How to get rid of a comb filter effect?

[Wegi76]

Hello there,

my desk is introducing a strong comb filter effect. The speakers are already angled up by ~7°and I also tried to put them onto a staple of books (15cm or so) ... Does not change much...

I get a broad dip around 900Hz, then again @2.5Khz and from 4K upwards there is the typical combfilter pattern. When I sweep a sine, I can also hear the flucutations in loudness, so it`s not just a cosmetic problem.

Most of this disappears when I put a plate of basotect on my desk. The 1ms impulse, the 900Hz mess, the combfilters... (see spl graph from right channel)... all gone more or less. Obviously I could not use the desk anymore, so not a real solution...

Please have a look at the data and i also uploaded a photo of the "normal" situation and with that basotect plate.

Do you have any advice how to get rid of these effects? Does it make sense to buy tall stands and to angle the speakers downwards instead of using the small stands? I guess, it`s the last thing I could try, right? Want to mention that I can`t angle the whole desk. This will have to remain in a 90° angle to the ground.

Thank you!


SPL comparison (1/24" smoothing), right channel only

Impluse (1ms reflection gone with basotect plate on desk)

Listening situation

Basotect test

 

[alex-z]

The desk will always have some influence. Taller stands, spread further apart, and angled down will help.

Although frankly that treble ripple is probably not easily audible without the sine sweep. I would be more worried about the 20-200Hz region which looks like the swiss alps in small rooms.

 

[Wegi76]

Thanks, Alex, for the advice.

20-200Hz region is reasonably smoothed. I have no space for bass treatment in my room but my MiniDSP & DIRAC already removed the peaks so that only the valleys (mainly 34 Hz) are left.

(OK, only if I sit in that single spot in the room i optimized the setup for, everyelse it`s a mess of course. But that`s how it is...

 

[Hipper]

I wonder if you could identify where the speaker is reflecting off the desk by moving some small piece of absorbent material on the desk until it affects the measurement?

If that works, the perhaps you can put some small pieces of absorbent material right in front of each speaker to catch that first reflection? For example, could such material be placed under the speakers on the angled stands and stick out enough to catch that reflection? Or even be used to make the stand like a wedge but stick out beyond the speaker.

 

[Wegi76]

Thank you too, Hipper, for that idea. Problem is, I really need all the space on the desk.

Guess, I`ll follow Alex`s advice and try taller stands and to angle the speaker downwards. Probably this model: https://www.thomann.de/de/gravity_sp_3102.htm

 

[dasdoing]

angling it down will probably make it worse cause the reflection will be more on axis.
the best would be to angle the desk down. I know...not realy an option

 

[Vladimir Filevski]

Do you have any advice how to get rid of these effects? Does it make sense to buy tall stands and to angle the speakers downwards instead of using the small stands?

It is impossible to get rid of the comb effect with conventional loudspeakers (like yours). It needs out of the box solution.
Here are my three efforts, plus one from Razer:
https://ieeexplore.ieee.org/document/1341680 (Brute force solution: oval or "racetrack" woofer in horizontal position touching the desk, and angled horn tweeter with narrow vertical dispersion)

https://www.aes.org/images/e-lib/thumbnails/1/3/13585_full.png (Elegant solution: dipole loudspeaker angled backward to “direct” the dipole null at the desktop to cancel the desktop reflection)

After this dipole desktop speaker, I was sketching my new invention on the drawing board. Tired, but happy with the new idea, I checked the news from Consumer Electronic Show 2007 when I saw this finished product from Razer with exactly the same solution as mine:
https://www.engadget.com/2007-01-10...5M4SgWlDs6PqsjDaIqizFNEKkij4Zpu1NWzZIPE7cTHDG

https://sudonull.com/post/186115-Razer-mako (Pay attention to the X-ray image of the ground plane/slot technology used in the satellite speaker)

https://www.aes.org/images/e-lib/thumbnails/1/4/14757_full.png ("Rabbit ears" solution: horizontally pivoting telescoping arm, one side attached to monitor top corner, other side attached to a small satellite loudspeaker)

 

[Wegi76]

Hi Vladimir,

Thanks for your input. You obviously thought a lot about this problem as the author of some of these papers. 1st link is behind a paywall I guess but I get the idea.

Can you maybe give an example how the "Rabbit ears" would look like please in practice? And where is the difference between a floating arm and a stand if the speaker is ultimately in the identical position realtive to the desk? (besides that it`s obviously easier to move the speakers and consumes no place on the desk) Are there commercial products that implement your idea?

Maybe something like one of these monitor arms & a vesa notebook tray would work too...

I better should have written "to reduce the comb filter effect" instead of "get rid of" in the title. So one last question: Do you think it`s worth it when swapping speakers is no option?

Thank you all!

 

[Hipper]

Just to show you what I was thinking:

The cardboard represents some sort of insulation.

I agree that some sort of wall mounted set up like the stand you pictured might work as it gets the speaker off your desk and allows variable geometry. Of course it's a question of weight and the further from the wall the speaker is the stronger the stand needs to be. You already seem to have lost some space on your desk by having the monitor so far forward. I suppose that can't be pushed back a bit.

 

[Vladimir Filevski]

Can you maybe give an example how the "Rabbit ears" would look like please in practice? And where is the difference between a floating arm and a stand if the speaker is ultimately in the identical position realtive to the desk? (besides that it`s obviously easier to move the speakers and consumes no place on the desk) Are there commercial products that implement your idea?

Maybe something like one of these monitor arms & a vesa notebook tray would work too...

Yes, that floating arm (as on your photo) will work too - extend it toward you ("very near" field monitor!) and high enough. This will reduce comb effect to a great extent.
Pictures and photos of the "rabbit ears" solution are on one of my old hard-disc drives, I have to look for it. There are no commercial products based on my idea. Basically it is floating arm attached at the monitor top corner - it demands very small satellite speaker, because of weight.

 

[Vladimir Filevski]

Just to show you what I was thinking:
The cardboard represents some sort of insulation.

Cardboard will introduce it's own reflection, better material is foam (1" thick) which will absorb frequencies above 1 kHz. Below 1 kHz it will be not effective.

 

[dasdoing]

you could make a little tower with that absorbive material and let it sit in front of the speaker, just below the center of the woofer. this will absorb the off-axis of the woofer at least

 

[dasdoing]

you could make a little tower with that absorbive material and let it sit in front of the speaker, just below the center of the woofer. this will absorb the off-axis of the woofer at least

here is what I mean. you have this:

now instead of putting absorbtion on the surface where the reflection ocours you can put it in the early path (so the table will continue to be free):

 

[Vladimir Filevski]

Absorbent block can not sit on the desk because it will occupy too much desk space. See photo at post #9 - bottom of the absorbent block should be at the cardboard plane (at bottom of the loudspeaker enclosure). This solution will work OK only if the absorbing block is protruding far enough in front of speaker toward you. If you put mirror on the desk, you will see that most of the reflection is coming just in front of you - where your arms are trying to type at keyboard. Because the absorbent block can be 2" thick in this case, it will work toward lower frequencies, but it still must be large enough in the horizontal plane (longer wavelengths will simply bypass the small block).

 

[dasdoing]

Absorbent block can not sit on the desk because it will occupy too much desk space. See photo at post #9 - bottom of the absorbent block should be at the cardboard plane (at bottom of the loudspeaker enclosure). This solution will work OK only if the absorbing block is protruding far enough in front of speaker toward you. If you put mirror on the desk, you will see that most of the reflection is coming just in front of you - where your arms are trying to type at keyboard. Because the absorbent block can be 2" thick in this case, it will work toward lower frequencies, but it still must be large enough in the horizontal plane (longer wavelengths will simply bypass the small block).

doesn't need to take all space. I would need a picture from the side to see how much material would be necessary, but if this is not blocking the tweeter path allready, the woofer path is blocked

 

[Vladimir Filevski]

Put mirror flat on the desk to help you - to work properly, the absorbent block must block (pun intended) the reflected images of the tweeter and woofer.

 

[Hipper]

Cardboard will introduce it's own reflection, better material is foam (1" thick) which will absorb frequencies above 1 kHz. Below 1 kHz it will be not effective.

I just used the cardboard to show how I would use insulation in Wegi76's situation. Of course I do not think cardboard would act as an absorber!

 

[aranido]

Hello, I'm not answering the question, but I realized that I do have the same setup (ADAM A7X over IsoAcoustics) and I do have space in my desk for absorption. My question is, Any recommendation for commercial "desk" panels?

 

[Wegi76]

Thanks guys. Guess I`ll improvise a bit with the "high-up/facing down" placement idea before buying something expensive and see if that helps to reduce the effect at all or if it makes it even worse (optics will suffer for sure). Really don`t have the space to implement the absorbtion materials ideas. Sorry.

Actually, when listening to music I`m not unhappy. And luckily sine sweeps are not my prefered taste of music