Yes. I did that with a wide dispersion speaker and the Nathan's. In order to send the same frequency response to the wall and the listening position the toe-out was 9°, i.e. the listening axis was 39° from the speaker's 0° axis.
Although the wide dispersion speaker had a more spacious representation, the Nathan's didn't sound tremendously spatially different. This was somewhat unexpected.
Here's the ETC for wide dispersion - Left, Right:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
...Nathan - Left, Right:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Are you sure your expectations are correct ?
An unbiased listener would expect that when recording a spoken voice in anechoic chamber and playing it back with your stereo system it would sound like a spoken word in anechoic chamber.
Of course it will not ! Stereo system is a complete failure to start with. It cannot reproduce the acoustic space of the original venue, not even a clue about it in this case. Whatever you do with your stereo system is just wishful thinking.
- Elias
When are you going to start doing wavelet plots ?
When are you starting to answer my question how a dipole line array can help with stereo speaker crosstalk?
Member
Joined 2009
Markus,
I want to point out that in the graphs above the left and right Natans don't produce similar reflections. It is likely that they are not turned at the same angle and I find that to audibly smear the stereo image.
Can you give us more information please? Do you know which reflection peak corresponds to the near sidewall and which to the far? (A mattress on the wall is one strategy to hunt it down).
I want to point out that in the graphs above the left and right Natans don't produce similar reflections. It is likely that they are not turned at the same angle and I find that to audibly smear the stereo image.
Can you give us more information please? Do you know which reflection peak corresponds to the near sidewall and which to the far? (A mattress on the wall is one strategy to hunt it down).
I'm not sure what you mean there. Are you saying that both setups project equal image width but are toed-in differently? Can you do a quick sketch of your setup please?
They are turned at the same angle but the left side is two windows and a glass door while the other side is a plain wall. So although the graphs indicate asymmetry, the perceptual result is symmetry. Our hearing groups reflections in time but an ETC doesn't
I'll post that tomorrow.
Markus, now you are talking ! To generate strong ipsis, aim speakers towards side walls. Left to left, right to right. Also maximises stereo "channel separation". And minimises IACC. All good !
Are you not aware this is strongly against the user manual !
Very complicated ! Better to measure with dummy head at the listening position. Get all the reflections tracked at once.
- Elias
The graph starts at 60ms ? Or is that really your zero reference time ?They are turned at the same angle but the left side is two windows and a glass door while the other side is a plain wall. So although the graphs indicate asymmetry, the perceptual result is symmetry. Our hearing groups reflections in time but an ETC doesn't
Assuming so, what's interesting (although not unexpected) is that the initial decay over the first few ms or so is much cleaner and decays faster on the more directional speaker - less baffle diffraction effects perhaps, as well as less floor reflection ?
After 10ms there is not a lot of change - slightly less at some delays, more at others, suggesting that the overall reverberent field is not hugely different in level. That makes sense because once the directional beam hits the far walls it will tend to scatter (especially if it hits a corner) and illuminate the rest of the room via multiple reflections.
So it seems in this instance that the directivity is very successful in cleaning up the first 10ms without killing the reflections beyond 10ms that a heavily damped room would do.
Look what made the news today--
https://www.nytimes.com/2011/09/06/science/06sound.html?ref=general&src=me&pagewanted=all
https://www.nytimes.com/2011/09/06/science/06sound.html?ref=general&src=me&pagewanted=all
"It turned out that the Immersive Audio Lab at U.S.C.’s Viterbi School of Engineering is dark, a bit dingy, and only 30 feet wide, 45 feet long and 14 feet high."
I would be happy (and getting better results) if I had a living room that size too!
Now lets see them get the same results in a room 11 feet wide by 15 feet long by 8 feet tall, like mine....
A room this size is a much bigger challenge to get good acoustics in, IMHO, and leaves very limited options for placement.
I'm afraid that is what they call a "notional diagram" rather than a real design. Simply stringing drivers together with inductors in between will not provide effective delay. Raising the inductance in the hopes of more delay will roll off the energy preventing it from going down the line. I've done the odd latice type all-pass before and gotten a little bit of practical delay for phase shifting a tweeter, but it is hard to achieve significant passive delay.
Assuming you did, wouldn't it be like having a pair of undelayed line arrays that aim diagonally across the room? I'm not sure what the unique benefit of that would be?
One-Digital (Pioneer) and Yamaha are doing interesting things with DSP delayed arrays of drivers, but it is more for firing narrow angle beams of sound off the side wall to create virtual remote speakers (this might interest you).
David S.
The ideal stereo speaker pattern is difficult to describe, each of us likes what we create in our listening rooms, but I have come to like omni-directional systems. My latest is the "Castle System" featured in the September Audio Xpress. It has low distortion and a very wide "concert hall' dispersion. It is also surprisingly good on solo performances.
I like what and how you said what you said here, but I'm curious about the high end frequency limit on timing cue perception. You stated 800HZ as the approx limit. I lean toward believing it's more like 1.8kHZ. My theory, which I have not proven experimentally, is that timing info is perhaps the primary imaging cue up to the frequency where the distance between the eardrums is more than a half wavelength, because above that frequency, the ear-brain mechanism has no way of knowing which period it's comparing. I chose half wavelenth instead of whole because I suspect that the ear-brain mech is effectively comparing zero crossings. Perhaps the transition from timing to amplitude and differential HRTF frequecy response variations is actually quite gradual. I could be wrong. Comments encouraged.
"It turned out that the Immersive Audio Lab at U.S.C.’s Viterbi School of Engineering is dark, a bit dingy, and only 30 feet wide, 45 feet long and 14 feet high."
I would be happy (and getting better results) if I had a living room that size too!
Now lets see them get the same results in a room 11 feet wide by 15 feet long by 8 feet tall, like mine....
A room this size is a much bigger challenge to get good acoustics in, IMHO, and leaves very limited options for placement.
I too was a bit challenged by the size of their room at USC. Since most reflective surfaces would be much further away from the listener and/or the speakers, the amplitudes of the reflections would be much lower in dB, but the time delays involved would be in the range of 30mS to 150mS, and therefore be particularly detrimental to speech intelligibility. Not at all like the listening rooms in the average home.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.