Hi
Dick was a brilliant man, I owe him much and now having gone through some of his unpublished work, there are many questions I would ask him if it were only possible.
I met him twice, once at JPL when I didn’t know who he was and again with Don and Carolyn Davis who had given me an excuse to leave the tradeshow booth at AES and go with.
I could only squeak out a few words that night at dinner, I was terribly aware of being in the presence of giants like Dick, Gene P, Don and others and me being about an inch tall.A friend was in charge of the Heyser library and it was for me a hair standing up on the back of my neck experience to go through that room and look at his test equipment and some of his unpublished work.
It was nice too eventually we were able to hire him and this month went on full time.
http://svconline.com/proav/acoustics_danley-sound-taps-top-acoustician/ In reading thorough his work, it is clear he was big on seeing things interchangeably from one perspective to another, much like how mag&phase are one view of the event while the impulse response is another view of the same exact event.
Personally I believe Dick has gotten less credit than he deserves.
In another form I explained what I saw / see starting with post #2195, #2200, #2209 here;
http://www.diyaudio.com/forums/multi-way/195124-what-ideal-directivity-pattern-stereo-speakers-220.html Anyway, to me there are two phases of the problem, the “capture” and the “reproduction” and a “real” image requires both be correct.
He was correct in what he was saying too but a hologram is not necessary as compared to light, we are working with TINY dimensions relative to the wavelengths. SO many things depend on the acoustic size. We hear in 3d BUT we are totally unaware of that process or how different that is compared to a perfect microphone. In Don Davis’s “in the ear” recordings shows with amazing clarity, if you measure what is inside your ear, you find the response has all kinds of changes and comb filtering depending on the incoming angle etc. Our ears are nothing like “flat” and everything about them would appear to be a”flaw” if see through an measurement eye.
We hear NONE of those things, rather, this is all we have ever known and so instead of hearing the comb filtering, inter ear delay and all that stuff as “flaws” we hear what direction the sound is coming from, how high it is and how far away it is, 3D perception from a two point system which normally could only resolve one axis, with processing and direction dependent errors, we hear in 3D. Playing with these effects can trick your ears to a degree but never as strong as the real thing. To me, the reproduction side was he side where the “light went on” for me about 12 years ago.
As the full range horns at work got to be more and more like one driver a weird thing happened.
Playing a voice though one speaker, it got to be harder to tell exactly how far away the speaker was when my eyes were closed. It was still easy to hear the direction BUT harder to hear the location in depth. When the tef measurements got to where it looked like one driver, the effect was stronger still. This was an irrelevant thing for 99% of where the horns are used at work however I used them at home and the effect on the stereo image was dramatic.
When a loudspeaker provides a depth position que, then this detract from the stereo image. In that case (as most speaker do) when you play a mono signal which should be a single apparent source dead center, there are two additional sources” you hear in the physical depth of the speakers. If a speaker doesn’t radiate a complex field, then your brain hears the sources less and the phantom more.
Now what kinds of sources do this? A number of hifi speakers have heard radiate only a small source identity, A Quad esl-63, a Manger on a flat baffle and to a degree some small loudspeakers where the drivers are small and close together . A hifi company has recently discovered the audibility of source identity also, google up the KEF blade concept designed around that criteria..
What radiates the identity i think is a complex field, one that provides ample clues in the differences between the right and left ear inputs. Many aspects can cause that too.
If what arrives to the R&L ears from one speaker is identical, there is no source distance information conveyed and if part of a stereo, the sources disappear into the image to a much larger degree. For a mono phantom, you do not want a Right, Left and Center sounding image.
For the skeptical, consider an experiment. Obtain a pair of small fostex full range drivers. These are necessarily limited at either end BUT if mounted on a large flat baffle, radiate as a simple source up to reasonably high frequencies. .
Move these away from the walls in the room and listen to the stereo image, these can be stunning in the depth and “real” feel of it (if present in the recordings). Part B is the capture.
This is where it really falls flat, most recordings are not a capture of then event but a re-creation of one in the studio. You can only hear how far it falls flat when you hear something more realistic. As we hear in 3d but measure from one point in space, one can make a recording of a loudspeaker and “hear” what it sounds like without the 3d brain processing.
We used generation loss recording of speakers in the early days as a reality check. I think people would be amazed, stunned and befuddled how very few generations a loudspeaker can be used in a closed loop audition where a truly faithful device has no limit. A really good microphone can be much more precise than any loudspeaker because of it’s dimensions and power flow amount and direction. If you use a measurement mic to make mono recordings, they can be chilling in their realism and better yet, you were there live so you have a big edge over someone else’s recordings.
The problem is you cannot easily record a live stereo image with two mics, there is no simple combination of spacing etc that “captures” the event and so most of what we hear was created with pan pots etc in the studio. I have an approach to the issue too, while it has been a back burner project, I believe this approach is more like the idea behind a hologram although not being a massively parallel system. This works in a 360 degree capture although that requires 5 channels, 8 to cover the “up” direction too. Pop on some headphones and try a couple of these recordings. Not the most exciting stuff but the setup is cumbersome and ugly right now and for me environmental sounds have the advantage that I know exactly what it sounded like and they are usually available. Also, these have no compression so they sound quiet and the fireworks will severely tax nearly any loudspeakers, is ruined by the mp3 process.
Try with good headphones first, FWIW, these are only the front two channels about the width of your vision.
Recordings at the bottom of this page; http://www.danleysoundlabs.com/technical%20downloads.html Best,
Tom Danley
Danley Sound Labs
|
