Subj: Page's Studio Tricks VI
Hello, all --
Here's what I said in my previous post, Page's Studio Tricks V:
"An interesting characteristic of phase shifting is that if you phase-shift one signal, a non-phase-shifted counterpart sounds as though it's phased, too. You can hear this during the verse section from 2:34 through 3:00. That clean sound off on the right souds phased because of its close proximity to the actually-phased warm sound next to it."
Phase-shifting, in the guitar sense of the term, is a "time-domain" effect. That is, phasing is brought about by using "delay." Remember my bit about the echoes in the Lemon Song? Another name for those echoes is "delay." Delay unto itself is a nice effect, and one you've heard almost every time you've ever listened to a recorded bit of music. Extremely short delay times can also be used to create cool effects, and phase shifting is one of them. Here's how it works (and this is a little complex, so bear with me):
Remember that all sounds are waves of vibration that vibrate at a certain frequency. This frequency can be graphed as a sine wave, which "wiggles" up and down across a center line. For a musical frequency, the distance from the peak of one wiggle to the peak of the next is a constant. This distance is measurable in actual units (e.g. inches), although I don't know what those numbers are. What that distance is isn't important, but the concept that it *is* a distance is important. Got me? Okay, all sounds, regardless of frequency, travel at a constant speed, something close to 700 miles an hour; again, the actual speed isn't important but the concept is. I'll be getting back to this in a minute.
What *is* important for the moment is that for any given amount of time, a sound is going to cycle through a given number of wiggles; the musical note "A," for example, will cycle 440 times a second. This frequency is called 440 Hertz (appreviated Hz).
This is next bit is crucial, so pay attention. "Phase" refers to the sine wave and it's postion relative to other waves of the same frequency. A sound is "in phase" with another sound if it's "peak" is where the other sound's peak is, and it's "trough" is where the other sound's trough is. A sound is "out of phase" with another sound if it's peak is someplace along the downslope or upslope of the other sound's wave. A sound is "180 degrees out of phase" if it's peak is where the other sound's trough is.
If a sound is coming from a single source (i.e. from one speaker), your ears are going to hear it in phase no matter what, because the waves of sound are reaching you one after the other, like waves reaching the beach. For that matter, your brain hears almost everything "in phase," because your brain is really, really, *really* smart, and it's really good at telling you that sound is happening and what it sounds like, regardless of phase. Plus, your ears aren't in the exact same place and have been hearing sounds out-of-phase since the day you were born, so your brain has gotten a chance to get used to it.
[As an aside, I was once under the influence of a common psychedelic fungus, and in walking past a television I could literally hear the phase relationships of the sound changing as my position relative to the TV changed. Believe me when I tell you that we're all glad the world doesn't sound like that all the time.]
Here's where the fun begins:
Suppose you place two microphones in front of a speaker, and instead of listening to the speaker itself you listen to what the two mikes are "hearing." Physics tells us that the two mikes can't possibly occupy the same space at the same time, so they're going to be hearing the sound waves at slightly different points in the wave's cycles. Remember that the sound coming from the speaker is an amalgamation of many, many frequencies, which by definition have different distances from peak to peak. If our two mikes were in exactly the same place, each would hear exactly the same thing, and all the frequencies would be in phase. But physics already told us that was impossible, so although each of the two mikes is hearing all of the different frequencies, for each specific frequency, while one mike is hearing a peak, the other mike will be hearing a trough, or the downslope of the sine wave, or whatever.
If you combine two signals of the same frequency that are in phase, you hear that frequency at its maximum volume. If you combine two signals of the same frequency that are partially out of phase, they begin to cancel each other out, and you hear that frequency at a reduced volume. If the combined sounds are 180 degrees out of phase with one another, what you hear is silence. Dead, absolute silence. Nada. Nothing. The void. This varying degree of "out of phaseness" is called "phase cancellation."
So, when we combine our two mikes, we're going to be combining all these many different frequencies. Some of them are in phase, some of them are out of phase, and some of them are 180 degrees out of phase. What you've done, essentially, is EQ the sound in an unusual manner, because you've partially or totally removed some of the frequencies that came out of the speaker. "So why not just use an EQ and make life easy?" you ask. Because the sound you've now created is very, very different from the sound you'd have gotten using an EQ, that's why. For proof, listen to the guitar sound on the song Houses of the Holy. That's what phase-cancellation does to an innocent guitar signal. Those righteous solos in IMTOD are heavily phase-cancelled. Page does this all the time, and that you know what you're listening for, you'll hear it. For more examples listen to some of Mick Ronson's guitar tracks with David Bowie. Glorious phase-cancellation.
This "stationary phase cancellation" set-up is sometimes called a "comb filter," although I don't know why. In Star Wars, both Darth Vader and C-3PO's voices were modifed with a comb filter.
Remember that we're creating this phase-cancellation using two stationary microphones. Frequencies are being cancelled, but those frequencies won't change unless one of those mikes is moved. Once you move a mike, a whole different set of frequencies will be cancelled. Why? Because those distances that each wave travels from peak to peak are constants, that's why, and now that you've moved the mike some of those once-cancelled frequencies are back in power and ready to rock. Of course, some of their brothers and sisters have been emasculated, but that's another story. For information of the phase wars, see Star Wars part seven.
Now, suppose you hire a little troll to sit in front of your speaker and gently move one of the two mikes around while the other mike stays in the same place. Your combined signal will now have a whole variety of frequencies coming into phase, going out of phase, coming, going... Like fashion, only faster. And the sound! It's swirling around, sounding like space munchkins playing "chicken" on Tau Ceti. Like the end of "Bold as Love" on Jimi's Axis Bold as Love (which was not made this way, but would still sound like that if it had been). Great! What a cool sound! Except that our poor troll is coughing into the mike and ruining our track, never mind that he's going deaf and demanding health insurance and clamoring for his own set of groupies.
So what are we gonna do? We've got this great sound but we don't want to pay the troll's exorbitant fee.
Well, guess what? Remember I said something about the waves each having a constant length *and* that they travel a constant distance in a given time? Time! That's our answer! Time! See, our two microphones are hearing those waves at different points along the waves because it takes the sound slightly longer to reach the second mike. Unlike light, which travels 186,000 miles a *second* --functionally too fast for your average human to play with -- sound is only going a measly 700 miles an hour. Well within our reach!
The sound takes a measurable additional amount of time to reach the second mike, somewhere between one and ten thousandths of a second, or milliseconds. Okay. Why not build a machine to *delay* the sound for that long? Why not, indeed? Consider it done!
Great. We've now built a machine to delay the sound, thereby cancelling phase, and we've got some neat-o guitar sounds. But what about the swirly sound we dug so much? The troll's long gone -- I heard he's working for Richie Blackmore's Rainbow these days.
Rememeber the moving mike? What was changing was the distance between the mikes, but what caused the difference in sound was the difference in the amount of time it took the sounds to reach the mikes.
Well, why not build a machine to vary the delay time? Consider it done!
If you're still with me, congratulations. You obviously haven't outgrown your thirst for knowledge or your sick desire for torture, whichever you please.
A "phase-shifter," then, is a machine which creates a duplicate of the incoming signal and delays it slightly. This delay is then varied across a spread of delay time, or "modulated." So what you end up with is a delayed signal cancelling a group of frequencies, then changing delay time and cancelling a different group of freqencies, and changing again, and so on. Swoosh, swoosh, swoosh.
Page used phase-shifters quite often: IMTOD, TYG, and Nobody's Fault But Mine come to mind. Now that you know what you're listening for, you'll notice everywhere. Back when I got my first phase shifter, I finally understood how Eddie Van Halen got that amazing swoosh on "Eruption." They're all over disco music.
If you're wondering what one looks like, watch TSRS. xx:xx from the moment Richard Cole opens the barred window thing in Grant's fantasy sequence, you can see a little orange box on the floor in front of Page. That's an MXR Phase 90, IMHO the finest phaser ever made. Phasers, being electronic, come in any shaped case you want.
There are other "time domain" effects that are similar to phasers. A "flanger" is almost identical, but the delay time ranges from 1 to about 20 milliseconds, and additional signal is pumpedback through it (regeneration) for added swoosh. Flangers often make guitars sound like airplanes flying overhead -- which is a perfect example of shifting phase! You've all been on an airplane, right? The engine sound is constant from the airplane's point of view, but from the ground it seems to swoosh. That's changing phase. [The way it drops in pitch is caused by the Doppler effect, which is another topic altogether.]
Flanging at it's finest sounds like the end of "Bold as Love," although, again, they didn't use a "flanger" to create that sound.
"Chorus," the last of the major time-domain effects, uses a much longer delay time -- around 50 milliseconds or so. Again, the delay sound is modulated a few milliseconds back and forth. This longer delay time doesn't cause much phase-cancellation; rather, it sounds more like a second instrument playing the same part. I can't think of any Zep songs that feature chorus, but when I do I'll let you know.
Have a good one,
This Month in