Hi,
Maybe for a comparison, generate the dipole with 2kHz low pass front wall reflection and for a monopole add floor and ceiling reflections.
- Elias
Maybe for a comparison, generate the dipole with 2kHz low pass front wall reflection and for a monopole add floor and ceiling reflections.
- Elias
I agree that it was reasonable but it was also misleading. His sound files demonstrate what a microphone "perceives" which is not what humans perceive.
The front wall reflection of my dipole line array can be seen here at 7 ms: There is practically no ceiling, floor nor side wall reflections.
But for a monopole box there will be ceiling, floor and side wall reflections that screws the response between 500Hz-2kHz. And below 400Hz it's just a big mess.
Note that at these low freqs (below 1kHz) reflections don't add to spaciousness but they only mask other sounds, upwards freq masking, backwards and forward temporal masking.
- Elias
But for a monopole box there will be ceiling, floor and side wall reflections that screws the response between 500Hz-2kHz. And below 400Hz it's just a big mess.
Note that at these low freqs (below 1kHz) reflections don't add to spaciousness but they only mask other sounds, upwards freq masking, backwards and forward temporal masking.
- Elias
We've discussed this before. You have a valid point if there is great angular disparity between the source and the reflection in question (only a single reflection is modeled). But in the case of sounds on the median plane (half way between your ears), sounds vertically displaced (the floor bounce) or sounds not far apart in angle, binaural hearing will not separate the two sounds. The reflection will, in that case, sound just like the simulation.
If the source were in front of you and the reflection to the side, then you might hear them as separate sources rather than a single source with response errors. Thats why I keep repeating that side wall reflections are better than front wall reflections.
David S.
Sorry, I must have missed this earlier.
I regularly measure at one meter, two meters and ten meters. Sometimes at other distances. I see the notch at all microphone distances. The only times I don't see it is when I do a ground plane measurement or a pseudo-anechoic with that reflection gated out.
Here's how my dipole measures in my room:
Then the monopole under similar, but not same conditions:
Not that these mean much of anything, but the dipole does look to have a better behaved midbass.
Edit: anyone want to care what their polar response looks like? 10 points for anyone who can get it right.
Dan
Then the monopole under similar, but not same conditions:
Not that these mean much of anything, but the dipole does look to have a better behaved midbass.
Edit: anyone want to care what their polar response looks like? 10 points for anyone who can get it right.
Dan
Last edited:
Just for you Elias, here is a sim with reflection filtered at 2kHz. So it is noise, plus a reflection with 11ms delay at -5dB and 4th order filtering at 2k. I've also included the previous -5dB unfiltered reflection case.
These are interesting. The unfiltered has the "tunnelly" sound plus extra high frequencys from the reflection. The filtered version just has the tunnelly sound. This probably makes sense if you consider that reflections add comb filtering and at at higher frequencies the teeth of the comb are so close that their comb structure (audibly) disappears. The high end part of the reflection sounds just like normal hiss. You sure you want dipole for low frequencies and not for high?
Again its 2 seconds off, 2 seconds on, 2 second off, etc,. and you must replace the ZIP with MP3.
Regards,
David
These are interesting. The unfiltered has the "tunnelly" sound plus extra high frequencys from the reflection. The filtered version just has the tunnelly sound. This probably makes sense if you consider that reflections add comb filtering and at at higher frequencies the teeth of the comb are so close that their comb structure (audibly) disappears. The high end part of the reflection sounds just like normal hiss. You sure you want dipole for low frequencies and not for high?
Again its 2 seconds off, 2 seconds on, 2 second off, etc,. and you must replace the ZIP with MP3.
Regards,
David
I suspect the speakers were about two feet off the ground in both cases. Both have the same floor bounce notch at around 150Hz. Any point source at ear level does this, dipole or monopole. Add a speaker nearer the floor and run it up to 200Hz or so and that will go away. Sort of a helper woofer to smooth that range.
Here's how my dipole measures in my room:
Then the monopole under similar, but not same conditions:
Last edited:
You are missing what I've already made explicit in a diagram for you... if you were to simulate even 2 rear reflections as illustrated in my very simple diagram, you would be just a little closer to the reality of a listening room.... but still quite a ways from the reality of multiple reflections of varying path length. (i.e. diffusion)
I used my 15" midwoofers as dipoles before I used the same drivers as monopoles. For reference, as dipoles I crossed them at the schroeder frequency to Fostex 3.5" full rangers. Now I cross just over an octave higher to waveguides.
The benefit of the dipoles was they revealed sound like I hadn't heard before. They seemed to sidestep some of the room modes and let the music come through in a balanced way, which I think we all agree in the midbass is so important.
Since treating my room with bass trapping and using multiple subs, I can also achieve this effect. I feel I have more practical control over it too, so I can take it to the nth degree.
The primary difference between the two methods is the echo/tunnel effect, which I initially preferred in the dipole once I settled in to believe that was the cause of the midbass improvement. Now I see it as the price to pay for that midbass improvement.
The benefit of the dipoles was they revealed sound like I hadn't heard before. They seemed to sidestep some of the room modes and let the music come through in a balanced way, which I think we all agree in the midbass is so important.
Since treating my room with bass trapping and using multiple subs, I can also achieve this effect. I feel I have more practical control over it too, so I can take it to the nth degree.
The primary difference between the two methods is the echo/tunnel effect, which I initially preferred in the dipole once I settled in to believe that was the cause of the midbass improvement. Now I see it as the price to pay for that midbass improvement.
A single reflection is a "reasonable" simulation of the complex environment of a listening room with multiple reflections? It's not an opinion that there are multiple reflections. Is is also not an opinion that the reality of the listening environment is infinitely more complex.
So, I'm also very unclear as to what can be learned by the simulation presented except to prove that reflections exist and can cause comb filtering, which of course none here would dispute.
You just set back the argument for using confuser simulations 10 years.
Reality definitely trumps your conclusion.
Your right... they don't mean anything because one graph has smoothing applied and one does not.
reality check... a quick comparison
So in the interest of presenting something that actually represents reality, I present a comparison of monopole vs dipole in the same room under the very nearly the same testing conditions. (one speaker beside the other... so slightly different distance to sidewall with monopole slightly farther away from sidewall. Both same distance from back wall of ~4ft)
Both tests shown are with 1/48th octave smoothing applied with a 200ms gate.
Anyone care to guess which is which?
Just in case you were wondering, here is the same test except that both are presented without smoothing applied (same 200ms gate on both)
dipole:
monopole
So in the interest of presenting something that actually represents reality, I present a comparison of monopole vs dipole in the same room under the very nearly the same testing conditions. (one speaker beside the other... so slightly different distance to sidewall with monopole slightly farther away from sidewall. Both same distance from back wall of ~4ft)
Both tests shown are with 1/48th octave smoothing applied with a 200ms gate.
An externally hosted image should be here but it was not working when we last tested it.
Anyone care to guess which is which?
Just in case you were wondering, here is the same test except that both are presented without smoothing applied (same 200ms gate on both)
dipole:
An externally hosted image should be here but it was not working when we last tested it.
monopole
An externally hosted image should be here but it was not working when we last tested it.
And what else did you expect? That was enough for me. Thats a lot for such a simple demo.
And pragmatism trumps your sarcasim. You expect too much from a simple demo and you fail to grasp its implications.
It has nothing to do with the simple demo, but rather, the sweeping conclusions he came to from such a simple demo.
The implications from the demo are rudimentary at best.
Actually, "implications" is too strong a term for the subject at hand.
I'm sure the "simulator" can replicate those measurements... no ?
Sometimes I'm puzzled with the efforts, arguments, etc. based on simulation when one can simply build and point a $3 microphone.
My response had nothing to do with beliefs. At best, I'm an agnostic on most everything in life, and the hobby of designing loudspeakers doesn't even qualify for consideration when it comes to beliefs. I have absolutely zero skin in the game.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.