[Housteau]

I am considering modifying the treatments to my room.  All of my efforts so far have worked out well to help minimize my room effects.  However, I still have two dips centered around 47 and 97Hz that I would like to improve upon.  According to my calculations they are both tangential in nature.  While I could boost up my corner treatments, which I may do anyway, I am not sure if that would help these particular areas.

My corners now consist of triangular cut sections of 703 from floor to ceiling.  I am considering adding 6 more inches to that total depth into the corner.  From the center of that diagonal treatment it would then be about 24" thick.

Another option would be to redesign my ceiling treatments, which I feel may work better on those tangential issues.  This is what I have now:



There are more photos under my page links.  These are 2 x 2 foot frames 4 inches thick of 703 that straddle my beam running the length of the cathedral peak.  They are spaced apart a bit to allow room for the light fixtures, and there is also a 2 x 4 foot panel used there as well.  I was thinking that to make improvements these changes would need to be fairly substantial.

My thoughts were running alongs these lines:  What if I straddled that cathedral peak with a drop ceiling a bit below that center beam.  That would give a decent space that could be filled with 703 supported by standard drop ceiling acoustic panels.

Am I on the right track here, or do I need to be locked up somewhere?

[Ethan Winer]

I still have two dips centered around 47 and 97Hz that I would like to improve upon.  According to my calculations they are both tangential in nature.

It's possible those nulls are tangential, but you really need to measure in several locations in the room to be sure they're not just positional. That is, you measure where you listen, then move the microphone six inches forward or back and see if the frequencies change. Then measure again two feet forward and/or back. If the frequencies change, they are not related to room modes.

Also, mode calculations are not as accurate as you might think. I've seen calculations be off 20 percent or more from what's actually measured. In this case measuring always trumps predictions. Further, your room is not a rectangle, so it may not even be possible to calculate all the modes correctly. Certainly not any modes that involve the ceiling.

What if I straddled that cathedral peak with a drop ceiling a bit below that center beam.

That can only help, but there's a more scientific way to identify the best place for bass traps as described here:

Pink noise aids placing bass traps

If you determine by measuring that the null frequencies are the same at different locations, then you can use sine waves at those frequencies instead of the bassy pink noise from that article.

--Ethan

[Housteau]

Thank you Ethan.  That is some very good information and why I am here.  The dip at 97Hz may very well be positional due to my listening position, but I will not be sure until I get back home next week and do some testing.

It's possible those nulls are tangential, but you really need to measure in several locations in the room to be sure they're not just positional. That is, you measure where you listen, then move the microphone six inches forward or back and see if the frequencies change. Then measure again two feet forward and/or back. If the frequencies change, they are not related to room modes.

To be sure I understand you correctly, if the null remains but changes in frequency, then it is positional.

[oneinthepipe]

I haven't taken measurements, but my room is boomier than I would like.  I downloaded Ethan's pink noise file and walked around the room, and there are areas that are noticeably boomier than others, particularly in the corners and the sides of the speakers.

I already have 6" thick OC703 panels floor-to-ceiling in the corners and 6" OC 703 panels behind the speakers.  The speakers are 3' from the side walls and 3.5' from the front wall.  My room is only 205 square feet, and I can't move the speakers out much further from the front wall.  I can move them inward from the side walls a little.

Any suggestions?  I can't enlarge the room.

[oneinthepipe]

I pulled my speakers out from the front wall.  The from baffle is 45" from the front wall but since the speakers are 17' deep, the back of the speaker is on 28" from the front wall.  I also moved the speakers away from the side wall.  The driver center's are 43" from the side walls, and the speakers' drivers are 74" apart.  I have moved my chair back, and the drivers are 74" inches from my ears, bit my ears are only 23" from the back wall.  There is less boom, but I am sitting very close to fairly large full range speakers, and because they are 16" closer, I am giving up quite a bit.

I was thinking about ordering (2) or (3) more 12 panel cases of 2" OC 703 tomorrow from Kamco in Woburn, MA.  I am not sure where I will place ( or (12) more 6" panels, but I assume that they can't hurt anything.  I don't know whether or not I should try to stretch some plastic over the front to make them into bass traps, of sorts.

[Scottdazzle]

This may not be possible in your room, but having your speakers the same distance from side and back walls reinforces certain frequencies.  Placing your speakers so that the right and left woofers are different distances from the walls can help smooth out bass response.  It is also a good idea for those distances to not be multiples of each other (e.g. 2' and 4' bad; 3' and 4' good).

[Ethan Winer]

To be sure I understand you correctly, if the null remains but changes in frequency, then it is positional.

Yes, exactly. Peaks and nulls due to modes will change intensity as you move the measuring microphone around, but the frequency will stay the same. I meant to link this article yesterday which provides a sort of case study:

http://www.realtraps.com/art_modes.htm

This video also addresses non-modal peaks and nulls:

http://www.realtraps.com/video_wave.htm

In fact, I made the above video after arguing in a forum with some "pro" acousticians who insisted that all peaks and nulls are due to modes.

--Ethan

[Ethan Winer]

I already have 6" thick OC703 panels floor-to-ceiling in the corners and 6" OC 703 panels behind the speakers.

Rectangle rooms have 12 corners, not only four. The more total corner surface you treat, the flatter and tighter the response will be. It's that simple. The front and rear walls are also good locations for more bass trapping. Though you don't have to go nuts like I did with 51 panels (so far) in my living room.

--Ethan

[bpape]

What boundaries are the potential tangentials involving?  If it's 4 walls, that's possible, or as Ethan said, it could be positional.  If the theoretical tangentials are involving the height at all, it's doubtful as your ceiling is not flat and parallel with the floor the way most of the calculations assume.

Bryan

[Housteau]

What boundaries are the potential tangentials involving?  If it's 4 walls, that's possible, or as Ethan said, it could be positional.  If the theoretical tangentials are involving the height at all, it's doubtful as your ceiling is not flat and parallel with the floor the way most of the calculations assume.

If I use this mode calculator: 
http://www.bobgolds.com/Mode/RoomModes.htm
and with my room dimensions plugged in, it highlights my trouble spots exactly how I measured them from my listening position.  Since I have a cathedral peak I did use an average of 10' for one calculation and 12' for another.  It would be 12' above my listening position.  Either way I did the calculation, the trouble spots were still in the same place and highlighted as tangentials.  If I change the ceiling height to 8' the dip at 45Hz still remains, but the one at 97 drops down to 95Hz as being the trouble spot.

However, my listening position is also 1/4 wavelength of one (front to back)and about 1/2 of the other (side to side).  So, it could be a combination of mode and position.  I guess I will not know until I get back home to do more testing.

[Kevin Haskins]

I'm not wise but I also would think about the proportion of bass traps to other reflected surfaces in the room.    I like a fairly live room for the mid range & up.     Unfortunately, as you add more absorption to tackle bass issues <300Hz, you are almost always adding more absorptive capability from the mid range on up.

So... I'd experiment before adding anything permanent.    That way you don't put in a drop ceiling, and end up with overall worse subjective results although have slightly better measured bass response.   

[Ethan Winer]

Rectangle rooms have 12 corners, not only four.

In response to a PM asking me to clarify:

There are four corners where each wall meets another wall, four corners where each wall meets the ceiling, and four more where each wall meets the floor. So that's 12 corners total, but eight tri-corners (which is something else).

--Ethan

[Housteau]

I'm not wise but I also would think about the proportion of bass traps to other reflected surfaces in the room.    I like a fairly live room for the mid range & up.     Unfortunately, as you add more absorption to tackle bass issues <300Hz, you are almost always adding more absorptive capability from the mid range on up.

So... I'd experiment before adding anything permanent.    That way you don't put in a drop ceiling, and end up with overall worse subjective results although have slightly better measured bass response.   

Agreed.  Most of my room treatments have been DIY and I did make them with what you had to say here in mind.  About 1/2 of them were made to be reflective of the higher frequencies.

[Housteau]

Speaker positioning directly effects the intensity of room modes.  My current placement offers the best balance with the flattest response.  Can the positions of the speakers also have an effect on non-modal peaks and nulls?

Is broadband absorption still the best answer to the non-modal problems?  Besides changing the listening position, which may only shift the null to other frequencies, is there a solution, or is this just one of those things to live with? 

[MaxCast]

What about a Helmholtz resonator?  Bryan mentioned these before.  They can be tuned to a specific frequency.

quote from Real traps:

Finally, RealTraps are designed to absorb a wide range of low frequencies. Some bass trap designs are based on a Helmholtz resonator which is tuned to a single problem frequency. While a Helmholtz trap can reduce standing waves at the one frequency it's tuned to, it does nothing for the harmonically related frequencies nor will it reduce low frequency reverb across the entire bass range. Since RealTraps products have a broader range they are appropriate for any size room, and they also reduce reverb time at all frequencies making music played in the room sound much clearer.

[bpape]

That's possible certainly - at least for the 90ish problem.  The lower one is still OK but approaching the range where Helmholz have their own issues.

Bryan

[youngho]

Speaker positioning directly effects the intensity of room modes.  My current placement offers the best balance with the flattest response.  Can the positions of the speakers also have an effect on non-modal peaks and nulls?

Yes, comb filter and adjacent boundary (SBIR) effects are non-modal and related to speaker position (also boundary +/- listener position), but they can contribute to peaks and nulls. This is what Ethan was referring to earlier.

Is broadband absorption still the best answer to the non-modal problems?  Besides changing the listening position, which may only shift the null to other frequencies, is there a solution, or is this just one of those things to live with?

It depends. Diffusion can be a relatively expensive way to deal with some of the first reflections (only a portion of the original reflection reaches the listener, since much of it may be re-radiated in a non-specular manner, depending the wavelengths involved and the diffuser design), but increasing depth is needed for lower frequencies. There's also what I consider to be diffraction, which is any obstacle to the sound that does not result in re-radiation in an essentially cylindrical or hemispheric pattern (obviously, specular = reflection). If the floor is carpeted or there is a thick rug, strategically placed and angled wall decorations (say, a larger glass-paned print or photograph) may reflect much of the higher frequencies away from a direct path to the listener, like downward into the floor covering or upholstered furniture. When it comes to the bass frequencies, besides broadband absorption and Helmholtz resonators, there are a few diaphragmatic absorbers on the market, like the RPG Modex Plate or Modex Broadband.

I'm not saying that any one is best,  just responding to the specific questions quoted above.

[Ethan Winer]

Is broadband absorption still the best answer to the non-modal problems?

Yes!

--Ethan

[Housteau]

I finally made it back home and have had a chance to do some testing.  The 47Hz dip is actually centered at 45Hz and is modal.  Mathmatically it is tangential from the four walls with my cathedral peak not being involved.  The dip at 97Hz was positional.

I was not able to do much about the dip at 45Hz, but by adjusting the positions of my listening position and speakers I was able to shift the dip at 97Hz up to 103Hz while improving it by better than half.  It is fairly narrow and only down by 5 db now.

My testing only went up to 120Hz and so I am sure I have more issues left to deal with.  But, the solution looks to be the same for them as it is for my one down at 45Hz.  That would be a room treatment solution.  My plan is to add more thickness to my front wall first reflection panel and possibly one behind my listening seat as well.  I am thinking that center wall treatments will help with that tangential problem at 45Hz.  Am I on the right track this time?

[Ethan Winer]

I am thinking that center wall treatments will help with that tangential problem at 45Hz.  Am I on the right track this time?

That mode is probably not tangential. It's more likely an axial mode, but not at the exact frequency predicted. Modes as measured are often different from what's predicted due to wall construction and absorption. I've seen them differ by 20 percent or more! Regardless, bass trapping is the answer, as always. The best way to find where a trap should go is to play a sine wave at that frequency (45 Hz) while you walk around with an SPL meter. Wherever the level is highest is a good location for a trap.

--Ethan