David Gilmour "Money" Tremolo

[vanessa]

I've read that David Gilmour used a "Kepex processor" to get the tremolo effect on "Money". I assume that this is a Valley People Kepex processor. I'm not familiar with these units and all I could come up with is that these are limiters/gate modules? What photos I could find did not seem to show a speed control that could be used to manipulate a LFO. Was there a LFO in these things? If not how do you think they used this effect. Any schematics laying around?

[slacker]

According to this site http://members.aol.com/uniquenyc/key19.htm
"The name Kepex was short for the keyable programable expander."
and
"the "Key Input" allows the gate to follow the dynamics and envelope of the program material that is feed into it"

So I guess he hooked up an LFO, possibly from the VCS synth to the key input. Which would turn the noisegate into a tremolo. They might even of hooked the key input up to a drum track to get a tremolo that was in sync with the rest of the track.

[Mark Hammer]

If I'm "keying in" to what you're after, this is the sound at the start of the tune where the tremolo seems to kick in a bit after the starting strum?  If so, then the essential ingredient is that the risk of a dip in volume at the start of the strum is methodically avoided.  How might this be done with a "keyed" (i.e., sidechain controlled) device?

One means is via ducking, which is likely one of the potential functions of the Kepex.  Imagine I have some sort of splitter.  I run the signal to a tremolo and the ducking unit input.  We'll treat the straight guitar as tantamount to the DJ's voice, and the tremolo version as tantamount to the music.  When the DJ talks, the music disappears, and when the DJ stops talking (depending on the threshold setting), the music fades in, only in this case it is a modified version of the same guitar, which only fades in after the input peak.

Does that make sense?  Does that fit with the actual sound as it appears on disc?  I'm working in the dark from audio memory here.

[d95err]

Could it be that he used a device with external trigger input in order to sync the tremolo with the music somehow?

[Mark Hammer]

Maybe, but you have to consider when the recording was done and what was available commercially atthat time.  Ducking had been available for at least 15 years at that point and was widely used in radio stations.  Heck, I imagine ducking was used when the DJ said "Brand new... from Pink Floyd, it's MONEY! on the state's number 1 rock station!", and the tune receded to the background til the idiot was finished babbling.  Syncing devices via triggering would not have been that widespread for a few years after that probably.

[TELEFUNKON]

maybe that Floyd-band was: "tight" ?

[Doug_H]

According to a gear pager who has one, the tremolo was done with an EMS Synthi Hi-Fli:

http://www.gilmourish.com/?page_id=77

[Mark Hammer]

If you look closely at this picture - http://www.sequencer.de/pix/ems/ems_hifly.jpg - you'll see that some of the modulation options on that second rotary switch from the right include modulation that increases or decreases over time.  So yeah, there is actually a pretty good chance that the Hi-Fly was used for that effect.....although you could do it the way I described.

[petemoore]

  It is quite possible that sliding 2 faders on a mixing console could have found the desired effect, tremolo added to one guitar track during or after recording, the other track 'plain' guitar, both with reverb?. Sliding down the straight guitar while sliding up the tremoloed track.
  I'm guessing musicians have approached designers asking 'can I have just that effect exactly [or close] in a less costly [so I won't have to pay someone to patch a split-trem and whip sliders on a mixer just right every time we play that song]. there may even be a digital patch [somewhere in the digi-domain] which is designed to capture this effect.
  Using a bright side for non-tremolo attack, and a darker tremolo sound [perhaps with compression], the quick fade of the bright side would allow the tremolo to have it's way soon after attack...[or at least that's the sketchy plan], might not be the exact sound, but could be stuffed in a box, and wouldn't require env. detection coupled to an attenuator.
  Then entertain the env. controlling volume swell or drop, on one side chain or the other for a more profound effect...might be pretty nifty to mix a bright attack sound reverbed, the a compressed or even swelled tremolo.

[vanessa]

If you look closely at this picture - http://www.sequencer.de/pix/ems/ems_hifly.jpg - you'll see that some of the modulation options on that second rotary switch from the right include modulation that increases or decreases over time.  So yeah, there is actually a pretty good chance that the Hi-Fly was used for that effect.....although you could do it the way I described.

I can't see a square wave selection in the photo. Just hypothetical but maybe it was used on the scratch takes (demos) only? And they wanted to make the effect more prevalent in the mix and had to use a different manner via the Kepex.

Or maybe better the Hi-Fly was used to create modulation (as the LFO) and ran into a heavily gated Kepex. When the note reached below a certain threshold the Kepex chopped it off. If that was the case the Kepex must have had a very fast attack/release sensitivity.

[Plectrum]

Syncing devices via triggering would not have been that widespread for a few years after that probably.

One issue here is that Floyd (and recording team) were fanatical about the (audio) technology, and IMHO had enough clout and resourcefullness to obtain or emulate thing's few people had yet heard of. I think they had a few "firsts" when it came to some effects/applications.

My guess is the triggered gate... Though:

I know with synth designs, you can sync the waveform start to a logic type trigger.
This is normally done from another oscillator. But I have used it (it's a well known practice) on my Nord Modular to sync the restart to a keyboard input: it can make basses more "punchy" sounding etc.
So if you did this to a trem's LFO, you could do the whole effect inside the trem, by sending it a logic pulse of some kind, at the intended downstroke of the chord (or just before, to have the LFO at it's maximum amplitude for the stroke. (hope I'm not teaching everyone to suck eggs here). Would it be easy to do this electronically?

[Doug_H]

I can't see a square wave selection in the photo.

3rd from the right- there's a selection that looks like a decaying square wave, which seems like it could be the "money" sound.

I thought the keypex stuff had something to do with noise-reduction, gave them fits during the Atom Heart Mother sessions IIRC.

[vanessa]

Magnifying the photo looks as though they are represented as sine and triangle waves. I think each setting is just type of modulation (speed fader) and the decay (modulation ramp) preferences.
Now that I know what the Kepex is. I believe its use on this track was to create a very warm (if you can do that), noise free, faux square wave tremolo effect.