If the objective is a donut shaped radiation patter, I just don't see that working. At every useful observation point you will have the combination of direct sound and sound off of the reflector. Plus, at a lot of upwards angles the reflector doesn't obscure the dome at all.
Signed,
Sceptical in Toronto
Signed,
Sceptical in Toronto
In my post I was only considering the frequency response of the direct signal affecting perceived image location.
Interesting suggestion of yours, but I'm not sure that errors in the frequency response of the side-wall reflection would have the same effect.
Even though the off-axis response of most speakers would skew the frequency response of the side-wall reflection, conceivably in a way that could trigger a change in perceived localization (compared to a spectrally neutral reflection) in isolation, it seems likely that with the presence of the direct signal the precedence effect would render this somewhat moot. Maybe there would be a small change in the ASW.
What would be quite interesting, if it were easily available, would be the HRTF model of a dummy head in software such that you could adjust the incident angle of a sound source in 3D and see exactly what frequency response modification this causes.
I've seen printed graphs of frequency response vs horizontal angle, but only in one axis, being able to select any arbitrary horizontal and vertical angle and immediately see the result would be quite neat, and informative. Has anyone come across something like this that's free (for non commercial use) ? 🙂
I must be lucky, or have a statistically "average" ear shape because most of the dummy head binaural recordings I've listened to are extremely convincing, I don't even have to close my eyes to convince myself I'm there.
Some of the software synthesized binaural recordings I've listened to also work very well, which are presumably based on computer models of statistically average head/ears.
Doesn't help people with more unusual ear shapes who don't perceive "standardised" binaural recordings well though...
As far as I understood the Smyth Realizer just simulates normal speakers in normal rooms. One would have expected something more intelligent from such a device.
with what source? - everything currently produced commercially is meant for loudspeaker/room playback
the goal of the Smyth SVS Realizer is to "improve" headphone reproduction Right Now, not be a blind alley like Ambisonics or other "theoretical" system that would be "ideal" if only the ingnorant recording companies recogized "the truth" and adopted "the right" spatial encoding/recording methods
the goal of the Smyth SVS Realizer is to "improve" headphone reproduction Right Now, not be a blind alley like Ambisonics or other "theoretical" system that would be "ideal" if only the ingnorant recording companies recogized "the truth" and adopted "the right" spatial encoding/recording methods
The Smyth Realiser does exactly what it is designed for: capture the acoustics of real speakers in real rooms and make it available for headphone playback.
Doesn't a cone driver behave like a diffraction device by itself? Doesn't a driver that starts to beam at higher frequencies show a diffraction-like effect?
If you want to get "real theoretical", yes, thats true. The response of a piston is exactly the same as the sum of a point source and a difrracting ring at the disks edge. This means that any finite size source has its own internal diffraction - thats what creates its polar response pattern. But this diffraction is very very close in timing to the "point source" at the disks center so the dealy is miniscule. Thats will not be the case with what you are showing.
By the way, an ideal waveguide acts like a point source with directivity, a perfectly shadded diffractionless point source. There isn't even a psuedo-diffraction from the edge of the disk source as in the case of a circular piston.
For a direct comparison I don't. I just opened a Sound Cloud account that I posted a partially a dummy head recording/partially DI. With headphones you should be able to tell what is what.
Alone in California 1 by dantheman-10 on SoundCloud - Create, record and share your sounds for free
Maybe not helpful,
Dan
Is this just a theoretical concept or would a porous absorber ring covering the outer parts of a cone actually reduce the driver's high frequency directivity?
Last edited:
Thanks Dan. It would be great to hear a comparison of the dummy head and your head. Cars passing by or ocean waves would probably be revealing.
So this is your DIY music ? I'm baffled ! Not being a specialist, this reminds me Neil Young.
For the SQ, I will say that the first part sounds better to my ears, but that's not very obvious.
For the SQ, I will say that the first part sounds better to my ears, but that's not very obvious.
Here's data from a bigger reflector (diameter about 34cm which is about double the effective diameter of the driver):
0°-90° vertical, without reflector:
0°-90° vertical, with reflector:
What are we looking at? Is this a horn with unusual shape and driver location?
An externally hosted image should be here but it was not working when we last tested it.
0°-90° vertical, without reflector:
An externally hosted image should be here but it was not working when we last tested it.
0°-90° vertical, with reflector:
An externally hosted image should be here but it was not working when we last tested it.
What are we looking at? Is this a horn with unusual shape and driver location?
Markus,
I would expect in this settup a certain amount of horn loading. Have some questions interpreting this:
1) 0-90 degrees vertical, does this mean 1st measurement on axis, etc.?
2) for the non-reflector case it is easy to understand which line should correspond to a further deviation from on axis. For the reflector case, the lines are garbled to the extent this is no longer possible. Would it possible to assign different colors?
3) both measurements start at around 0dB. Have you normalized this by adjusting volume of distance from measuring mic? Because in case of horn loading, it would be expected that ceteris paribus the SPL is higher in the range where this horn is most effective.
The answer to 3) will give a direct indication of the amount of horn loading going on.
vac
I would expect in this settup a certain amount of horn loading. Have some questions interpreting this:
1) 0-90 degrees vertical, does this mean 1st measurement on axis, etc.?
2) for the non-reflector case it is easy to understand which line should correspond to a further deviation from on axis. For the reflector case, the lines are garbled to the extent this is no longer possible. Would it possible to assign different colors?
3) both measurements start at around 0dB. Have you normalized this by adjusting volume of distance from measuring mic? Because in case of horn loading, it would be expected that ceteris paribus the SPL is higher in the range where this horn is most effective.
The answer to 3) will give a direct indication of the amount of horn loading going on.
vac
I am hardly surprised that an active proponent of constant directivity speaker systems would be quick to disparage multidirctional systems whether they are created by passive reflections as in the example of posting #1 here or by use of multiple active drivers. These seem to be entirely opposite views of what a loudspeaker system should do.
You'd have to admit that the multidirectional approach has had a very large number of advocates too. The example in post #1 is probably a 4" or 5" driver which makes a very poor tweeter and would not be expected to radiate much energy in the highest and even second highest octave. Proponents of multidirectional speakers come in many camps. The reflector idea is an old one tried by many especially in the quadraphonic heyday of the 1970s. Others have pointed out some of the more popular ones such as those widely distributed by Zenith. But whether passive or active, according to my own model all of them I'm aware of have serious execution flaws.
Full range (except sometimes for low bass) bipolar speakers both planar exotics like electrostatic panels and more conventional speakers like some models from Mirage are one camp. Very wide dispersion direct radiator models like those from Acoustic Research, especially LST and its Cello clone, those from Allison and Soliloquy are another. Then there's the famous (infamous) Bose 901 which is a camp all by itself. Then there's the direct radiator with a rear firing tweeter. Lots of highly regarded speakers were in that camp. Snell AIIIi, Vandersteen 5, Revel Salon Ultima (first version) just to mention a few. I think it was Olson who postulated that the ideal speaker would be a pulsating sphere. Bose started out with an eighth sphere placed in a room corner, the assumption being that this would immitate a perfect spherical radiator (he was probably right about that, not that it mattered.) Like his 901 effort his design IMO was flawed for a number of reasons, not the least of which was that a 4" driver made a lousy tweeter and the speaker didn't seem able to produce high frequencies, at least not to my ears. I almost left out the highly regarded Soundstream speaker which was yet another variant.
It's interesting to see how actual acoustic instruments radiate sound. Most are multidirectional or indirect radiators. These include pianos, string instruments, and percussion instruments. Those you'd think of as more directiontional, horns which include brass and woodwind instruments are directional but rarely aimed directly at the audience. Normally they are held vertically(clarinets and oboes are examples), at an angle downward (trumpets and trombones are examples), or with their mouths straight up (tubas) or sideways (french horns) so that even the first arrival at the listener is mostly reflected sound. Not only that but their spectral balance that affects tone doesn't vary much with direction. This can easily be experienced at a piano bar or near a street busker by walking around and noticing what change in tone you hear. Usually not much. Now put your speaker in the middle of a room or better yet outside away from any walls and walk around it while it's playing and see what changes you hear. A very different effect.
1. The driver facing the mic is 0°. For the box without the reflector, the listening axis would be 0°, for the box with reflector it would be 90°.
2. Here you go:
An externally hosted image should be here but it was not working when we last tested it.
3. No further processing. The rotation axis is probably not at the acoustic center that's why we see some variation at the low end.
Marcus, you are quick!
From the fact that there is no increase in SPL at any frequency, I have to conclude there is no horn loading.
Rather, a lot of destructive interference appears to be taking place.
I also can see that the reflector does help to disperse sound in the 90 degree region, but at a high price (wild FR swings). This while the speaker without reflector is admirably flat. Within +/- 2 dB for the region below 9K is very nice. Which driver are you using?
From the fact that there is no increase in SPL at any frequency, I have to conclude there is no horn loading.
Rather, a lot of destructive interference appears to be taking place.
I also can see that the reflector does help to disperse sound in the 90 degree region, but at a high price (wild FR swings). This while the speaker without reflector is admirably flat. Within +/- 2 dB for the region below 9K is very nice. Which driver are you using?
Response-wise you still have more energy in the 0 degree (upward) direction. Even the treble seems to be difracting around the cone. From 50 degrees or so you start to see the lateral energy come up and contribute from about 4 kHz up. Still, the response is nowhere like the donut shaped ideal that the marketing types of such products would suggest.
It seems like the direct response of the tweeter still dominates until you are far off axis. The only way to get to the idealized omni pattern is to increase the bounced energy and reduce the conventional axial ienergy I know a number of these units are shaped more as 360 degree radial horns. This might pick up response issues that you don't want, but it would likely send more energy outwards and less upwards.
David S.
Last edited:
- Status
- Not open for further replies.