Aren't those all reflection issues? It would make sense to not mix reflection with diffraction - although both have many things in common.
😱 a simple box for a 12" woofer would look like a big dipole baffle
but basicly you could just ad material to the edges
no, all of what I mentioned are very often sharp edges
if rounded you are ofcourse right about it
and one reason why I did this
Attachments
Yes, I agree....but I think this would only count for tweeters where diffraction or reflection is pushed up very high in frequency, and the time smear is small, also very high in frequency. And we are talking about the surround for the tweeter, it is so small that it will only apply to frequencies well beyond our hearing. The screws and edge of the faceplate is flush with the baffle.
If we are talking full range drivers, well high frequencies are beaming and don't really impact the surround too much.
The problem is when the frequencies of the driver has 180 degree dispersion at full power response, then hits the edge of the baffle. That's where you will get a cue that the sound is coming from a box. It may not be a severe ripple but, it may be enough to have that effect......psychoacoustically.
As a rough general principle, I work from the input end of my system in assigning priorities. On that basis, a well behaved loudspeaker enclosure comes ahead of room size, treatments, speaker stands etc. When diffraction effects (speaker and nearby surfaces) are well controlled, I can then deal much more effectively with room acoustics. In a sense a well behaved enclosure assists diffraction and therefore throws more energy out to the sides where the room acoustics and treatments have to deal with it.
I have to disagree from my experience in my living room with engaged, but not "critical" listening. My measurements confirm my listening, in that I can measure the additional peaks and hear the bright and edgy sound with a sharp edge vs an otherwise identical speaker with a large radius in the same position. The initial peak just above the baffle step point is by far the most dominating.
I usually use 3/4 MDF and that is sufficient for a 3/4 inch radius. I will be adding a backing strip when I try a one inch.
IMHO, I can give a $100 speaker the sharp edge. In a $2000 speaker, no. No excuse. In any speaker irregardless of price,there is no excuse for not doing some modeling and prototyping with edge/position simulation. So often I see offset tweeters, but the more critical midrange is symmetrical where the hump at step could be reduced by more than a dB.
Again, don't take my word for it. Don't even take Olson. Just do the test. None of us hear the same thing nor do the clues the speaker produces all tell us the same thing. Psychoacoustics; fascinating subject.
Egg speaker. Several problems. Expensive unless they can be molded in vast quantity. They would not fit the decor or space of as many rooms. Very worth a try for DIY. Compare identical system in alternative cabinets.
I would describe the radius as distributing the incidence angle of diffraction, neither hindering or enhancing diffraction itself.
One thing that appears to be overlooked in the issue about edge treatment size (roundover/chamfer) is that the dimension itself is but one part of the equation. Diffraction is better considered as a function of wavelength vs. total distance traversed (i.e. time-delay) from a driver point to an edge point. Take a 1/2" roundover on a very wide baffle and you may see almost no impact at any frequency. Use that same roundover on a baffle barely wider than the tweeter and you see a massive impact (not kidding) vs. a square edge. But you can't adequately evaluate the impact by measuring on a single axis.
This is essentially part of what is happening when a tweeter is offset on a baffle. The baffle edge treatment closer to the tweeter has more impact than the edges farther away. You can't focus on edge treatment dimension alone.
Many of you know my devotion to using felt on any baffle. I made many a test combining baffle edge radii/chamfer with felt. In almost all cases, surprising to me, I found that a square edge provided the best result.
Why you may ask? I attribute it exclusively to having more area for felt coverage. With the exception of what I think of as excessively large baffle radii, felt is optimal. Roundovers/chamfers reduce the distance through which felt can be applied, since the wave must pass through the felt to be effective. Placing felt on a chamfer did not provide much benefit if at all IME, it was better to have a square edge with more felt. It also lowers the frequency above which diffraction is no longer an issue, since diffraction is a time-delay vs. wavelength issue.
It does help to both offset a tweeter and use felt (or other edge treatment) primarily due to the increased randomness of distance, but it isn't all good. You have to examine a fairly wide arc, since particular combinations may exacerbate an angle that might just be the axis used if a system is placed with some off-axis angle as the direct response. This is largely why most offset designs are used with the tweeter offset edges (for mirrored pairs) closer to the listener, the variation is usually reduced whereas the variation usually increases in the off-axis on the longer distance side.
OK and fair enough comment on money.
I was speaking from my current impoverished perspective where I am building $100- speakers. Never really thought much about building up the edges tho, would that work well enough to give it a try? SWMBOs speakers are next in the pipeline and I do want them to be better than OK
My biggest router bit is a half inch and that's usually all I can use due to using only 18mm MDF
I was speaking from my current impoverished perspective where I am building $100- speakers. Never really thought much about building up the edges tho, would that work well enough to give it a try? SWMBOs speakers are next in the pipeline and I do want them to be better than OK
My biggest router bit is a half inch and that's usually all I can use due to using only 18mm MDF
I've used 1" roundover bits on 18mm.
If you want good cheap bits, get grizzly bits.
http://www.grizzly.com/products/category/161
1" roundover.
http://www.grizzly.com/products/1-r-Roundover-Bit-1-2-Shank/C1191
If you want good cheap bits, get grizzly bits.
http://www.grizzly.com/products/category/161
1" roundover.
http://www.grizzly.com/products/1-r-Roundover-Bit-1-2-Shank/C1191
guys, I hope you read this
It should be noted that a beveled edge is almost as good and can be made much larger in size with a lot less effort.
And agree with tin/bon- dealing with diffraction makes everything else easier to manage, since you're not dealing with different FR at different points in the room.
And agree with tin/bon- dealing with diffraction makes everything else easier to manage, since you're not dealing with different FR at different points in the room.
I would like to point out that just because something has a "smallish" effect on the frequency response does not mean that the audible effect is the comensurate. There is a lot of data to say that time delayed effects are far more audible than minimum phase ones and cabinet diffraction is a time delayed effect (as are HOMs). So one cannot simply look at the result in a simulation and write off its importance.
I use large radi on my enclosures. Is it a night and day effect?, probably not, but at the hi-end even little thing can help (well thats not really true, its a cost-tradeoff). I will say that these radi on the edges, waveguide and woofer hole make it impossible to laminate finishes and this IMO is why we don't see large radi in the marketplace. Lamination is the only high volume finish that is practical.
In many speakers there are far bigger problems than cabinet edge diffraction, so in that case it probably isn't worthwhile. But as you reduce all of the other problems, the edges slowely become the major ones. It also makes no sense to radi the cabinet edges if you are not going to minimize the diffraction from the objects nearby the speaker.
So clearly, in most cases radi on the edges is not a big thing, but in extreme situations I believe it is.
I use large radi on my enclosures. Is it a night and day effect?, probably not, but at the hi-end even little thing can help (well thats not really true, its a cost-tradeoff). I will say that these radi on the edges, waveguide and woofer hole make it impossible to laminate finishes and this IMO is why we don't see large radi in the marketplace. Lamination is the only high volume finish that is practical.
In many speakers there are far bigger problems than cabinet edge diffraction, so in that case it probably isn't worthwhile. But as you reduce all of the other problems, the edges slowely become the major ones. It also makes no sense to radi the cabinet edges if you are not going to minimize the diffraction from the objects nearby the speaker.
So clearly, in most cases radi on the edges is not a big thing, but in extreme situations I believe it is.
Yeah, as I said earlier....
It may not be dramatic, but psychoacoustically it may be noticeable.
Some things effect us more than others. Maybe we perceive the size of things by which band of frequencies are diffracted off of the object?
Maybe? I don't know....just speculating 🙂
It may not be dramatic, but psychoacoustically it may be noticeable.
Some things effect us more than others. Maybe we perceive the size of things by which band of frequencies are diffracted off of the object?
Maybe? I don't know....just speculating 🙂
Delicate spatial cues(purist recordings) are probably most effected by rounding or not rounding.
Thanks- that's very useful for demonstrating that even a small 2.4mm roundover has a significant positive effect.
The part of interest in the sheet posted by JIKO is the off-axis responses for each edge profile. One can see improvement in response with edges that have a larger radius. The profiles that show improvement do so by removing some of the sharp peaks off-axis ... these peaks, as Dr. Geddes stated, are caused by "echos" reflected back off the untreated edge.
- Status
- Not open for further replies.