The birth of flanging

My late uncle Leo Kulka, who had storied career in sound recording going back to the 1950’s, often claimed that that he invented flanging. As he told it, he was working at Hollywood’s legendary Gold Star studios in 1959, A-B’ing mixdowns on two sync’d tape decks. Instead of switching between the outputs of the decks, he accidentally summed their outputs, producing the “jet plane effect” that we call flanging. Further investigation showed that the effect could be heightened by manipulating the aluminum the reel flanges. Placing a finger on the supply reel created drag, caused the machine to slow down, which increased the delay and lowered the pitch of the notches. The sound could be swept upward by doing the opposite — touching the take up reel and speeding it up slightly. Not surprisingly, the effect got everyone’s attention that day and they decided to utilize it in “The Big Hurt” by Miss Toni Fisher. This was the first recording to use the flanging effect. The second was “Itchycoo Park” by the Small Faces. (Rod Stewart joined the Faces years later, but wasn’t on Itchycoo Park.)

Leo was a wonderful storyteller. Most of the tales weren’t 100% accurate, but usually his audiences didn’t mind, because he was so damned charming and entertaining. Well, a few years ago, I happened to be with Gold Star founders Stan Ross and Dave Gold, and engineer Larry Levine at a breakfast get together, and asked for their version of the story. With apologies to Leo, all agreed on the basic premise of the story — the accidental mixing of two decks that were in sync — but were quite sure that Larry was running the board that day, though Leo, who often engineered there, may have been assisting or working in another room that day. So while the fable of accidentally discovering flanging is probably true, and the 48 year old The Big Hurt is the first song with that cool sound, I don’t think my dear uncle Leo can really take very much of the credit.

What is flanging?

Let’s start with the Doppler effect, which everyone has experienced. You’re standing at a street corner as a car approaches with its horn wailing. As the car approaches the listener, the sound waves strike his ear more and more rapidly, and the pitch goes up. The moment the car passes and begins to move away, the sound waves become “stretched out” and the pitch goes down.

If you had a remote control and could joystick the car back and forth with the horn blowing you’d hear the pitch increase every time the car got closer, and decrease each time it moved away. If you have a Time Modulator and turn the Time Delay knob back and forth you’ll hear the same thing. In fact, it sounds exactly the same. Steve got pots with custom tapers, so that natural sounding doppler and flange effects could be produced by manually sweeping the knob from end to end. (Of course you could do “automatic” flanging by using the VCO, but manual mode was carefully optimized so that the operator could “play” the Delay control in a musical way, timing the sweep to fit the length of a chorus, a solo, etc.)

The day my uncle Leo or Larry Levine pushed the wrong button and discovered flanging, a kind of Doppler signal was produced by the additional deck, which, owing to worn mechanical parts and Murphy’s law, would inevitably have been running at a slightly different speed. The outputs of the two decks were summed, with a little time delay on one deck. The delay caused cancellation at certain frequencies, and Larry or Leo, or whoever could sweep the cancellation — the pitch of the jet plane effect — by letting his finger gently rub the reel flange, and then maybe pushing the flange to speed it up a tad. This produced a swept comb filter effect. The harmonics and overtones were “notched” up and then maybe down. It sounded just like a jet plane taking off from stage left to stage right. The jet plane is a good analogy because it’s sound, a little like pink noise, contains a smorgasbord of frequency components that, combined with Doppler, create a kind of flanging sound.

Itchycoo Park came and went, tape flanging was a neat engineer’s trick, but it tied up two tape machines and was way unwieldy. In 1971, new company Eventide released the PS101 Instant Phaser, a black box that roughly approximated flanging. It’s anemic 20 db null (vs. the 35 db from traditional tape machines flanges) wasn’t too impressive but the real problem was: phasing is just a coarse approximation of flanging. Eventide released the FL201 Instant Flanger in 1976. It was a great improvement over the Instant Phaser. (Was this the first “flanger in a box”?)

Conceptual difference: flanging is about time delay, phasing is about phase shift. Phase shift is simpler and cheaper to implement, but is limited to 180 degrees of delay, and doesn’t sound as good. Unlike phase nulls, flange nulls are harmonically related and sound more musical. For example, 440 Hertz “A” above middle C has harmonics at 880 Hz., 1320, 1760, 2200, etc. With a delay of 136 milliseconds the fundamental and all odd harmonics will cancel out, leaving only the even harmonics. Compared to phasing, flanging produces more frequency nulls. Flanging is more difficult to achieve, but is more pronounced and natural sounding.

Of course the MTM was a flanger, not a phaser. It was distinguished by the extreme depth of its nulls — up to 85 db, far more than tape deck flanging could accomplish. This was accomplished by creating the effect within a dbx type II compression/expansion loop, which doubled the intensity of the nulls, producing an amazing sound that had not been heard before.

Positive, negative, and through zero flanging

The effect that Larry Levine or uncle Leo discovered and used in The Big Hurt was positive flanging. The dry and delayed signals were combined in phase. With very short time differences between the two decks, cancellations occurred at notches in the higher frequencies, producing the jet plane effect.

The guys at Gold Star could have produced negative flanging, a much different effect, simply by reversing phase on one of the decks. (I wonder if they tried it?) With this set up you find that at the “top” of the flange, when delay between the two signals is minimal and only the very highest frequencies are notched all the lower frequencies cancel out, revealing the flanged high frequency sounds in their naked glory.

With through zero flanging, the outputs of two flange circuits are combined out of phase, allowing momentary total cancellation (no sound at all) as the sweep progresses. (I have never heard this effect but I checked the AMS DM2-20 manual and the unit, which has two discrete sections that can be combined antiphase, will indeed produce this effect. I believe that Steven St. Croix’s product information states that his second generation MTM, the 5402, does produce this effect. We have a few in the shop but I’ve decided not to explore that unit until I know the 5002 inside out, so I can’t yet confirm whether it does TZF.

The Time Modulator produces other amazing and baffling effects. The arpeggio effect is very trippy. Pitch quantizing, which gives vocal sounds a metallic constant pitch (with sibilants and fricatives preserved) is mind blowing; it’s hard to imagine that a little bit of delay and modulation can produce such a profound effect. We plan to build and fix lots of Time Modulators in the coming years, so I should have plenty of time to learn the nuances. Again, we’re very thankful to Steve’s wife for choosing us to take this on. It’s important, fascinating, and certainly a lot of fun.