Emo-Fan
09-04-2010, 04:51 PM
Greetings, all!
I thought we could get a discussion going about crosstalk and channel separation.
Some years ago, I purchased Rhapsodies, with Leopold Stokowski conducting (RCA Victor 09026-61503-2), and was very disappointed with the sound: It came across like there were separate recordings coming out of each speaker. There was a hole in the middle; the recording sounded edgy and boxy, like the musicians were locked inside the speakers.** One day, I had what I thought was a brainwave: I ran my CD player into my mixer and gave each of the two channels its own track, then turned the pan pots far left and right respectively. It increased the channel separation, but sounded even more edgy. so I turned the pan pots a bit toward each other, thus decreasing the channel separation but removing the so-called hole-in-the-middle.
My observation is that, if this had been an LP played with a really decent phono cartridge, the channel separation values would be--What?--around 25 to 35 dB, right? Yet when the original master tapes were transferred to CD, the resulting channel separation is over 100dB.
My premise, then, is that ONE reason for digital edginess that audiophiles abhor so much must be channel separation.
For synth music, when you oftentimes NEED a hard, edgy sound, what is supposed to be in the left channel stays there, and likewise for the right. But at a symphony, you don't hear the first violins only with your left ear and the contrabasses only with your right--you hear the whole orchestra with both ears. You're awash with sound. The recording should reflect that listening experience.
I'm prompted to write this thread because on my Hammond XK3c organ, there is variable crosstalk to correspond to older Hammond organs that had so-called "leakage", that is, the sounds from one drawbar "leaked" into the sounds from another (analogue technology of the 1950s). Older organs had that "warmth" that newer synth organs never had, not because the technology wasn't good enough, but ironically because it was TOO good: No leakage, no crosstalk. Music sounded TOO clean, TOO...well...antiseptic.
I realize that channel separation isn't the same as crosstalk, of course, but both phenomena involve "leakage" of audio material getting into someplace where it doesn't belong. My point is that crosstalk and low channel separation values aren't necessarily bad things.
Any thoughts out there?
**A newer CD player with upsampling solved this problem with higher rez. The recording sounds great with both good channel separation and loss of the hole-in-the-middle, but I still contend that lower channel separation is closer to what the engineers in 1960 and 1961 had to work with, at least with regard to the then-current state-of-the-art technology.
I thought we could get a discussion going about crosstalk and channel separation.
Some years ago, I purchased Rhapsodies, with Leopold Stokowski conducting (RCA Victor 09026-61503-2), and was very disappointed with the sound: It came across like there were separate recordings coming out of each speaker. There was a hole in the middle; the recording sounded edgy and boxy, like the musicians were locked inside the speakers.** One day, I had what I thought was a brainwave: I ran my CD player into my mixer and gave each of the two channels its own track, then turned the pan pots far left and right respectively. It increased the channel separation, but sounded even more edgy. so I turned the pan pots a bit toward each other, thus decreasing the channel separation but removing the so-called hole-in-the-middle.
My observation is that, if this had been an LP played with a really decent phono cartridge, the channel separation values would be--What?--around 25 to 35 dB, right? Yet when the original master tapes were transferred to CD, the resulting channel separation is over 100dB.
My premise, then, is that ONE reason for digital edginess that audiophiles abhor so much must be channel separation.
For synth music, when you oftentimes NEED a hard, edgy sound, what is supposed to be in the left channel stays there, and likewise for the right. But at a symphony, you don't hear the first violins only with your left ear and the contrabasses only with your right--you hear the whole orchestra with both ears. You're awash with sound. The recording should reflect that listening experience.
I'm prompted to write this thread because on my Hammond XK3c organ, there is variable crosstalk to correspond to older Hammond organs that had so-called "leakage", that is, the sounds from one drawbar "leaked" into the sounds from another (analogue technology of the 1950s). Older organs had that "warmth" that newer synth organs never had, not because the technology wasn't good enough, but ironically because it was TOO good: No leakage, no crosstalk. Music sounded TOO clean, TOO...well...antiseptic.
I realize that channel separation isn't the same as crosstalk, of course, but both phenomena involve "leakage" of audio material getting into someplace where it doesn't belong. My point is that crosstalk and low channel separation values aren't necessarily bad things.
Any thoughts out there?
**A newer CD player with upsampling solved this problem with higher rez. The recording sounds great with both good channel separation and loss of the hole-in-the-middle, but I still contend that lower channel separation is closer to what the engineers in 1960 and 1961 had to work with, at least with regard to the then-current state-of-the-art technology.