[JohnR]

Seems like a good candidate for eg a miniDSP, or more accurately  vice versa 

Hi Rick, I'm quite fond of active/DSP systems, as I expect you know, but I was able to get flat response with a passive crossover to a dipole woofer using parts I had on hand. It is a bit tricky, because of the capacitive nature of ESL panels. Below 100 Hz on the woofer I still needed some active mojo, but that's life with dipole woofs. I'm going to have another go at the crossover though to reduce the nr of parts.

[rick57]

Thanks, JR

(I see you are a Sydney boy: I was too for 35 years)

Any further thoughts on the panels? . . see my next post for a broader  question   

Cheers

[rick57]

Ron from ER emailed me today, that the – 3 dB angle of vertical dispersion is +/ - 15 degrees.
Ie very limited, but I would you agree have thought would be enough for a seated solo listener. And often we can be audio bachelors, yes?

Narrow dispersion would not suit speakers in eg the kitchen, where we walk around - but be fine for great clarity in home theatre/ TV, unless you are prone to jumping from yours seat in the exciting bits of a film, or when your team is pulling ahead in tight game of your favourite sport . . and fine for listening critically in one sweet spot

How big should the sweet spot be?

an older audio engineer said to me today that even when seated, 
"wide dispersion in both directions sounds 'better' than narrow.  This seems to be a pretty common opinion"

While as the self taught enthusiast, I  think that a bigger dispersion would increase reflections and room effects generally. I would have thought that a bigger dispersion is more often a dis-advantage.

What is right?

Cheers

[TRADERXFAN]

John,

The graphs showing the comb filtering in the upper ranges look offputting. But you were saying something about raking back the panels helped. Any more comments/updates on how this is progressing? How does upper end sound? Are you happy without a dedicated tweeter?  I see you agreed they had some special qualities which makes me curious about these for a near field speaker I am thinking about. Would combine with my ob servo subs from Danny at GR.

-Tony

[rick57]

Tony

at 200 Hz, the ER Audio mini panels have their peak distortion.

What frequency do the ob servo subs from GR go up to?

Cheers

[TRADERXFAN]

Said to be 300hz...

[rick57]

Then response could be fine, on axis

Bu might be worth finding out the directivity of the OBs. At the crossover, there would be a big jump from the narrow directivity of the mini panels, to the OBs

I've read a few times that big jumps in directivity are not good.

Though I'm not clear on how it's audible    ..

[josh358]

Then response could be fine, on axis

Bu might be worth finding out the directivity of the OBs. At the crossover, there would be a big jump from the narrow directivity of the mini panels, to the OBs

I've read a few times that big jumps in directivity are not good.

Though I'm not clear on how it's audible    ..

A lot of the sound we hear is from room reflections. If polar response isn't consistent, either the amplitude response of the direct sound is wrong, or of the reflected sound, or both, and the brain can to some extent anyway distinguish between them, so there's no way to EQ it out. Harman International says it has an algorithm that does an excellent job of predicting subjective quality on the basis of anechoic measurements at various points around the speaker. There's a lot in Floyd Toole's book about research he and his associates did on the importance of having both good power response and on-axis response, and a critique of the common problem of poor power response at the crossover point.

[JohnR]

The graphs showing the comb filtering in the upper ranges look offputting. But you were saying something about raking back the panels helped. Any more comments/updates on how this is progressing?

Hi, technically I think that's lobing, not comb filtering. I don't remember if a different baffle shape would make any difference here, will have to redo some sims when I get a chance.

The raking back is so that the sound doesn't change so much between seated and standing. I'm not sure what effect it would have on off-axis response, to be honest... good point...

How does upper end sound? Are you happy without a dedicated tweeter?  I see you agreed they had some special qualities which makes me curious about these for a near field speaker I am thinking about.

Nearfield is probably the easiest to work with. What distance are you thinking of?

As said, I'm still messing with the crossover (when I get a chance - not often). So I've only got one running.

Would combine with my ob servo subs from Danny at GR.

-Tony

The panel has a very strong resonance at 105 Hz. (Although less so if the sock is left on apparently.) Not sure how you are planning to do the crossover but if at 300 a fairly steep slope is probably needed.

[rick57]

Nearfield is probably the easiest to work with. What distance are you thinking of?

John

As – 3 dB vertical dispersion is only +/ - 15 degrees,
unless you keep your head still, I would have thought they'd be only for the mid-far fields?

                             m      feet      m     feet      m   feet

Listening distance   0.5      1.6      2    6.6               4   13.1

 - 3 dB                   0.13     0.44      0.54    1.76      1.1   3.5

[tabrink]

Curious as I am very interested in this product for nearfield application.. are you actually using these in your space for a basis of your comments? 
Or reacting and posting from  information given you.
Best,
Tom

Ron from ER emailed me today, that the – 3 dB angle of vertical dispersion is +/ - 15 degrees.
Ie very limited, but I would you agree have thought would be enough for a seated solo listener. And often we can be audio bachelors, yes?

Narrow dispersion would not suit speakers in eg the kitchen, where we walk around - but be fine for great clarity in home theatre/ TV, unless you are prone to jumping from yours seat in the exciting bits of a film, or when your team is pulling ahead in tight game of your favourite sport . . and fine for listening critically in one sweet spot

How big should the sweet spot be?

an older audio engineer said to me today that even when seated, 
"wide dispersion in both directions sounds 'better' than narrow.  This seems to be a pretty common opinion"

While as the self taught enthusiast, I  think that a bigger dispersion would increase reflections and room effects generally. I would have thought that a bigger dispersion is more often a dis-advantage.

What is right?

Cheers

[JohnR]

Rick, your math doesn't work because they are not a point source. 440mm high.

[rick57]

John

I know they're not a point source, but didn't know that made a difference to the maths . . 

What is the maths for a driver of that size?

[JohnR]

I know they're not a point source, but didn't know that made a difference to the maths . . 

They're like a short line source.

Could you check your PMs please...

[JohnR]

I decided to give the "nearfield" listening setup a try and placed the panels for a listening distance of 1.2 m. Whew that's close! With a bit more messing with the crossover with parts I have on hand, here's the in-room response:



I've got the panels tilted slightly back slightly. I'll post the crossover schematic when I get time to draw it, but the highpass is 4.7R in series with 12 uF, and then 3.3mH to the negative. The woofer is a similar topology - nominally second order with the panel inverted.

This is the response of the individual drivers:



It's not exactly "textbook" but it's what I could get with parts I have on hand. I did have the effective crossover lower but didn't have enough large-value parts for two speakers, so I went with a higher frequency for now.

Got another thing to try with this, more later.

[rick57]

re these drivers, I'm trying to understand/ extrapolate as much as theory tells us, not just in their response on axis

I'm not that familiar with line sources, but had thought that with width to height of only about 0.4 they wouldn't behave like one.

But IIRC, compared to ‘normal sources’, line sources have a directivity that is higher. Is that correct?

Cheers

[JohnR]

As far as I'm aware, you can consider the horizontal and vertical fairly much independently. An ideal point source radiates equally in all directions. However real drivers are not points, and as the driver size increases relative to wavelength, the radiation pattern narrows. So the "directivity" of a driver (in a flat baffle) is related to its size and the frequency that you are considering.

As the size of the driver increases further, higher frequencies will come into the "near field". This is discussed in detail in the Jim Griffin line array white paper. In the near field, the radiation pattern (of a line) is cylindrical; in the far field, it is spherical. At a given frequency, the transition from near field to far field is determined by the length of the line - the longer the line, the lower the frequency of the transition. It turns out that if you want to be in the near field down to say 150 Hz in a real room, you need a line from floor to ceiling, so the reflections from the floor and ceiling effectively make the line longer.

So, two "ideal" configurations are a point source, or a long but thin line.

For anything in-between, where the line is shorter and not very thin, i.e. most panel speakers, you have some sort of compromise. Frequency response (on and off axis) varies with listening height and distance to the panel. I suspect this is the reason why placement of panels is generally considered so critical, because the listener is trying to balance these things in their room.
 
If smooth off-axis response is a primary goal, then panels are probably the wrong place to look.

Getting back to the mini panels, I think that it's better nearer than further. Since you're in Melbourne (right?), I believe Involve have a pair set up there, you could go and listen for yourself?

[James Romeyn]

error, deleted

[JohnR]

Well, you know, I think it's a lot simpler (and cheaper) to just use the ESL panel under discussion. Here in the Planar Circle...

[josh358]

http://jamesromeyn.com/home-audio-gear/ramca-romeyn-ambiancemode-cancelling-array/

It's an interesting web page that discusses some concepts that I'm also interested in. Since I don't want to take the discussion off topic or hijack the thread -- I've sometimes wondered whether/if one could do something like this with planar dipoles, e.g., direct sound to the first sidewall reflection with a >10 ms delay. It isn't something that dipoles want to do, of course, because of the dipole null. But suppose you used two dipoles? You could then apply delay you wanted to the ambiance dipole if the path length weren't sufficient, or maybe just put it closer to the center of the room. The Magnepan on-walls are designed to direct the backwave to the side walls, and I've wondered if this effect has something to do with the enhanced spatial performance of Magnepan's tri-center. I wonder if this could also improve the power response of speakers made with beamy ESL panels. With small panels like these, it would be easy to mount an angled ambiance panel above or below the on-axis one. However, off the top of my head, I'm not sure what the effect would be -- I imagine you'd end up with a quadrapole, and four nulls.