Killing 77 Hz... Helmholtz... membrane?

Hello! I have quite a large null at about 77 Hz in my control room that I need to absorb. I believe this one is a mode between floor/ceiling but also between two of the walls. What kind of trap would I use for that? I already have corner chunks in all wall-to-wall corners. First, I thought that a membrane/panel absorber would be easy to build but I understand that it's difficult to mount the panel to the frame correctly... or is it? Anyway, someone recommended me building a perforated Helmholtz absorber. But do I need a sealed cavity for that? I never really understood that. Also, when I use Chris Whealy's spreadsheet calculator, I can't seem to get values that are useful. From what I understand, we're talking lots of holes but to get down to 77 Hz, I need either really few holes or a really large air gap. Right now, I'm thinkng about building a frame over which I put a 20 mm MDF board in which I drill a few but larger holes. From what I understand, it's the total area of the holes, the depth of the holes (in this case 20 mm) and the volume inside the box that determines the freqency? Does any of this make sense and what (correct) formulas would I use in calculating them?

I'd start with more broadband absorption in more room corners (wall/ceiling corners, or even wall/floor corners such as behind the desk). What are your room dimensions? It'd be interesting to see which dimension (if any) has a null at 77Hz, which translates to a 14.67' (4.46m) wavelength. If this is the front-rear dimension of your room (and actually even if it isn't) then I recommend quite a bit of absorption on the rear wall. Also, you can experiment with layout and placement, it's possible your listening position or speakers are in a null point. If after all these steps you still have an issue at 77Hz, then I'd consider a pressure absorber such as a Helmholtz. There are calculators on this site.

jwl wrote:

What are your room dimensions? It'd be interesting to see which dimension (if any) has a null at 77Hz, which translates to a 14.67' (4.46m) wavelength.

That is my first thought. Plus reflections ~7 ms longer than initial sound. Andre

4.20 x 3.20 x 2.50 m Right now, the rear wall is covered by a shelf, the studio window and a fireplace in the corner which is the reason the studio windows is behind me. I don't know if I used the optimal insulation for the corner chunks though. I took whatever insulation I had lying around as I don't have a lot of money to spend at the moment. My speakers are out about 85 cm from the wall (the baffle) and have a height of about 110 cm (the woofer). Also, the calculator to which you linked only calculates slat type Helmholtz resonators. From what I understand, it's hard to get down to 77 Hz with these. I moved the desk right now another 10 cm out from the wall (can't move more as there's a door on the left side that will be blocked then. I'm gonna measure again (I don't have a measurement mic but my Oktava will have to do... I can at least compare it to previous measurements and I do see the peaks and nulls clearly). About the 77 Hz mode, when playing a 77 Hz tone and moving around the room it gets louder if I lie on the floor or stand on a chair with my head to the ceiling. It also gets louder close to the front and rear walls. As you might understand, this room is far from optimal for use as a control room but it's what I have.

shadewind wrote:

My speakers are out about 85 cm from the wall (the baffle) and have a height of about 110 cm (the woofer).

And what height is your listening/measuring position?

Also, the calculator to which you linked only calculates slat type Helmholtz resonators. From what I understand, it's hard to get down to 77 Hz with these.

Which is what most Helmholtz absorbers are. Play with the numbers in the calculator. For lower frequencies, use deeper slots. Andre

Edited my last post some... My listening position would something like 160-ish cm from the wall and the height of course 115-ish. So slats are the way to go? My idea of a box resonator (sort of like a ported speaker cab without a speaker) would be an ineffective design? EDIT: It seems like moving the desk and the speakers acutally made things worse. Moving a bit too close to the monitors (no equilateral triangle that is) actually makes the null not so deep.

shadewind wrote:

It seems like moving the desk and the speakers acutally made things worse. Moving a bit too close to the monitors (no equilateral triangle that is) actually makes the null not so deep.

Now you have me completely confused. The null, which was your original concern, is reduced, but you write things are worse. :shock: Andre

Sorry about that. It's like this: When I moved the desk out 10 cm it didn't get any better and it might have gotten a little worse over all in the spectrum. BUT... it seems that when the microphone (or my head for that matter) is a little closer to the monitors than the sweet spot (that is, regardless of the distance to the wall) the null gets a lot better, the tips seems almost chopped off... looks weird. What I'm trying to say is that the monitor position does mostly nothing but the mic position in relation to the monitors does something allright. But then again, this might be my inexact placement or something else... dunno... you tell me :)

You are now confusing me more. If the response by the movement is improved, then it is improved. Please explain "away from the sweet spot." Andre

Nulls cannot be solved with absorption,,,, so change your location of spakers and listening point. Peaks can be solved with a membrane or other resonant absorbers

AVare wrote:

You are now confusing me more. If the response by the movement is improved, then it is improved. Please explain "away from the sweet spot." Andre

Sorry again :) Moving the monitors themselves doesn't do much of a difference. Moving the listening position in relation to the speakers does difference. That is, if I get closer to the speakers the null is chopped off. Conclusion:

*If I move both the speakers and the listening position forward, not much happens. *If I move the listenting position forward or backwards, the null is chopped off somewhat.

Nulls cannot be solved with absorption,,,, so change your location of spakers and listening point. Peaks can be solved with a membrane or other resonant absorbers

That sort of contradicts everything I've learned about physics and acoustics. Nulls are caused by standing waves (room modes) which form when a wave is reflected between two walls. In some places along this standing there will be a null at that frequency and at some places a peak at that frequency. The frequency is thus determined by the distance between the walls. By absorbing these standing waves you get rid of the peaks AND the nulls. At least that's what I have been taught.

Now you know where your "sweet spot" isn't! Yes listening position is critical also. Andre

andrebrito wrote:

Nulls cannot be solved with absorption,,,, so change your location of spakers and listening point. Peaks can be solved with a membrane or other resonant absorbers

Depends on what you mean by "solved". Nulls can certainly be reduced with absorption. Though I agree experimenting with layout can be extremely helpful.

Well, what I could do is space my monitors further apart which would move my sweet spot back as well. When I say sweet spot, I mean the stereo sense... that is, at the tip of the equilateral triangle formed by the monitors and the listener. I did an interesting experiment though. I placed a total of five 1,5 l empty PET bottles in the corners of my room. These have a resonant freqency of about a 100 Hz. I did measurements with and without them in the room with the mic and me in the exact same position and surely, the difference was clearly visible as the small peak there was smoothed out. Actually it was between a peak and a null so you could see them both smoothed out. Which speaks in favor of my idead that nulls can be smoothed out by absorption. Maybe I'll try and build these boxes with holes in them. I'll also remove the rear shelf and put a slat resonator there. Of course more broadband absorbers might help a bit as well as everybody's saying.

Why do they drill these holes into the sides of the frames? Do these holes have a calculated size?

andrebrito wrote: Nulls cannot be solved with absorption,,,, so change your location of spakers and listening point. Peaks can be solved with a membrane or other resonant absorbers Depends on what you mean by "solved". Nulls can certainly be reduced with absorption. Though I agree experimenting with layout can be extremely helpful.

_________________ That was an unlucky comment, I was sleepy and writing garbage. What I meant is that nulls and peaks in a small room are usually quite accentuated just solved by absorption, particularly broadband, which does not act in a specific bandwidth. Therefore the only options you are left are : move the speakers/listeners location and as a last solution use EQ. I'll post a picture of empty room measurements with accentuated peaks.

Here it is... an empty room and its frequency response measured in 9 points. SPL is not calibrated but it is quite evident the peaks and valleys and its separation in terms of magnitude. After the bass traps (on corners and walls/ceilings) the room was not as bad as shown above but those peaks and valleys still existed. The negative consequence was that the room was quite dead at that point. http://img46.imageshack.us/img46/9826/48574464jt7.th.jpg