Radius must be at least 3/4" to reduce edge diffraction

0 Members and 1 Guest are viewing this topic. Read 7593 times.

audiojerry

  • Full Member
  • Posts: 1355
Danny, you may not subscribe to audioXpress, (although I've seen a new ad for the Criterion in it), but I found a very interesting article in it on edge diffraction in the latest Feb. issue, by Jim Moriyasu, who did a pretty extensive study on cabinet edge diffraction and how to reduce it. As far as I know, it may be the most thorough study anyone has done. One important finding was that any radius less than 3/4" is a waste of time.  He also made several recommendations for ideal and compromise solutions for cabinet edge radii.

Have you seen the article or are you aware of this study? His data seemed pretty credible. If you like, I can fax or mail you a copy.

The website is www.audioxpress.com. It's a magazine for audio do-it-yourself'ers.

Danny Richie

edges
« Reply #1 on: 7 Feb 2003, 09:51 pm »
I do not have the Feb. issue, but I am aware of the effects of a radius on the edge.

I would not say that a radius of less than 3/4" is a waste of time.

Obviously this is an esthetics issue too.

The smaller the radius the less the effectiveness for sure.

Even a 1/2 inch radius would be effective from about 18kHz and up though (16kHz and up maybe). Not real effective though huh?

I think even more important than adding a big radius would be to just have a a narrow baffle.

The criterion for instance is only 7.5" wide so edge diffraction is about as small as you can make it.

audiojerry

  • Full Member
  • Posts: 1355
Radius must be at least 3/4" to reduce edge diffraction
« Reply #2 on: 7 Feb 2003, 10:02 pm »
With the Criterion, assuming that a 1/2" radius is merely for aesthetics, do you think squared off corners would have a negative effect. As a matter of taste, I really like the look of precise corners ala ProAc. I equate it to precision cabinet craftsmanship.

Relative to aesthetics, would moving the tweeter to one corner have any negative effects?

doug s.

  • Full Member
  • Posts: 6572
  • makin' music
Radius must be at least 3/4" to reduce edge diffraction
« Reply #3 on: 8 Feb 2003, 04:38 am »
jerry, note that the proacs' have their tweeter offset, & the distance to the cabinet's sides is also different than the distance to the cabinet's top.  this means no one single frequency's diffraction is reinforced, & the ear is not wery sensitive to the diffraction, when it's averaged out in this manner.

danny, have ya ever tried offsetting yer tweets in this manner?

doug s.

Danny Richie

tweeter offset
« Reply #4 on: 9 Feb 2003, 01:23 am »
Quote
danny, have ya ever tried offsetting yer tweets in this manner?


Been there done that many times...

It works great with dome tweeters with wide baffles.

The negatives are a twisted lobing pattern and different off axis responses.

It doesn't work as well with ribbons and doing this with the Criterion would have also increased the center to center spacing which is critical in this design due to the relatively high crossover point.

audiojerry

  • Full Member
  • Posts: 1355
Radius must be at least 3/4" to reduce edge diffraction
« Reply #5 on: 9 Feb 2003, 04:41 am »
Danny, you always have a knowleadgeable answer, but what about the squared off corners?

JohnR

Radius must be at least 3/4" to reduce edge diffraction
« Reply #6 on: 10 Feb 2003, 07:13 am »
Doesn't the effect of roundoffs vary with frequency?

Did the article elaborate more than "waste of time"?

JohnR

audiojerry

  • Full Member
  • Posts: 1355
Radius must be at least 3/4" to reduce edge diffraction
« Reply #7 on: 10 Feb 2003, 04:52 pm »
Did the article elaborate more than "waste of time"?
Yeah, way more! That was just my encapsulation  :wink:

JohnR

Radius must be at least 3/4" to reduce edge diffraction
« Reply #8 on: 10 Feb 2003, 07:32 pm »
Thanks Jerry, very helpful  :roll:

theranman

  • Jr. Member
  • Posts: 20
Radius must be at least 3/4" to reduce edge diffraction
« Reply #9 on: 20 Feb 2003, 01:02 am »
Interesting info.

Everything I've ever read suggested that a radius of less than about 1.5" does squat to ameliorate the audibility of edge defraction. Does the article state whether the 3/4" radius has an "audible" difference or merely a measurable one?
 Aesthetics is a whole nuther story in MY book...even a 1/2" radius looks waaaaay better on speaker edges.

Randy

Brian Bunge

Radius must be at least 3/4" to reduce edge diffraction
« Reply #10 on: 20 Feb 2003, 06:40 am »
Actually, the author states the best results came from using a 22 degree bevel and then a 45 degree bevel together.  It was mainly stated that a 3/4" radius does very little to help reduce diffraction, stating that 3/4" is equal to an 18KHz wavelength and will do nothing to help diffraction artifacts between 1KHz-18KHz.

theranman

  • Jr. Member
  • Posts: 20
Radius must be at least 3/4" to reduce edge diffraction
« Reply #11 on: 20 Feb 2003, 04:20 pm »
As for the combined(?) bevels, does the article suggest a minimum width for each. Surely a 1/8" 22 degree bevel combined with a 1//8" 45 degree bevel won't do anything, right? Or am I missing something?

audiojerry

  • Full Member
  • Posts: 1355
Radius must be at least 3/4" to reduce edge diffraction
« Reply #12 on: 20 Feb 2003, 04:29 pm »
Quote from: Brian Bunge
Actually, the author states the best results came from using a 22 degree bevel and then a 45 degree bevel together.  It was mainly stated that a 3/4" radius does very little to help reduce diffraction, stating that 3/4" is equal to an 18KHz wavelength and will do nothing to help diffraction artifacts between 1KHz-18KHz.


Yes, that was the crux of it, and I give the author credit for doing some really in-depth research on the subject. It would be nice to see more studies like this. It would be difficult to prove or disprove, but based on the amount of effort he put into it, I will choose to accept it as good advice.

This has a definite impact on the cabinet style of small monitors, since achieving an optimal radius would require a much larger cabinet. I'm glad this is true, because aesthetically I prefer hard edges to rounded. To me a cabinet with very sharp edges requires greater cabinetry skill, and represents a a more furniture-like quality of appearance.

What do you think, Brian?

Brian Bunge

Radius must be at least 3/4" to reduce edge diffraction
« Reply #13 on: 20 Feb 2003, 04:34 pm »
Randy,

They were pretty large.  One was about 2" and the other 1" or something.  I'd have to re-check the article.

Jerry,

Square boxes are boring to me, but that's just a personal opinion.  We start out with very accurate corners and then round over afterwards.  I can tell you that it's more time consuming to veneer a single side at the time, so in some respects it does take a bit more care to try to hide seams.

doug s.

  • Full Member
  • Posts: 6572
  • makin' music
Radius must be at least 3/4" to reduce edge diffraction
« Reply #14 on: 20 Feb 2003, 04:49 pm »
hi all,

well, the speaker that flat out disappeared better than anything else i'd ever heard - and by a wide margin - was the diapason adamantes.  also a fabulous-looking bookshelf speaker, imo.  it's too bad i've never been able to get a pair at a give-away price, to try in my own listening room...   :?

http://www.diapason-italia.com/eng/intro.htm

hey danny, tink ya can offer the criterion w/a cabinet kit similar to the adamantes', for, say ~$1500?   :)

regards,

doug s.

KevinW

  • Full Member
  • Posts: 322
Radius must be at least 3/4" to reduce edge diffraction
« Reply #15 on: 25 Feb 2003, 06:28 pm »
Quote from: audiojerry

This has a definite impact on the cabinet style of small monitors, since achieving an optimal radius would require a much larger cabinet. I'm glad this is true, because aesthetically I prefer hard edges to rounded. To me a cabinet with very sharp edges requires greater cabinetry skill, and represents a a more furniture-like quality of appearance.


Hey Jerry,
Since I am starting the path of learning about speaker design, with the goal of producing speakers for sale, I have spent a fair bit of time learning about edge diffraction and baffle step response.  Here's what I've learned so far...

Edge diffraction and baffle step are very similar.  Essentially they are both consequences of the following physical phenomemon:
1.  Bass waves are non-directional in closed rooms, because the wavelength is very large compared to the size of the room, and compared to the size of the speaker.  A 20hz wave has a wavelength of 56 feet, and a 100Hz wave is 11 feet.
2.  High frequencies are very directional, and very short compared to the size of the speaker. 5kHz is about 2 inches long.
3.  The above differences mean that high frequencies are radiated in only 180 degrees, while bass waves radiate in 360 degrees.  This means that highs are "reinforced" by the presence of a wall behind the driver.  This "wall" is the cabinet baffle.

The reinforcement is called "baffle step response", and is a real headache for speaker designers.  It essentially causes a 6dB rise of the highs over the lows at some point in the audio spectrum, which needs to be compensated.

But where does the edge shape come into this discussion???  :)  Well, the highs are reinforced by the baffle as if it were an infinite plane, except that eventually those frequencies will encounter the edge, where they diffract.  Here is where the edge shape is critical.  A sharp edge will cause all kinds of weird diffractions, which will mess with the sound in odd ways by creating many jagged harmonics.  A smooth shape will cause many less diffractions, and in fact a perfect sphere will have a very smooth transition into the "stepped" frequencies.  Our ears will tend to ignore smooth transitions, while sharp ones will be quite apparent.

Regarding cabinetry skill, it is the simplest thing in the world to make sharp edges.  All you need to do is cut sheets of wood with a table saw, join the edges, and sand the joins smooth.  Making curves is much harder, especially large, complex radius curves.  That is where skill comes into play.  The curves also have a much better effect on the sound quality, because there are not a lot of "sharp" edge diffraction and baffle steps mixed into the music.

So from a designer's perspective, curved cabinets equate to more skill, because they are better for sound quality.

audiojerry

  • Full Member
  • Posts: 1355
Radius must be at least 3/4" to reduce edge diffraction
« Reply #16 on: 28 Feb 2003, 02:15 am »
Kevin, I've been on vacation so I'm just now reading this. Sharp edges with veneer I believe have to be difficult.

I think the ProAc 1SC, which has sharp edges and looks great to my eyes, sounds great also, better than nearly all mini monitors of that size that I have heard, so I guess it sounds great despite the dharp edge, or the designer brilliantly found a way to compensate for it.

KevinW

  • Full Member
  • Posts: 322
Radius must be at least 3/4" to reduce edge diffraction
« Reply #17 on: 28 Feb 2003, 06:06 pm »
Quote from: audiojerry
Kevin, I've been on vacation so I'm just now reading this. Sharp edges with veneer I believe have to be difficult.

I think the ProAc 1SC, which has sharp edges and looks great to my eyes, sounds great also, better than nearly all mini monitors of that size that I have heard, so I guess it sounds great despite the dharp edge, or the designer brilliantly found a way to compensate for it.


Perhaps I was a little too strong in my previous post... you can tell when I get excited by the way I write... I didn't mean to imply that sharp edges mean that a speaker is junk.  (I don't think I did, actually)  And there are ways to compensate for sharp edges using networks or the inherent characteristics of a driver.  But based on my understanding of the diffraction properties, a sharp edge on the front baffle is an inherent liability in the design of a speaker.  The effect is almost certainly not the most crucial aspect of a speaker design, but it may have some significant impact.  How significant?? I don't know, but I'm doing some empirical testing of the effect, and will report back in a couple of months when I have some hard data.

doug s.

  • Full Member
  • Posts: 6572
  • makin' music
Radius must be at least 3/4" to reduce edge diffraction
« Reply #18 on: 28 Feb 2003, 06:57 pm »
as i mentioned earlier, having the tweeter offset from the edges (a la proac) counteracts the diffraction issues.  but, as danny richie mentioned, this introduces another set of issues.  which is why speaker building is as much of an art as a science, it seems to me...  i'd still like danny to design a criterion-based speaker w/a diapason adamantes styled cabinet, to retail in kit form, for ~$1500...  ;~)

doug s.

massappeal85

  • Jr. Member
  • Posts: 40
Radius must be at least 3/4" to reduce edge diffraction
« Reply #19 on: 25 Mar 2003, 02:34 am »
Question: How does any of this about edge diffraction apply to horns?