Thanks!
Yes the pivot point is not quite on the front baffle axis, it is about 1" behind it, which does introduce a little error. I could move the speaker back a little further but then I start to get diffraction effetcs off the platform.
edit: I guess you picked that up from the impulse response? I time zero locked it so that time of flight was absolute between the measurements
Tony.
I noticed that there seems to be significant directivity even in the low hundreds, which is unusual. Therefore I thought that maybe it was caused by the increased distance traveled for the greater angles. One inch can't explain the effect. Any idea?
That depends on the measuring distance. If you're measuring at several feet away, I doesn't matter much, but indoors you might have to measure at a closer distance, maybe 3 feet. Then 3 inches becomes too significant to just ignore.
If you are measuring at distances so close that a 1" offset is an issue then you are certainly too close for valid far field data.
The gating was at a bit over 300Hz so anything below that is not to be trusted at all
at 500Hz the difference between off axis and 90 degrees is 6db and it seems that the variation is reasonably constant depending on anhle. As this is the first speaker I've measured properly at multiple angles (and I've not compared this measurement to other peoples measurements as yet) I don't actually know if that is a typical result. At the 30 deg mark I think that they are excellent up to close to 5K being less than 2db difference across the frequency range to that point.
Tony.
Yes, and the frequency can remain quite well matched up to probably 80deg each way, but there are tradeoffs in the process of obtaining this.
The B&O design is actually licensed, at least during the early days. For the Beolab 5 as example, that kind of design relies on combined reflections. What this does is that it does make the sound a bit diffused. But if the CSD of the drivers stretch out, it seems that it is not a critical factor. You need to place these type of speakers far away from side walls though.
0 and 30deg SPL and Phase (speakers are the ones in my avatar)
Driver polar plot calculated from a Klippel scan at 20KHz
From the calculation reports I have seen, the directivity suggested by Earl is probably the most reasonable in rooms where side walls cannot be further away from speakers than the back walls.
An externally hosted image should be here but it was not working when we last tested it.
Driver polar plot calculated from a Klippel scan at 20KHz
An externally hosted image should be here but it was not working when we last tested it.
From the calculation reports I have seen, the directivity suggested by Earl is probably the most reasonable in rooms where side walls cannot be further away from speakers than the back walls.
Last edited:
Not too brite... hehe
Hi Dan...
Yeah... that's the curve b4 I used Audessey for in room correction during movies, etc. Contrary to what seems to be the collective perspective here pooh poohing such methods, my results have been pretty positive. i.e. i hear definite improvements in localisation and sound stage on appropriate material after correction vs. 2 channel or matrixed surround modes w/o Audessey. Personally, I believe Audessey works best if you don't expect it to do too much... which I believe I've achieved with this setup.
These are in a pretty size-able room, 2 story slanted cathedral ceiling, maybe 15,000 ft^3 with lots of diffusion, so the treble has plenty of time to dissipate
When I re-run these and do the inverted phase, I'll re-scale to smaller values and also do some smoothing as those were run with minimal back when I first modded the xover.
John L.
Hi Dan...
Yeah... that's the curve b4 I used Audessey for in room correction during movies, etc. Contrary to what seems to be the collective perspective here pooh poohing such methods, my results have been pretty positive. i.e. i hear definite improvements in localisation and sound stage on appropriate material after correction vs. 2 channel or matrixed surround modes w/o Audessey. Personally, I believe Audessey works best if you don't expect it to do too much... which I believe I've achieved with this setup.
These are in a pretty size-able room, 2 story slanted cathedral ceiling, maybe 15,000 ft^3 with lots of diffusion, so the treble has plenty of time to dissipate
When I re-run these and do the inverted phase, I'll re-scale to smaller values and also do some smoothing as those were run with minimal back when I first modded the xover.
John L.
Last edited:
I think this is the one that did some calculations based on ITD/IID
AES E-Library Optimum Loudspeaker Directional Patterns
AES E-Library Optimum Loudspeaker Directional Patterns
http://decoy.iki.fi/dsound/ambisoni...udspeaker directional patterns_Kates_1980.pdf
Copy on the web.
David S.
Copy on the web.
David S.
Kates is looking for a means of compensating for image shift as a listener moves off the mid line of a 2 speaker stereo setup. He solves for polar patterns that give delay compensation at low frequencies and level compensation at high frequencies for position shift (2 patterns). He shows that the polar pattern needs to vary in directionality based on the spread between the speakers (more narrow pattern needed as the speaker spread gets wider).
Since he ends up with directional speakers heavily crossed in front of the listener, I'm not sure why Earl would declare it "not very relevant". Read it yourself and decide.
David S.
Since he ends up with directional speakers heavily crossed in front of the listener, I'm not sure why Earl would declare it "not very relevant". Read it yourself and decide.
David S.
Almost looks like for the ultimate image, you are going to need some sort of dipole/WG combo with a fully absorbed front wall. I believe it was Gainphile who talked about dipoles in a garage with the door open having amazing imaging. The pattern of each section needs to vary though depending on position. Would be something interesting to do. Now I'm getting ideas. This could be dangerous.
Dan
Dan
Not Earl but I think Kate's view is too simplistic and the paper also only looks at a center image in an anechoic environment. He also believes that direct sound equals localization and reflections equal "ambience enhancement". It can easily be shown that reflections can do both, shift the image (and not just broaden it) and add spaciousness. Build-up and breakdown of precedence is a complex issue.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.