(I'm back!) Is baffle step loss a poorly developed concept?

[OlderMongrel]

Hi folks,

I've been just a lurker for a while. I was posting some time ago as OldMongrel and lost my login. Apparently the old stuff cannot be recovered at this point, so I will just start over. This is mostly just a test post. I might be flooding the forum with stuff regarding recent thoughts and projects, so watch out!

Baffle step is something I have been wondering about for a while. The concept is probably sound, and maybe I just don't get it completely, but even so, maybe better documentation is needed. Many of you probably recall the roller-coaster response curve for a driver mounted in the end of a cylinder, shown in John L Murphy's well known article on the subject at...

http://trueaudio.com/st_diff1.htmhttp://trueaudio.com/st_diff1.htm



John shows lots of references, but it looks like it's all theory. I have not been able to find any measurements of something like this. It seems to me that this should be easy to confirm if you have an anechoic chamber, maybe not otherwise, but there ought to be something empirical. Does anyone know of any measurements that show an actual curve like this?

Another thought: Doesn't the first half waveform (180 deg) arrive intact, and the loss only kicks in after that as the diffraction begins to arrive out of phase with the second half of the directly radiated wave?

Is this out-of-phase diffraction, or an actual loss as the longer wavelengths "wrap back" around the baffle? Are these not two different things that are happening simultaneosly?

Thanks for any clarifications,

John

[dlr]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

Baffle step is something I have been wondering about for a while. The concept is probably sound, and maybe I just don't get it completely, but even so, maybe better documentation is needed. Many of you probably recall the roller-coaster response curve for a driver mounted in the end of a cylinder, shown in John L Murphy's well known article on the subject at...

http://trueaudio.com/st_diff1.htmhttp://trueaudio.com/st_diff1.htm



John shows lots of references, but it looks like it's all theory. I have not been able to find any measurements of something like this. It seems to me that this should be easy to confirm if you have an anechoic chamber, maybe not otherwise, but there ought to be something empirical. Does anyone know of any measurements that show an actual curve like this?

Another thought: Doesn't the first half waveform (180 deg) arrive intact, and the loss only kicks in after that as the diffraction begins to arrive out of phase with the second half of the directly radiated wave?

Is this out-of-phase diffraction, or an actual loss as the longer wavelengths "wrap back" around the baffle? Are these not two different things that are happening simultaneosly?

Thanks for any clarifications,

John

I may have responded to your inquiries under your old moniker, I don't recall. In any case, you'll not often find anyone with comparable measurements since much of it is academic due to there not being many cases of that example actually being built. There are a lot of other examples of real-world cases for more typical baffles. I've got a set of design pages for my 2-way that is hosted by John over at Zaph Audio. I included a fairly detailed examination of models vs. measurements for baffle step/diffraction as just one part of it.

The Chameleon

The page on the low end splicing of near-field and 1m responses demonstrates the good correlation and this is in a home environment. Measurements in an anechoic chamber would show even closer correlation, I'm sure.

Let me also suggest, if you have not yet seen it, this article at my site:

Understanding Cabinet Edge Diffraction

It is theoretical, yes, but it is just this theory that is used in software such as the Baffle Diffraction Simulator that has very good correlation to empirical examination.

The actual mechanism is not a purely half or full wave issue, it's one of pure time delay of the diffracted wave as it is formed from the distributed interaction with the baffle edge. The result is complex constructive/destructive inteference patterns across the bandwidth below the frequency where the direct wave is purely 2-pi.

dlr

[augerpro]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

I believe the example shown and all the others in the LDC were based on actual measurements by Olson? So those are not theoretical plots by some precursor to Edge or BDS, for example.

[OlderMongrel]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

Thanks for the input. Is compesation based on just axial early arrival? What effect does it have on power response? It seems like there are lots of tweaks for axial response which neglect the power effects.

John

[dlr]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

Thanks for the input. Is compesation based on just axial early arrival? What effect does it have on power response? It seems like there are lots of tweaks for axial response which neglect the power effects.

John

That's a mistake often made, though it occurs less often now in the DIY community since there's so much information available relating to on-axis vs. off-axis. You'll find a lot of past threads here at PE that include the topic of axial vs. power response.

dlr

[spasticteapot]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

I will very soon have access to the materials and tools required to produce some spherical speakers. Does anyone have any data on this type of speaker? And how large should the radius of the sphere be?

[cixelsid]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

Hi folks,

I've been just a lurker for a while. I was posting some time ago as OldMongrel and lost my login. Apparently the old stuff cannot be recovered at this point,

166 old posts all are available.

http://techtalk.parts-express.com/member.php?u=2934

[OlderMongrel]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

166 old posts all are available.

http://techtalk.parts-express.com/member.php?u=2934

I should have clarified: what I mean is that I no longer have the log in and could not get it.

[OlderMongrel]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

That's a mistake often made, though it occurs less often now in the DIY community since there's so much information available relating to on-axis vs. off-axis. You'll find a lot of past threads here at PE that include the topic of axial vs. power response.

dlr

Dave, what mistake are you referring to? Changing the axial response will change the power response, correct? I think I also had in mind that it will of course change the room response as well.

I've been reading a lot of posts here for a long time, and yes, many of them mention power response or axial response. This usually has to do with what might sound better for one or the other (or for in-room), but I don't recall seeing anything about dealing with these issues successfully all at once. Do you have any posts in mind that deal with some methodology for balancing the 3: axial, power, room?

[dlr]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

Dave, what mistake are you referring to? Changing the axial response will change the power response, correct? I think I also had in mind that it will of course change the room response as well.

I've been reading a lot of posts here for a long time, and yes, many of them mention power response or axial response. This usually has to do with what might sound better for one or the other (or for in-room), but I don't recall seeing anything about dealing with these issues successfully all at once. Do you have any posts in mind that deal with some methodology for balancing the 3: axial, power, room?

I meant to say that the mistake is to focus too much on the axial response and not consider the off-axis and by extension the power response. It was in reply to your statement "It seems like there are lots of tweaks for axial response which neglect the power effects." with which I was agreeing.

The axial response can be changed without a lot of change in power response. Each case is different and there are different aspects to consider. For multi-way systems you have driver directionality, baffle dimensions, crossover type, order and Fc, driver spacing, omni/dipole/bipole, etc. A change in axial response doesn't say anything at all as to the power response impact. Each case has to be considered individually.

Check out the page at john k's site related to power response. You'll find a lot more information there with some examples of some of the issues.

dlr

[lhwidget]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

Hi Oldermongrel,
You might want to check out Linkwitz's study on diffraction/baffle step phenomena. It looks to be exactly what you are looking for. I'm like you, I tend to think of baffle step and diffraction as the two separate phenomena they are. I suppose they are usually lumped together because they are both governed by the baffle's dimensions.

http://www.linkwitzlab.com/diffraction.htm

I'm like Augerpro, I thought Olsen's theories were tested before he published. Linkwitz experiments ran to the same conclusions as Olsen's.

Spastic,
You need to talk to Curt Campbell (Exclamations).

http://speakerdesignworks.com/Exclamations.html

[dlr]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

Hi Oldermongrel,
You might want to check out Linkwitz's study on diffraction/baffle step phenomena. It looks to be exactly what you are looking for. I'm like you, I tend to think of baffle step and diffraction as the two separate phenomena they are. I suppose they are usually lumped together because they are both governed by the baffle's dimensions.

Step and "diffraction" are discussed separately, but they are both due to the same phenomenon. That's what confuses folks, thinking of them as separate phenomena. They are one and the same.

I'd again suggest reading this page. You'll see examples that show the baffle step as being just one part of it. It's all determined by baffle dimensions and wavelength, both step and the rest of diffraction.

dlr

[OlderMongrel]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

Thanks to all for the tips and links. I've already checked out some of this, but need to study more. I tend to have contrary opinions and might come back to bounce (diffract :D) some more of them off you.

One of these days I hope to develop a Grand Unified Axial-Power-Room Response Driver-Crossover-Baffle-Vent-Stuffing Power-Compression Floor-Bounce Pressure-Mode Room Modification and Treatment, Measurement and Evaluation Tool with a software interface that scans your room, asks you a few questions, and voila, out pops all the design info and setup requirements, like Head Vice adjustments, which music to play and how loud, etc.

Naaaaa, that would take all the black magic out of it....

John

[Robert Dunn]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

With all due respect to a prior poster, I see baffle step loss and baffle edge diffraction as being very different, albeit related, effects. Step loss yields a drop in on-axis low frequency pressure while edge diffraction yields a high frequency comb effect. I think much of the confusion about the topic comes from loose phrasing such as "baffle step diffraction."

Just my two cents....

[dlr]

Re: (I'm back!) Is baffle step loss a poorly developed concept?

With all due respect to a prior poster, I see baffle step loss and baffle edge diffraction as being very different, albeit related, effects. Step loss yields a drop in on-axis low frequency pressure while edge diffraction yields a high frequency comb effect. I think much of the confusion about the topic comes from loose phrasing such as "baffle step diffraction."

Just my two cents....

The math that describes the interaction of the wave at the baffle edge covers both "cases". It's nothing more than a continuum of a single phenomenon. The "step" is nothing more than the area in which the full transformation from 2-pi to 4-pi is completed. The transition region is simply where the acoustical properties of air create wave movement that differs with frequency. The concept of "step" is just an easy means to visualize it.

dlr