Baffle Diffraction

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Here is AudioeXpress article from Ralph, more text about felt treatment
Diffraction Doesn’t Have to Be a Problem | audioXpress

He seems to have something like the worst scenario as "no treatment", so the difference dosn't come from felt alone, but perhaps even more from different edge distance distritbution - so at least I take this single article with a grain of salt...

https%3A%2F%2Fwww.audioxpress.com%2Fassets%2Fupload%2Fimages%2FDiffractionPhoto5.jpg
https%3A%2F%2Fwww.audioxpress.com%2Fassets%2Fupload%2Fimages%2FDiffractionPhoto6.jpg
 
Last edited:
Who said diffraction only happens at one frequency? Actually I don't know what the hell you're talking about. Are you arguing with me or something in your own head? By the way, the guy in your sig. looks like a freak.

Show us how to avoid diffraction with a steep crossover filter... ;)

I'm sure Diamond Dave would take that as a compliment....
Here's that freak in action:
YouTube

Just look at the last few seconds of that clip, I'm sure it will look familiar...
 
Last edited:
Felt or sound absorption is going to work best with smaller wavelengths but it will help soak up extra sound energy before the cone beams. Break up involves a shift in power response that is partly why its so annoying. But also we are designed to talk to people and pick up non verbal stuff and resonances. So having your face in a speaker for hours will likely be fatiguing.
 
Last edited:
Is there a theoretical optimal shape for a roundover? Is a quarter circle better than, say, an elliptical profile that starts shallow and gradually steepens towards the side wall? I think the elliptical curve might at least look nice.

KEF gave the LS50 a gently curved baffle that terminates in a fairly sharp edge. Does that provide the maximum benefit for short wavelengths, then give up trying when the WL's get too long for any practical roundover to help?

bf-ls50c-item-background.jpg
 
Show us how to avoid diffraction with a steep crossover filter... ;)

It has nothing to do with "show". It's just a math problem. If diffraction happens at 3.5KHz, put a steep filter at around 1.7KHz on the woofer, whatever happens at 3.5KHz … gone … nada. Sure the tweeter still may suffer a bit, but at least you eliminate half the problem. Tweeter will be happy singing on its own without the woofer.
 
KEF gave the LS50 a gently curved baffle that terminates in a fairly sharp edge. Does that provide the maximum benefit for short wavelengths, then give up trying when the WL's get too long for any practical roundover to help?
What do you think ? Harry F Olson showed us that spherical shapes are better for dispersion and the LS50 is attempting at softening the reflection buy creating a more gentle roll-off - see measurement

KEF LS50 Anniversary Model loudspeaker Measurements | Stereophile.com
 
Last edited:
It has nothing to do with "show". It's just a math problem. If diffraction happens at 3.5KHz, put a steep filter at around 1.7KHz on the woofer, whatever happens at 3.5KHz … gone … nada. Sure the tweeter still may suffer a bit, but at least you eliminate half the problem. Tweeter will be happy singing on its own without the woofer.

Which shows me that you lack to give an example. It doesn't "happen" at 3.5 KHz, it happens over a frequency range. You'd be better off shaping the cabinet to avoid it so the tweeter won't have to suffer either.

Diffraction examples:
olson-baffleshape-fr.gif


Even though these are simple examples drawn to show trends, nothing more, they do show we are not talking about a problem to fix with crossovers. Do you know what diffraction does for perception? Probably even more so on the tweeter you mention as "may suffer a bit"?
First stop building rectangular boxes. You can't solve this with crossovers, unless you'd want to listen to that woofer low passed steeply at 1.7 KHz and forget about the tweeter (which will most likely have the bigger diffraction problem anyway!).
 
Which shows me that you lack to give an example. It doesn't "happen" at 3.5 KHz, it happens over a frequency range. You'd be better off shaping the cabinet to avoid it so the tweeter won't have to suffer either.

Again, nobody said at one frequency. I used 3.5KHz as an example. I assume most people here know enough to understand diffraction does not happen at one frequency. If you keep going on and on about obvious stuffs, good luck.

You keep posting stuffs that are rudimentary and academic that not helping anyone.
 
You are the one claiming one could solve the problem by crossing below diffraction. But I'd assume one would want some sound above that crossover too. So how do you solve the diffraction from the tweeter above it? That's all I'm pointing out. Crossovers aren't a solution to a diffraction problem.
Nor will diffraction be the same if one moves a little off axis. So solve it as it can be solved, by avoiding it.
 
Last edited:
You are the one claiming one could solve the problem by crossing below diffraction. But I'd assume one would want some sound above that crossover too. So how do you solve the diffraction from the tweeter above it? That's all I'm pointing out. Crossovers aren't a solution to a diffraction problem.
Nor will diffraction be the same if one moves a little off axis. So solve it as it can be solved, by avoiding it.

Whatever ...
 
Is there a theoretical optimal shape for a roundover? Is a quarter circle better than, say, an elliptical profile that starts shallow and gradually steepens towards the side wall? I think the elliptical curve might at least look nice.

KEF gave the LS50 a gently curved baffle that terminates in a fairly sharp edge. Does that provide the maximum benefit for short wavelengths, then give up trying when the WL's get too long for any practical roundover to help?

These are good questions and I would suggest that anyone who thinks that they know the answer would likely be mistaken. The rule in diffraction is how large is the surface curvature relative to the wavelength. Hence this, of course, means that all frequencies see the edge in different ways. I have thought about this in every way that you can imagine and they all seem to be complex enough that it is hard to guess at the answer and I have not seen a study of such, so there is no data that I am aware of. Hence, I just stuck with a circle.
 
As put forward by F Toole.

Sound Production vs Sound Reproduction.
Sound Production: The performance is the objective. It is ART happening in real time.
No two performances may be exactly the same.

Sound Reproduction: Reconstructing a captured performance for the benefit of the listeners,
whenever and wherever they choose to press "play". The goal is to do so with minimal change,
thereby preserving the "ART".
 
Have you considered how diffraction relates to open baffle speakers?

It could be quite low if the two path lengths are identical. But slight variations could be a real problem. For two point sources there is no edge diffraction if they are placed symmetrically (and exactly out of phase.) But any variance will show up as diffraction. I've never really studied the problem as the dipole inefficiency at LFs is just not acceptable to me and the HFs are handled with a waveguide so the baffle isn't a huge issue.
 
Last edited:
As put forward by F Toole.

Sound Production vs Sound Reproduction.
Sound Production: The performance is the objective. It is ART happening in real time.
No two performances may be exactly the same.

Sound Reproduction: Reconstructing a captured performance for the benefit of the listeners,
whenever and wherever they choose to press "play". The goal is to do so with minimal change,
thereby preserving the "ART".

Well said and oh so true!

My son is working as an intern at Harman this summer.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.