How Do You Deal With Baffle Edge Diffraction?

I recently purchased new bookshelf speakers and set up my system in a new room, and after implementing my usual tweaks (power cables, outlets, sound treatments, etc.) I found that the highs were a tad irritating at higher volumes, which caused listener fatigue. My setup was mostly the same as before, except for the speakers. “Perhaps these speakers are just brighter than the last ones,” I thought to myself. But still, bright or not, a decent speaker should not be causing listener fatigue, especially with decent gear upstream. 

I remembered that on my last setup I had made custom speaker grilles out of 1” thick wool with cutouts for the speaker drivers and wrapped it in an acoustically-transparent cloth. The idea behind these is that the wool absorbs the higher frequencies on the face and edges of the baffle, which knocks down baffle edge diffraction. I dug those out and affixed them to my new speakers. It solved most of the problem! They don’t perfectly align with the old speaker driver locations so I need to remake them, but for now I am a happy camper.

I’ve read that felt around a tweeter works very well for this purpose but for whatever reason (probably looks) very few manufacturers implement this. Is anyone else using custom built grilles to help knock down baffle edge diffraction? How do you all deal with this issue? Are you all aware of this issue? If you are currently less than happy with your speakers’ higher frequency performance or struggle with listener fatigue, I highly recommend you try this. In my case, I had all of the materials already, so besides time it was a free upgrade and one of the better tweaks I’ve implemented. 
wow, great idea.. i think i am going to try a wrap of some material around me tweeters to see if there is a benefit
Try it out and let me know what you think. Two challenges are affixing it to the speaker so it doesn’t cause damage (tape not leaving residue, etc) and making it look presentable. :) I’m currently supporting mine with some dental floss but I will work on a better solution. 
I think that it could be a bad thing too, depending on how the speaker was designed.  I owned a pair of TDL Monitors years ago, a large 4 way floorstander, and I experimented with doing that around and between drivers, and to my surprise the blend of the drivers then became disjointed, I ended up doing it lightly just around the tweeter with some benefit.
John Bau did the same on his Spica loudspeakers. Richard Vandersteen puts a thin felt sheet around the drivers in his Model 2 and 3.
Post removed 
Unless you are using a waveguide,felt is good.

Madisound offers felt rounds for just such a purpose.

You should also experiment however with the area between your speakers. I have often found that harshness in the highs comes from not enough damping there. Try throwing a blanket there and see what happens. :)

Erik, do you mean the area directly between the speakers or the back wall behind the speakers?
I saved up for a pair of used Avalon Acoustics whose baffle design seems very popular among speaker designers. 
We were using this technique in the late 70's.  It worked very well then and should do the same now.  
John Dunlavy used this alot in his speakers. You might take a look at some of his speakers for some idea's. 
Hey gang,

Just FYI, the "baffle step" effect which @Elizabeth linked to is different than I think the OP meant to ask. Here

Both are real.

Baffle step has to do with how the baffle affects low frequencies of a speaker without rear wall reinforcement. It affects how quasi-anechoic measurements are done, and is related to "baffle step compensation" which is done in the crossover. Basically, speaker designers have to lower the woofer's efficiency to extend the bass low enough.

Here is an example of a speaker deliberately lacking in BSC because it is designed for rear wall placement:

See figure 6, and note the steep ramp up to 100 Hz. JA's failure to place the speaker as recommended, and failure to understand what he was looking at makes this review otherwise severely inadequate.

Edge diffraction, which I believe the OP was talking about is a different thing. It is prevalent in the mid to high frequencies. As the hemispherical wavefront goes from driver to the edge of the baffle, the corners can cause the signal to re-emit. Instead of hearing just a tweeter:


You hear the tweeter plus the edges:


with the edge emissions being delayed.It is similar to the idea of first reflection points in a room.  There are many approaches to this issue:

  • Felt
  • Rounded corners
  • Beveled corners
  • Minimum width baffle
  • Maximum width baffle
  • Wave guides

You’re correct Erik. I am referring to baffle edge diffraction. Both that and baffle step are important concepts. Good information all around. 

It’s my understanding that edge diffraction will cause the higher frequencies to “re-emit” from multiple points around the speaker which causes a number of issues. Perhaps worst of all, this “re-emittance” happens out-of-phase with the original sound wave created at the tweeter which really mucks up the sound. I believe this contributes to listener fatigue because the sound doesn’t quite make sense to the brain. (That’s my personal theory. Not sure if tests have been done to verify it.)
@elizabeth, it may be of interest to you (if you don't already know) that Danny Richie is largely responsible (along with Brian Ding) for the design of the Rythmik/GR Research OB/Dipole Sub that mates so successfully with planar loudspeakers like your MG20.7's. He's a very knowledgeable and talented speaker, sub, and crossover designer.
Part of it is the tweeter.  It is brighter or the crossover has to be tuned a hair brighter because of how it interacts with the cabinet.  

Felt is a nice solution.  Try adhering with something temporary first to see if it helps.