Diffraction can be audible. No doubt about it. In most cases I'm sure what you hear is just the effect of diffraction on amplitude. In most practical cases the diffracted sound arrives at the ears so incredibly soon after the direct sound, that it just can't be perceived separately. I'm not saying that amplitude is all there is with respect to diffraction, but I just haven't yet seen any strong evidence that suggests otherwise.
As loudspeaker designers it is our task to ensure a smooth response that doesn't change considerably with changes in mic/listening position. Diffraction can mess that up.
What can you do?
- Waveguides can direct the soundwave away from sharp edges
- A large radius can smooth the diffraction
- A source almost the size of the baffle will smooth diffraction
- Sound radiated from a driver on a baffle that is much smaller than the wavelengths, will just wrap around the baffle and the loss is easily compensated for in the crossover (baffle-step compensation)
Below you see a photo from during the construction of my current speakers. It's a 3-way with a 10" woofer (not shown).
The second image shows the measurements at 0 to 90 degrees horizontal (gated, indoors).
As loudspeaker designers it is our task to ensure a smooth response that doesn't change considerably with changes in mic/listening position. Diffraction can mess that up.
What can you do?
- Waveguides can direct the soundwave away from sharp edges
- A large radius can smooth the diffraction
- A source almost the size of the baffle will smooth diffraction
- Sound radiated from a driver on a baffle that is much smaller than the wavelengths, will just wrap around the baffle and the loss is easily compensated for in the crossover (baffle-step compensation)
Below you see a photo from during the construction of my current speakers. It's a 3-way with a 10" woofer (not shown).
The second image shows the measurements at 0 to 90 degrees horizontal (gated, indoors).
Keyser, when you sweep the mic across the listening window, the distance from the source and baffle edge will change causing the reflected sound to arrive at different times. This means, on a listening window curve, you will likely not see the effect of diffraction. If you've read Toole, you know what constitutes a resonance. He also talks about adapting to this type of coloration.
I'm an engineer - language is very important to me.
When I am working on a project, if I get the terminology wrong, or use words like "all" when I really mean "some", things break. Things literally stop working.
So when I say that ""I see a lot of Dunlavys and B&Ws in studios" I do not mean that "B&Ws are the most popular studio speakers."
I never said that "B&Ws are the most popular studio speakers" and I chose my words very carefully.
I hate multiquoting too. Sorry.
Right. It is directly dependent on baffle size and the wavelength being emitted. When the baffle becomes too large for the wavelength, it will start diffracting. Whether it is audible or not is another issue.
2 inches corresponds to 6810 Hz. Quite high.
Agreed that mouth radius on a horn is useful. There it really does create a reflection that's audible, especially on round horns.
I'm not providing anecdotal data. All this is well established science. We continue to ignore the factors that do make a difference and latch on to something that may or may not have an influence just so that we can explain something that we may or may not have heard. All this started with PB claiming that the reason he feels more relaxed with the Summas compared to the Vandersteens is because Summas tackle diffraction. All I'm saying is that there are countless other differences in the two speakers that can contribute to his perception. Not to mention how these two very different speakers interact with this room. How can you ever claim you know this is due to diffraction?
Right. It is directly dependent on baffle size and the wavelength being emitted. When the baffle becomes too large for the wavelength, it will start diffracting. Whether it is audible or not is another issue.
I highly doubt you will hear the combing, especially at high frequencies. The peaks and dips are so close together. Besides, it is a type of coloration that once you get used to, you learn to ignore. Lastly, the power response will not show it as a resonance. Another reason the ear may not perceive it.
That's fine. You can choose to disagree with it. You can put together your hypothesis, like PB has done. But you cannot conclude that this must be it. There are several other factors that are known to be much more audible that should be investigated first.
2 inches corresponds to 6810 Hz. Quite high.
Agreed that mouth radius on a horn is useful. There it really does create a reflection that's audible, especially on round horns.
Again, the audibility is determined by the listening window response. If that is smooth, and the combs are narrow, just don't see how we could hear it.
Could be. Should be investigated further.
I'm not providing anecdotal data. All this is well established science. We continue to ignore the factors that do make a difference and latch on to something that may or may not have an influence just so that we can explain something that we may or may not have heard. All this started with PB claiming that the reason he feels more relaxed with the Summas compared to the Vandersteens is because Summas tackle diffraction. All I'm saying is that there are countless other differences in the two speakers that can contribute to his perception. Not to mention how these two very different speakers interact with this room. How can you ever claim you know this is due to diffraction?
Ugh I wish I'd known the gas mileage was that good.
I had two Accords in a row, and then I felt that the V6 Accord was a bit sluggish. Not slow by any means, but just felt *heavy.*
So I considered the Civic SI, but passed on it because the reported MPG was mediocre. (25/31 mpg) Also looked at the BR-Z, Elantra, BMW 3 series, Mazda3, and VW Golf.
Ended up getting a Mazda6 and it's kinda a dud. It's got a great transmission and the mileage is amazing (35-40 on the highway) but the interior is kinda cheap and it just feels kinda low budget compared to the Accord.
Here's your post in full but without the pictures. You also quoted Earl's post that talked about his study being the only real data point on diffraction audibility (reproduced below). Can you see why it may have made me think you are claiming B&Ws are popular in studios because of their diffraction reduction strategies?
Anyway, let's not get caught up in semantics. Do you agree that there may be reasons other than diffraction for the popularity (real or not) of a speaker?
Anyway, let's not get caught up in semantics. Do you agree that there may be reasons other than diffraction for the popularity (real or not) of a speaker?
The concept of using the 1/4 wl vs. radius on a non-circular baffle is badly flawed. Consider the pathlength when the wavefront from the waveguide hits the baffle on a speaker like the summa 10" below the forward axis (Z axis) of the waveguide- the pathlength of transition across the roundover is much increased, and is additionally a longer pathlength across the flat baffle, lowering both SPL and applying 2pi directivity across a larger bandwidth. Exactly how much depends upon the particulars, but it's not as simple as "what's the radius".
Our study showed that delay of the diffraction was a factor as well, although absolute SPL was the greatest controlling variable. Amplitude was second.
Correct, Thanks.
It's curious how people often make mistakes to their advantage in a discussion. This usually means to me to "just pass on it because the same person might just make up something else if you try and continue.
The 1/4 lambda radius is a good rule of thumb for where the radius begins to become effective. It is the midpoint frequency for the transition from no effect to almost no diffraction that does happen over some bandwidth.
See the driver as an infinite number of point sources. Each of them has a different distance to the edge, and a different distance to the edge in each direction (except for the one single point in the center). The result is that the effect of diffraction on the amplitude response is smoothened.
EDIT: Good example of a commercial offering that makes use of this principle:
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
..he can't with any real degree of confidence, it's why I mentioned the minor/easy test of simply putting a small amount of absorption/material in varying diffractive positions to test the assertion.
It doesn't even have to be a "blind" condition for a forum response.
IF done (with a lossy enough material that doesn't limit the dispersion pattern to any significant extent), and the result is a confirmation then I'll accept that.
..he can't with any real degree of confidence, it's why I mentioned the minor/easy test of simply putting a small amount of absorption/material in varying diffractive positions to test the assertion.
It doesn't even have to be a "blind" condition for a forum response.
IF done (with a lossy enough material that doesn't limit the dispersion pattern to any significant extent), and the result is a confirmation then I'll accept that.
Here is one way to determine if diffraction treatment is effective:
1) take an existing speaker
2) and add a roundover
Then observe the difference.
Would it be better if we had measurements in both the frequency and time domain? Of course it would.
But if you believe that diffraction treatment doesn't work, read the Diyma thread. There are literally dozens of people all over the world trying this, and reporting back their findings. It's been universally positive.
Now I concede that we could apply the same yardstick to silliness like Enabl treatment, and come up with the same positive results.
My argument against that is that there *has* been a few people on the Diyma thread who have demonstrated that the diffraction treatment improved the performance and proved it with measurements.
Go check it out! I know that most of this forum doesn't frequent the car audio forums, but there's a lot of smart creative folks on Diyma.
This misses the point.
I *do* believe that diffraction treatment can make a difference - so you are "preaching to the choir" here.
What I do NOT automatically believe is that the fatigue you have mentioned is the result of diffraction with the Vandersteen 2C.
It *might* be the reason, or *part* of the reason.
-BUT there are any number of other possibilities as well when compared with your Summa's (..which btw, have plenty of diffraction - if at a lower freq.)