38-40Hz suckout, GONE. Little dip at 115-118Hz: GONE. All-pass filters to clean up 40Hz were causing issues in the 45-55Hz region -- with AP filters not needed these problems are also GONE. I've also had trouble in the 70-100Hz region caused by my too-small mains whose natural rolloff begins at 100Hz -- the subs could never fill this hole in a way that sounded right. Well, they're filling it now and it sounds 100% natural.

Sub delays with mains held at 16ms: 5, 6, 2, and 1ms. My next task is to eliminate one sub without screwing any of this up.

As for locations:

• PEBCAK/disgruntled audiofool throne is roughly 18" behind and 18" to the left of room center (this is the likely source of many problems but this desk is something special aka. heavy and moving it is the last resort).
• Main sub on floor directly in front of desk, centered with desk.
• Aux sub 1 behind PEBCAK, ~5ft off the floor, offset 1ft to the left which puts it at the rear wall's 1/3 mark longitudinally.
• Aux sub 2 behind PEBCAK and to the right, on floor, against rear wall at its 2/3 mark.
• Aux sub 3 to the right of PEBCAK and 1ft forward, on floor, against desk.

This is not as random as I'd like. Main sub and sub1 are almost longitudinally aligned, sub1 and sub2 are transversely aligned, sub2 and sub3 are almost longitudinally aligned, and main + sub2 + sub3 are all on the floor. I'd like to get sub1 above the room centerline and maybe lift another off the floor by ~2ft.

As a hopeless basshead I'll also try placing the main sub in whatever location provides maximum "room gain" under 35Hz.

I'm thankful I have only one listening position here in my office. I don't know what this sounds like 3ft away and I'm afraid to try it out.

 

[DaveBoswell]

An update to my post from Sunday...

I've had a breakthrough that I don't fully understand. Set mains to 19ms delay as my baseline. Then I advanced every sub by one full wavelength at 80Hz (12.5ms). From there I muted all but one sub and the mains. Playing an 80Hz sine wave I incrementally advanced further the sub-under-test until max volume was realized (i,e, that sub was maximally "helping" the mains). Repeated this for each sub+main combination -- I have 4x subs right now.

When done I unmuted everything, went to 40Hz (where the room's worst null exists) and muted one sub at a time to ensure all were contributing and none were subtracting. Also tried every combination of sub pairs together. Found one of the four in "conflict" so I reversed its phase, went back to 80Hz and repeated the mains agreement exercise on this guy. With this new timing I went back to 40Hz, made sure everyone was in agreement.

From here I tested 30Hz since it's at the very low-end of what my "auxiliary" subs can reproduce -- just to make sure none are fighting against the "big" sub.

Whatever I've stumbled upon IT WORKS. The 58-72 phase flip bullshit swirling around my head: GONE. Sharp 38-40Hz suckout, GONE. Little dip at 115-118Hz: GONE. All-pass filters to clean up 40Hz were causing issues in the 45-55Hz region -- with AP filters not needed these problems are also GONE. I've also had trouble in the 70-100Hz region caused by my too-small mains whose natural rolloff begins at 100Hz -- the subs could never fill this hole in a way that sounded right. Well, they're filling it now and it sounds 100% natural.

Sub delays with mains held at 16ms: 5, 6, 2, and 1ms. My next task is to eliminate one sub without screwing any of this up.

As for locations:

• PEBCAK/disgruntled audiofool throne is roughly 18" behind and 18" to the left of room center (this is the likely source of many problems but this desk is something special aka. heavy and moving it is the last resort).
• Main sub on floor directly in front of desk, centered with desk.
• Aux sub 1 behind PEBCAK, ~5ft off the floor, offset 1ft to the left which puts it at the rear wall's 1/3 mark longitudinally.
• Aux sub 2 behind PEBCAK and to the right, on floor, against rear wall at its 2/3 mark.
• Aux sub 3 to the right of PEBCAK and 1ft forward, on floor, against desk.

This is not as random as I'd like. Main sub and sub1 are almost longitudinally aligned, sub1 and sub2 are transversely aligned, sub2 and sub3 are almost longitudinally aligned, and main + sub2 + sub3 are all on the floor. I'd like to get sub1 above the room centerline and maybe lift another off the floor by ~2ft.

As a hopeless basshead I'll also try placing the main sub in whatever location provides maximum "room gain" under 35Hz.

I'm thankful I have only one listening position here in my office. I don't know what this sounds like 3ft away and I'm afraid to try it out.

If you are up for it and are looking for a much faster way to iterate through infinite permutations to find optimal solution, I highly recommend you use Multi Sub Optimizer (aka MSO). This will completely eliminate the manual and very cumbersome trial and error approach and can run through tens of thousands of permutations of filter settings including AP and LPF/HPF filters in only minutes. https://www.andyc.diy-audio-engineering.org/mso/html/index.html

There numerous threads here and on AVS pertaining to MSO to help you out.

 

[richardm]

There numerous threads here and on AVS pertaining to MSO to help you out.

I need a cheap mic. I can't find a US source for that Primo electret capsule they're buying on the other side of the pond...