[sebrof]

My room is 13' x 12' x 9'high, dedicated spare bedroom, 1 door and 1 closet.
Pasted below  is a shot of my response as measured with the Rives Audio test disc measured with a RS digital meter about 8' from the speakers (I ran tracks 32 and up which are the ones corrected for the RS meter). I baselined at ~85db @ 1KHz because that's about where I normally listen to music.
The suckout beginning at 125Hz is what I'm asking about - I assume this is something that should be addressed. Any suggestions?
What about the rest of the spectrum - Good, decent, bad?
I can't make the 125Hz better or worse, I've moved all treatments out, moved them all over the room, I opened the door, moved my speakers around, etc. Can't change it by more than 2 db or so.
I measured the response of the 100Hz to 125 Hz about 2' from the speakers and the drop is 2 to 3db (vs. -13db @ 8').

Thanks for any suggestions.

[Jeffrey Hedback]

Hello sebrof,

A null only has so many possible caiuse (good news), but these causes can be tricky to identity/fix (bad news).
Cause include:
- spkr to boundary cancellation: if your speakers are ~ 2'-3" from a surface, this could be the cause.  Moving the spkr approx 1' and remeasure with mic in same location would tell a great deal.
- listener to boundary cancellation: if your ears are ~ 2'-3" from the rear wall, this could be your trouble.  Moving the mic approx 1' with spkrs in same location would give insight on this.
- Node(s): either/both your speakers or ears could be in a node location.  This would most certainly be the case if the above two are not the cause.  At ~125Hz, acousticall traps can have great results.

Here's a nifty visualization tool of room nodes:
http://www.hunecke.de/en/calculators/room-eigenmodes.html

[neekomax]

I had the same kind of thing going on, right at 150 Hz. Very annoying, considering I could really hear it  .

In my case, it was my speakers. With my new speakers, I have no such problem. Do you have a second set of speakers you could test and see if you have the same issue?

[sebrof]

Jeffery - I did move my speakers around two or three times by a foot or two each time.
But I will try moving the mic into the room a little. My head is at the back wall, 2 GIK 244s hanging behind me.

Neekomax - I do have another pair of speakers I can try. Good idea.

Thanks guys for the responses.

[neekomax]

Let us know what you figure out, and good luck!

[richidoo]

That it worsens when you measure further away from the speaker says that your speakers are fine, and the dip is room mode related.

Realtraps' Modecalc says that there is a mode at 125Hz in the height axis (9',) and closeby is another mode at 130Hz in the length axis (13'.) There are other medium strength modes above the dip. Tips on using modecalc. 

Change the height of the measurement mic at the listening chair or near a speaker to see how much role height is playing. If it is, then adjust the height of the speakers. Tweeter height anywhere between 2" below your ears to 4 or more inches above your ears is not too objectionable to me, but only you know what you can tolerate. Try to find the axis which affects the dip the most, and don't forget reflections, and the room's mechanical resonances like floors, walls, hollow doors and big windows ringing in sympathy.

Only test one speaker at a time when looking for the cause of modes. Put the speakers wherever the modes are least, then adjust your chair and their toe to form your direct sound triangle with tweeters equal distance from your ears. It is easy to treat high freq reflections to cure unsymmetrical listening triangle but harder to fix low freq modes with passive treatments.

http://www.mcsquared.com/wavelength.htm
http://www.libinst.com/SynRTA.htm

[Jeffrey Hedback]

Given the size of the room the 2,2,0 Tangential is a highly likely culprit.  That mode would be damped greatly by mast typical vertical corner traps.

The 2nd order height axial mode would place ears/mic at ~ 2'-3" (hey we have a theme going huh) above the floor.  It's more likely his ears are 3'-ish above the floor where that mode is not as likely to cause such a dip.

[sebrof]

So at first I had the meter on a tripod sitting on my chair, 3'8" above the floor. This is actually about 6" higher than my head when I'm seated (didn't think that height would matter so much). So I lowered to my head height (~3'2").
Whether at my head height, the original high height, or at gut level, there's still a dip of ~ -10db but the freq of the dip changes. So I assume that if I change the height of the speakers I'll just change where the dip happens.

Jeffrey, if you could explain the 2,2,0 Tangential I'd appreciate it, or if it's in the links provided by richidoo I'll find it.

I have 2 GIK corner traps at the front wall behind speakers and 2 GIK 244s hanging behind the listener.
There's a door at the corner behind and left of the listener, and the other back corner is truncated (think of a 2' board straddling the corner, I assume there's some sort of electrical duct or something) so no chance for corner treatments in back corners. I am considering 2 more corners to sit on top of the existing corner traps.

Thanks again for any input

[Jeffrey Hedback]

Hi sebrof,

If you convert your room dimensions to meters, go to the link I sent previously and run the calculator you will see a visualization of high pressure (dark) and low pressure (white) areas of each mode.

Axial modes travel back/forth between two surfaces.  Tangential wrap/bounce around four surfaces.  The 2,2,0 is the second order tangential mode with the width and length dimensions of your room.

[richidoo]

Given the size of the room the 2,2,0 Tangential is a highly likely culprit.  That mode would be damped greatly by mast typical vertical corner traps.

I had read about that before but didn't remember what it meant, so I googled it and found this basic explanation of modes.

And this cool mode calculator. The legend explains things a little.

The 2nd order height axial mode would place ears/mic at ~ 2'-3" (hey we have a theme going huh) above the floor.  It's more likely his ears are 3'-ish above the floor where that mode is not as likely to cause such a dip.

Jeff is that because the null of 2nd harmonic is at quarter height? Then what would we expect to find (in theory) at half height and at the reflection points? Thanks
I like your website!
Rich

[JohnR]

I've provided an explanation of modes in the theory section of this article: http://www.hifizine.com/2011/09/prototyping-dipole-bass-system/

[sebrof]

I'll pursue fixing this, just not so easy to keep playing tones and measuring when there are other people in the house like tonight.

If we take a step back - How does the room look overall? I think it sounds pretty good, if I smooth out the suckout how much will I notice?
I assume that many if not most rooms look as bad or worse?
The measurements above about 1KHz look like they jump around but I've always found the hi freqs to be finicky regarding mic placement and also when I lean in to look at the meter my body affects the reading by several db.

[richidoo]

I've provided an explanation of modes in the theory section of this article: http://www.hifizine.com/2011/09/prototyping-dipole-bass-system/

Thanks John!

[doug s.]

since your room is a dedicated listening room, you can try asymmetric speaker placement.  (i am assuming you have not tried this.)  i.e.:  vary the distance your left speaker is from the left wall vs the right speaker's distance from the right wall.  and/or vary the distance the left speaker is from the rear wall vs the right speaker.

you can also try a diamond-type arrangement; i.e.: having the corners of the room in front of you, behind you, and to the sides, instead of the flat walls...  and again, slight asymmetry may still help...

doug s.

[Danny Richie]

The number one most common mistake I see when people try to take a room measurement is that they try to measure both speakers at the same time. This often causes peaks and dips due to the comb filtering effects created by distance variations between the two of them.

Be sure to only measure one of them at a time.

[InfernoSTi]

The number one most common mistake I see when people try to take a room measurement is that they try to measure both speakers at the same time. This often causes peaks and dips due to the comb filtering effects created by distance variations between the two of them.

Be sure to only measure one of them at a time.

Danny,

Thank you for your comment. Would you be able to please explain further?  Theoretically, I can "get the room right" one speaker at a time but then when I listen with both, wouldn't I have the same problems I've avoided seeing by only measuring one at a time?  What am I missing to understand this, please?  I'm sorry I'm a newbie at the room treatment/measurement thing....

John

[Danny Richie]

Danny,

Thank you for your comment. Would you be able to please explain further?  Theoretically, I can "get the room right" one speaker at a time but then when I listen with both, wouldn't I have the same problems I've avoided seeing by only measuring one at a time?  What am I missing to understand this, please?  I'm sorry I'm a newbie at the room treatment/measurement thing....

John

When you have a signal coming from two separate sources and it is the same mono signal then you will have out of phase cancellations all over the place. And then it is hard to pin point what is causing it. You don't know if it is a left side side wall reflection, a right side wall reflection, or a cancellation between the two speakers. It can easily be a ceiling or floor reflection causing the cancellation two.

Just move to the left or right with your mic and you can watch the responses change as you get closer to one speaker and farther away from the other. You can use this wavelength chart to help figure out the distances. http://www.soundoctor.com/freq.htm So you then know that higher frequencies can be canceled out with comb filtering effects by small movements as the wavelengths are short. Longer wavelengths take a greater distance differential to cause the cancellation.

Things are different for music play back because the left and right signal can be very different. We listen to a stereo signal and not dual mono. So cancellation effects are not all in one spot based on distance and wavelength because the signal is often shifted to one side or the other. This signal variation to the left and right channels is what gives us the sound stage layering and depth that we all love so much.

So, measure one speaker at a time. And don't be afraid to vary your measuring distance so that you can see a pattern form and figure out what the cause and effects might be. It might look very different than when you shot them both together.

[Jeffrey Hedback]

I would actually disagree or better put...like to add more context.

The biggest problem (to me) is not at all that people measure both speakers, it may be that they only measure both speakers and expect that to show the full story.

The measurement of both speakers is valid and representative below ~250Hz (if you say 300...no worries).  At those wavelengths there are not going to be cancellations that are purely measurement anomalies.  The data would be representative of the experienced in-room LF Response.

Above that frequency zone, measurements must be one speaker only to review frequency response, ETC...etc for the exact reasons Danny states.

To truly find cause/effect, having both single speaker and two spkr LF data is VERY helpful.  Much can be done in comparative studies when going deeper into time domain aspects, excess group delay & various gate lengths.

[rab1234]

Phasing?

[Danny Richie]

Hey Jeffrey,

My speakers are about 9 feet apart. So if I go to my first reflection point on the wall (either wall) and from that point one speaker is at least 8 feet (nearly 9 feet) further away than the other. That is a wavelength of between about 125 and 140Hz. So it is real easy to see that there can be cancellation effects going on in the room response that is a result of both speakers playing the same signal that are easily down into ranges of 100Hz or even less in a big room.