hypothetically

[noddin0ff]

Hypothetically, if someone made a CD where the left channel and right channel information was exactly the same _except_ that one was coded exactly out of phase with the other...

Could one then play the CD, have different interconnects for left and right, then at the end of the interconnects merge them with a y adaptor and see if a signal comes out. My thinking being if the interconnects were identically transparent you'd get perfect cancelation. If they differed then you'd get a signal which you could feed to an amp and listen to. Then you'd hear what color one might be contributing to the signal.

Just curious.

[superpanavision70mm]

Try it out! Sometimes that's the way the best ideas are born and implemented. I'd be curious on your findings.

[noddin0ff]

ehem, hypothetically!

[Smokey]

Could one then play the CD, have different interconnects for left and right, then at the end of the interconnects merge them with a y adaptor and see if a signal comes out. My thinking being if the interconnects were identically transparent you'd get perfect cancelation.

This test only show how two cables are transparent to each other, rather than the system. They might have the worst IC specifications in the world

[GMichael]

Hypothetically, you should get no sound. But in the real world it doesn't come out the same. Just use a Y splitter sending the same music to a pair of identical speakers and switch the polarity of one of them. Thin and lifeless, but still sound. Maybe that's because you can never have two identical speakers.

[noddin0ff]

In this case, the Y-splitter would be to merge to signals into one path where they could cancel out. Then the one path would go to a mono amp and one speaker.

The point would be to see how reletively transparent the interconnects were to each other. The $2 vs $200 interconnect comparison, with the rest of the path the same.

(L) + PHASE >------IC1---------<\\
. . . . . . . . . . . . . . . . . . . . . . . ------------- mono amp >-----> Speaker.
(R) - PHASE >------IC2----------

That little diagram probably wont come out, but what the heck...

[Swerd]

There might be computer software that lets you do this digitally without inverting the phase of the two identical signals. Imagine two frequency response curves, recorded through interconnects A or B. Have the software subtract one curve from the other and display a plot of the difference.

Frequency response curves are only a suggestion, other types of response display such as waterfall plots or impulse responses may work better at this. Then again, it is quite possible that no difference will be detected.

[noddin0ff]

I was just trying to think of the simplest, most analog way someone could actually 'hear' the difference with. I don't think there would be a difference, or at least, I'm not an expensive wire believer. But if someone wanted to know if they should spend big money on cables, this could be an easy way to determine that, and to learn what to listen for...if there was a difference. Especially for the subjectivists who don't believe computers and data plots can capture the experience of listening. The catch, at least to me, would be making a CD with + and - tracks. Once the CD existed, anyone could conceivably do this on their system just by downloading a file and buying a Y connector. Would it work though?

[Swerd]

It's actually not a bad idea. I can't imagine what it might sound like if there were a difference. I would still like to see both a "difference spectrum" by subtractive graphic plots and hear differences due to phase cancellation to see if anything correlates.

I don't think it will convince anyone - yeasayer or naysayer. They have different belief systems and speak different languages.

[kexodusc]

I don't think the channel being out of phase with its opposite channel is all there is too it. If your speakers are out of phase a bit you wouldn't get total cancellation. And speakers aren't in perfect phase for the whole spectrum.

I could be wrong, but it would seem to me total system phase would have to be perfectly opposite from the left to right channels. Though I would expect output to be greatly reduced.

[Geoffcin]

Hypothetically, if someone made a CD where the left channel and right channel information was exactly the same _except_ that one was coded exactly out of phase with the other...

Could one then play the CD, have different interconnects for left and right, then at the end of the interconnects merge them with a y adaptor and see if a signal comes out. My thinking being if the interconnects were identically transparent you'd get perfect cancelation. If they differed then you'd get a signal which you could feed to an amp and listen to. Then you'd hear what color one might be contributing to the signal.

Just curious.

One of the major advancements in electrical systems is the negitive feedback loop. It's used to identify distortion, and by way of feedback, negate some if it.

[noddin0ff]

I don't think the channel being out of phase with its opposite channel is all there is too it. If your speakers are out of phase a bit you wouldn't get total cancellation. And speakers aren't in perfect phase for the whole spectrum.

I could be wrong, but it would seem to me total system phase would have to be perfectly opposite from the left to right channels. Though I would expect output to be greatly reduced.

I'm thinking that since all the cancellation would happen at the speed of electrons, before anyone needs to deal with speeds of paper cones, that the only way they would mis-phase is if there was a delay difference between the L and R channel on the decoding to analog end.

[kexodusc]

I'm thinking that since all the cancellation would happen at the speed of electrons, before anyone needs to deal with speeds of paper cones, that the only way they would mis-phase is if there was a delay difference between the L and R channel on the decoding to analog end.

Well, in a computer you'd probably be right. But that electron speed is going to terminate at the speakers in a system...and the phase properties of speakers are determined as much by the circuitry of the crossovers through which the electrons flow (at a constant speed for all intents and purposes) as it is the speed of the cones.

The speakers are going to move...the sound waves will be made...as you move off of axis or out of phase even 1%, you'll lose cancellation...In a room the noise would bounce around and you'd still hear something, probably just reduced in level.

It's like setting 2 speakers back to back - in theory they should cancel out and nothing be heard - in practice that doesn't happen.

[noddin0ff]

Well, in a computer you'd probably be right. But that electron speed is going to terminate at the speakers in a system...and the phase properties of speakers are determined as much by the circuitry of the crossovers through which the electrons flow (at a constant speed for all intents and purposes) as it is the speed of the cones.

The speakers are going to move...the sound waves will be made...as you move off of axis or out of phase even 1%, you'll lose cancellation...In a room the noise would bounce around and you'd still hear something, probably just reduced in level.

It's like setting 2 speakers back to back - in theory they should cancel out and nothing be heard - in practice that doesn't happen.

I'm still not getting your argument. I think I didn't explain the set up well, or I am dense. In this hypothetical arrangement, the interconnects for L and R coming out from the CD player will end in a Y adapter and merge into one (cancelled) path. The single merged path goes to the Amp and then to one speaker. The point being that IF there is any signal at all, its a signal that represents the differential capabilities of the interconnects (or differential DAC/preamp processing). In a perfect ideal system you would hear absolutely nothing. If the interconnects didn't match, you'd hear something. You could have the left channel of Beethoven's 9th on the left and the inverted phase of the left channel put over into the right channel. In the ideal system you'd only hear silence.

[kexodusc]

I'm still not getting your argument. I think I didn't explain the set up well, or I am dense. In this hypothetical arrangement, the interconnects for L and R coming out from the CD player will end in a Y adapter and merge into one (cancelled) path. The single merged path goes to the Amp and then to one speaker. The point being that IF there is any signal at all, its a signal that represents the differential capabilities of the interconnects (or differential DAC/preamp processing). In a perfect ideal system you would hear absolutely nothing. If the interconnects didn't match, you'd hear something. You could have the left channel of Beethoven's 9th on the left and the inverted phase of the left channel put over into the right channel. In the ideal system you'd only hear silence.

LOL...my bad...I did read your post, but somehow selected out the part of "one" speaker..
I'll go crawl back under my rock now....

[Kaboom]

here's the file. both channels are left, and the right one is phase inverted.
through two speakers it sounds as if the sound was coming from everywhere. impossible to locate a center or to pinpoint any instruments. have fun


http://www.yousendit.com/transfer.ph...473D86507AB7D8

[noddin0ff]

Hey, Cool Kaboom! Now who out there has a real cheap and a real expensive IC to do the experiment for us?!

[Kaboom]

screw that! who has a y-splitter and two identical interconnects to test the actual theory?

[noddin0ff]

screw that! who has a y-splitter and two identical interconnects to test the actual theory?

Good point. We do need some rigour here. I would hope that a pair of expensive interconnects would be identical...