Tube Amps / Music Electronics
|For current discussions, please visit Music Electronics Forum.|
|7/7/2000 12:17 AM|
||Re: Would you add Blumlein to that list?|
"Your brain can't localise a negative-phase signal in the stereo field, so it ends up sounding like it's coming from just about anywhere, off to one side, behind you, whatever."
This is largely why the use of time delay made such a dramatic improvement. To simply use strict phase inversion results in a constant phase difference between the channels. But if what we are atempting to simulate is an *arrival time* difference, then the actual phase difference should *vary* with frequency, not be constant.
For example, if I have a 1msec delayed version of a signal fed to the other channel, a 1khz tone will arrive one complete cycle (360 degrees) later in the other channel, but a 2khz tone will be shifted by twice as much (720 degrees), and a 500hz tone shifted only a half cycle (180 degrees).
Although it is true that we *can* detect phase differences, you have to remember that there are limits to how fast nerve cells can fire (generally under 100hz in many instances), so directly detecting phase (rather than time) differences would likely be confined to relatively low frequencies. Being able to split the difference between a 12khz tone arriving in one ear a half cycle later than the other is unlikely to happen since it presupposes a potential neural firing rate and conduction speed which is generally impossible in the equipment we expect it from.
A time-delayed signal in the desired range permits one to use both phase difference cues in the low end, and arrival time differences in the upper registers. This is probably why time-based imaging devices seem to make more audible "sense" than strictly phase delayed attempts.
|And now, a word from our sponsors:
|7/7/2000 10:17 AM|
hmm... I suppose you have a point, after all I did sit with one of my hi-fi speakers hooked up backwards for several months and never noticed %-P I thought it was just bad acoustics in the odd-shaped room.
|7/11/2000 12:46 AM|
I sure wish I had George Martin's email address! (Eddie Kramer's too!)
Sir George did a lot of cool stuff with phasing, flanging, and "3D" sound on the Beatles' records. I bet he could give some useful insight.
For example: In "Nowhere Man", there's a tamborine that sounds like it's behind your head if you have headphones on, or if you're in a good stereo room.... Spooky.
Kramer did a lot of neat stuff on Hendrix' records with 3D sound.
For example: in "EXP", the feedback guitar "spaceship noises" sound like they're looping in a huge circle that extends behind the speakers and the listener, and a good way beyond each side of the stereo field (based on speaker placement). I sure would like to know how he did it....
|7/4/2000 4:30 PM|
I'm going to have to try and get a hold of one of these do-hickeys
|7/4/2000 7:13 PM|
If Maplin/Velleman doesn't have one in kit form, I'd be surprised. Radio Shack had a graphic EQ with a one-button-one-knob version of this effect (located on the right hand side) a couple years back. I see these units repeatedly in second hand stores and pawn shops for under $25. RS called the effect IMX (for IMage Xpander, I gather).
|7/5/2000 12:20 PM|
I just had a cool thought. I don't need any hardware to try the effect, I can do it with an audio editor on my PC. All I need to do is make a copy of a stereo file, swap the channels, low-pass filter it, and then mix it with the original file, moving it slightly to create the delay. With today's computing power I can easily do it to a whole track off a CD.
Could you suggest some ballpark parameters for level, filter frequency and delay time to get me started?
|7/5/2000 7:00 PM|
Typically, you are not looking at a very big delay at all. Likely less than 3-8msec, with no recirculation.
As for lowpass turnover, I think you want something that takes out the definition. If you can do it, a 3db/oct rolloff from about 2k up, and a steep rolloff (18db/oct) after about 6k or so seems like it ought to work. I'm just guessing here, though.
You may also want to roll off the low end as well, since low end normally tends to be relatively spatially nonspecific (which is why single subwoofers can exert little impact on stereo imaging). So I'm think that the ideal passband should be stuff in the 150hz to 4khz range...for starters.
For level, you probably want the delayed crossfeed about 6db down.
Again, all these values are second guesses, but they should give you at least a reasonably valid starting point.
Just out of curiousity, such units or algorithms usually have a fixed delay. Given that the reflecting surface at live venues fluctuates (i.e., moving people), I wonder if making the delay vary ever so slightly will give a "live-er" feel? So instead of processing it through a fixed delay, patch it through a chorus algorithm, tweaked for a bit of variation, say about +/- 1-2msec off whatever the basic value is (e.g., from 6-9msec, with basic delay of 8msec).
A lot to tweak here.
|<<First Page||<Prev||Page 8 of 15||Next>||Last Page>>|