Axiom Home Page Forums General Discussion Technical Questions Big ugly room resonances

I rearranged the room a few days ago to make it more "people friendly" but the sound doesn't seem as good after the move. I popped in the Avia disk and ran through the sweeps with the SPL meter going and had some bad peaks (>10dB) around 100 and 160 Hz, even with the M2s which I *know* are flat in that region.

I measured the room (24 x 13 x 7'4) and stuffed the results into a mode calculator in case the room resonance harmonics were adding up but didn't see any particular problem :



The room is in the basement, floor plan follows. Relative to the following picture, the room has patio doors behind heavy curtains on the left, concrete block behind wood panelling on the right and top, largely open at the bottom.



It's been 25 years since I have run around with test equipment to measure room resonances so everything is rusty but I'm used to seeing *some* kind of relationship between calculated room resonances and observed response. I'm not seeing any kind of pattern here.

Other than sticking in some mid-bass traps, any ideas ? I'm going to do the "subwoofer crawl" to try and fix some low bass peaks but really not sure what to do about the 100 and 160hz peaks. Any advice appreciated.

Thanks,
John

John, looking at those numbers it doesn't appear that room modes would explain the peaks which you measured. Assuming that the measurements were accurate(did you actually hear the peaks at those frequencies?), it possibly could be related to boundary reinforcement(which is entirely separate from room modes)at those frequencies. The sound bouncing off the nearest room surfaces(floor and two walls)reflects back to the woofer cones and can either increase or partially cancel output at frequencies related to the distances of the cones to those boundaries. For this reason the distance of the woofer cones to the floor and two walls shouldn't be nearly identical, but rather as dissimilar as possible, so that the boundary peaks and valleys don't pile up at the same frequency. Maybe this isn't the factor involved, but you could temporarily move the speakers to appropriate distances to see if the peaks are smoothed out.

Thanks, John. That might explain why the sound seemed worse in the same room after I rearranged everything...

The peaks were REALLY obvious to the ear. Quiet... louder... louder... REALLY LOUD... quieter.. quieter.. no, louder, louder, HOLY %$^%$, quieter. I grabbed the SPL meter just in case I was imagining things or my ear's response was that far off.

Just curious. Doesn't a room mode calculator only work with a square or rectangular room?

If there are any alcoves, hallways, or other rooms open to the area wouldn't they need to be taken into account for?

Yep... all you can really do is draw out all of the possible opposing walls where standing waves can build up (including angles and corners if you're bored) and work out the frequencies for all of them.

The most common thing, though, is not so much "measure the room to predict the behavior" but "measure the frequency response then look at the bad spots, convert those frequencies into distances, and go figure out what you can re-arrange to attack the problem".

In my case I think I committed the crime of having many speakers roughly the same distance from a wall. Not 100% sure but after I shuffled distances and moved the surrounds to the right (on the pic above) away from the opposing walls things got a lot better.

I bought a measurement mic and the TrueRTA software to calibrate my sub with. I have a peak at 63Hz that wasn't there a few months ago. Drove my self nuts trying to fix it with the sub EQ and positioning. Turns out that the M60s were causing it. Fixed it by moving them 5", but all I did was move the peak to 80Hz. The hump at 63 sounds better, so I moved them back. My room stinks. I need some bass traps, I think.

What's the formula?

Freq = distance / speed of sound ?

>>My room stinks. I need some bass traps, I think

That seems to be the only general solution, although sometimes you get lucky and can shift a peak over enough to fill a hole related to a different speaker/wall distance (sort of like landscaping). I'm planning to move in 6 months but am tempted to at least play with some traps now.

>>What's the formula? Freq = distance / speed of sound ?

Wavelength = speed of sound / Freq, or Freq = speed of sound / distance.

If the round trip distance between the walls is one wavelength (ie your walls are 1/2 wavelength apart) you get a peak, ditto at 2x, 3x, 4x the frequency etc... If the round trip distance is half a wavelength (or 1.5, or 2.5 etc..) you get a null.

Same deal with speaker-to-wall distances -- if the speaker is 1/2 wavelength from the wall (ie 1 wavelength ROUND TRIP) you get a peak.

The thing I'm trying to wrap my head around is that the peaks and nulls don't seem to follow the same pattern, eg. :

Speaker is 5 feet from the side wall :

- peak when 1,2,3... wavelength = 10' (2x5'), ie 114 Hz, 228Hz (2x), 342 Hz (3x) etc..
- null when 1/2, 3/2, 5/2... wavelength = 10', ie 57 Hz, 171 Hz (3x), 285 Hz (5x) etc...

Your 63 Hz peak corresponds to having the speaker about 9 feet from some wall. I wouldn't expect that moving 5" would shift the peak that much unless you were talking about moving them relative to the back wall, where 5" makes a HUGE difference in some rooms but that is a more subtle interaction than simple room resonances and I don't know the math for that

Take all of the above with a grain of salt. I haven't done this stuff in 25 years so there's a LOT of rust. If anyone jumps in and disagrees they are probably right...