Finally got round to listening to both of those. Fun and educational. What could be better with a beer on friday night.
Does make me realise that what had until now been the best hifi demo I have heard was at least partly faked. Ah well, couldn't afford it anyway.
Yes, I heard a system with Audio Research Reference 600s a couple of times - this was capable of being viciously brutal to the eardrums, just plain mediocre, or inspiringly musical - just adding money to the picture doesn't guarantee a good outcome, every time ...
Yup.
But for some people, their ego and vanity won't allow them to come to terms with their humanity.
se
Think i'd prefer the hovering mozzie Have you tried having your ears syringed? I'm told it's like getting your whole system upgraded, might be worth a shot
Hi Fas
I think your spot on but would add that in addition to reproducing what is there and all the pitfalls in that chain, a big problem loudspeakers have is they produce “free sound” in addition to what and when the signal would tell it.
That could be say harmonic distortion or noise BUT one property that wrecks the disappearing speaker act when it measures well is that it radiates an interference pattern or rather what arrives at each ear is enough different for your brain to localize that perturbation in physical space as the source. This is an acoustic effect that happens (often) very shortly after the sound leaves the transducer.
Things that cause this are often found acting as second sources of radiation at a different point in space. As easy to imagine one is a tweeter where there is a step in the mounting. The sound radiating away from a small source should be thought of as if it were in a horn, a step in the area and so acoustic impedance in either a flat baffle or horn can cause a diffraction or radiation as if it were a weak source, delayed in time which radiates producing a weak combing, an interference pattern. Later in time comes the baffle edge radiations if present and near wall reflections, none of which are part of the electrical signal but all arrive at your ears.
Conversely, if what reaches each ear is identical, you hear direction but not the distance cue that the type of loudspeaker radiation had produced. On top of all the other stuff, there is how a loudspeaker radiates and partly why a small full range driver on a flat baffle (which radiates simply) may measure poorly but image wonderfully.
Hi George,
I would buy you a beer just to hear how you came to that 13mm figure. I could also tell you what I am doing haha. Seriously, the object of that system is to pick up the sound from a single point in space and derive the vector for each channel. It is not a binaural recording or dummy head etc. That train recording was 3 generations ago and I remember the mosquitoes were horrid setting that up in the way back, ideally it should have looked like one point in space but …….. Unfortunately it is or was then a klugy setup that looked weird and begged to be tipped over so it limited my choice of things to record. For that parade recording in Deerfield il, i had it about 75 feet back from the road for that reason.
Since one can produce that “between the speaker phantom image” on the right loudspeakers, the goal is to capture and reproduce another environment, much like I heard at Don Davis’s farm long ago but reproducible for a large crowd.
A sort of step one very large scale audio hallucination is being installed at a happy place in Fla now, I hope to hear it if operational by the Infocom trade show in June or if they will demo it for me at night.
What kind of bike do you ride? I just started again after a long hiatus (well actually not now it’s winter here haha).
Best,
Tom
I think your spot on but would add that in addition to reproducing what is there and all the pitfalls in that chain, a big problem loudspeakers have is they produce “free sound” in addition to what and when the signal would tell it.
That could be say harmonic distortion or noise BUT one property that wrecks the disappearing speaker act when it measures well is that it radiates an interference pattern or rather what arrives at each ear is enough different for your brain to localize that perturbation in physical space as the source. This is an acoustic effect that happens (often) very shortly after the sound leaves the transducer.
Things that cause this are often found acting as second sources of radiation at a different point in space. As easy to imagine one is a tweeter where there is a step in the mounting. The sound radiating away from a small source should be thought of as if it were in a horn, a step in the area and so acoustic impedance in either a flat baffle or horn can cause a diffraction or radiation as if it were a weak source, delayed in time which radiates producing a weak combing, an interference pattern. Later in time comes the baffle edge radiations if present and near wall reflections, none of which are part of the electrical signal but all arrive at your ears.
Conversely, if what reaches each ear is identical, you hear direction but not the distance cue that the type of loudspeaker radiation had produced. On top of all the other stuff, there is how a loudspeaker radiates and partly why a small full range driver on a flat baffle (which radiates simply) may measure poorly but image wonderfully.
Hi George,
I would buy you a beer just to hear how you came to that 13mm figure. I could also tell you what I am doing haha. Seriously, the object of that system is to pick up the sound from a single point in space and derive the vector for each channel. It is not a binaural recording or dummy head etc. That train recording was 3 generations ago and I remember the mosquitoes were horrid setting that up in the way back, ideally it should have looked like one point in space but …….. Unfortunately it is or was then a klugy setup that looked weird and begged to be tipped over so it limited my choice of things to record. For that parade recording in Deerfield il, i had it about 75 feet back from the road for that reason.
Since one can produce that “between the speaker phantom image” on the right loudspeakers, the goal is to capture and reproduce another environment, much like I heard at Don Davis’s farm long ago but reproducible for a large crowd.
A sort of step one very large scale audio hallucination is being installed at a happy place in Fla now, I hope to hear it if operational by the Infocom trade show in June or if they will demo it for me at night.
What kind of bike do you ride? I just started again after a long hiatus (well actually not now it’s winter here haha).
Best,
Tom
There's an online test somewhere, i posted a link a while ago, after level matching i couldn't hear much above 14k and nothing above 16.
Tom, that all makes excellent sense, and I agree that those factors must influence whether the effect occurs or not, in any particular situation. But the remarkable thing, in my experience, is that the quality of the reproduction can still override those confounding aspects - I have a speaker with that very step in the tweeter assembly you mention, the Technics SB-5000 - good angle shot here, VNAV • Xem ch? ?? - Nh? các bác t? v?n v? loa Technics SB-5000 - and they are able to do the trick beautifully even with one's head hovering around the step area!
The transducer must be able be pump out the frequency not too far down in level of course - I have a test CD that I put on repeat replaying those high frequencies, and you can hear the tone kick in as it starts to repeat. With the highest stuff one is aware that it is there, or not there - rather than "hearing" it ...
How does that work? Are they presupposing that the transducers you're using are flat up to the frequency limit of the test?
se
With regard to localisation of sounds - whether it is learnt or comes naturally,
False dichotomy that's antithetical to real understanding.
Good question. I remember a funny mandatory visit to a doctor of Occupational medicine. He made audibility tests at high frequencies with headphones that I knew for being unable to reproduce them.
It's a toy at best no matter what your transducers are capable of. If i remember right you set the volume level at various frequency steps by ear so it sounds aproximately level across the given range. Then you listen to a tone up through the frequency steps till you cant hear it anymore. Thats then supposably the maximum frequency you can hear anything at. The test also gives you a minimum but i wouldn't bet much on that result either.
For those who are into "vintage" speakers - the big daddy version of my Technics, the SB-9500 ...
Technics SB-9500 ??? fff?ppp???? 5??? 7????? - ??????????KENRICK SOUND? - JBL????? 43XX???? ?? ???????????
122dB SPL, 190kg ...
Technics SB-9500 ??? fff?ppp???? 5??? 7????? - ??????????KENRICK SOUND? - JBL????? 43XX???? ?? ???????????
122dB SPL, 190kg ...
We were tested in school, that was the first time i came across Stax headphones, impressive sounding ugly cans.
For those who are into "vintage" speakers - the big daddy version of my Technics, the SB-9500 ...
Technics SB-9500 ??? fff?ppp???? 5??? 7????? - ??????????KENRICK SOUND? - JBL????? 43XX???? ?? ???????????
122dB SPL, 190kg ...
PA kit with a fancy finish, you might need to buy a castle to get the best from them
Yes but notice they use a cone tweeter which if were a 4 inch, would narrow to about 90 degree angle at around 3Khz and halve the angle each octave upward to breakup mode. That directivity would reduce the reflection that a small dome tweeter would produce mounted that way. Not just that but directivity would reduce what bounces off your side walls, also a good thing.
Part B, note your tweeter is set to the rear. In the "named crossover types above first order, while the amplitude sums flat, the phase has an "all pass" phase behavior where the low frequency end is behind (n order times 90 degrees) the high frequency end. Loudspeaker are what they call dispersive in time, different frequencies produced at different times, as if from different distances as Dick Heyser described it.
It may be that the design in your speaker accommodated that normal crossover phase rotation (rarely done and not easy) such that on axis, it time the sum is more like or even like a single driver.
For fun though, grab a pair of thick socks or 2 towels, listen to your most revealing musical passage. Then place one sock or folded towel flat on each shelf under the tweeter and see if there is a detectable change.
DNR again. These plots are for -80db full scale at 24bits and 32bit floating point. Remember this is just the DNR possibility of a particular numeric representation, a purely math problem. The second undithered, yes the numerical noise floor of 64bit double precision math is -300dB or so. The first is dithered at 1 LSB with a triangular probability distribution, what you actually can save as a 24bit file. I still see nothing that indicates more bits has any benefit at all.
Attachments
- Status
- Not open for further replies.