Further this discussion from the other day:
Alan Ratcliffe wrote:
Paul E wrote:
Can you elaborate on ocataving one of the signals? Would this be in one take having 2 octaves?
Yeah, if you have two or more pickup systems (or mics) with separate outputs, it's pretty easy to put one through an octaver/pitch shifter and drop it down a full octave, then you mix the two together and you get an extended low end.
Bonus is that if you do this with a magnetic pickup it ends up with something like a fat, round electric bass tone, if you do it with the piezo, you get a snappier almost plectrum bass tone. Drive the octave signal through a bass amp (or bass amp sim/modeller) for some real backside booting bass. Add distortion for crunchy rock bass, etc., etc., etc... Take a modelled nylon down an octave and you get an almost upright fatness and roundness. Best of all, unlike synth basses, they respond exactly to how you play - dig in and the sound gets brighter, snap or slap the string and the sound changes in kind, and all with no delay (or only very little if you use a digital pitch shifter).
I'll often filter out the highs of the octave signal with a lowpass filter, then it becomes like a sub frequency generator that basically adds extended low end to the sound without sounding like a second, bass guitar and has less effect on the higher strings than the lower. If the lows get muddy, I'll add in a high pass filter to the normal sound (at around the same cutoff point as the lowpass on the octave) to filter out the common frequencies.
Taking it even further, I often use the "12-string" part of my GR-55 to split the strings - shifting the bottom three down an octave, while shifting the top three
up an octave. Then it sounds like a bass on the low end and a mandolin on the upper end. Or the same on the low but just shift the top three a few cent sharp or flat (usually alternating) to fatten up the high strings when mixed with the original sound and without obvious detuning/chorus effects. If you're interested, I can capture a few examples for you...
Of course, these all make for wonderfully massive sounds, but are best suited to solo playing. They can be too big in a band mix.
OK... This can apply to live or recording. Recording you might do more "in the box", while live you'd process as much as possible before going to the amps or desk.
Take any guitar with two outputs - can be a magnetic + piezo outputs, can be a hex output + summed output (as in the examples). If you don't have two outputs, you can even simply split the signal so you have two outputs. If recording, it could be just a cloned track. Live you can even use a bass amp for the octave signal for a real big low end.
Here is a recording of a syncopated guitar part. This is just the combined piezo (jack) output from my Carvin nylon:
http://ratcliffe.co.za/music/octave/octave1.mp3
Now I take the second output/track and drop it down an octave with a pitch shifter, which should sound something like this:
http://ratcliffe.co.za/music/octave/octave2.mp3
Note that this is best done with a digital pitch shifter or harmoniser - many analog octave pedals can only process one note at a time, getting confused with two or more notes. These analog octavers also produce a very fuzzy square wave. Also note that modern digital octavers are just specialised pitch shifters, so they will work well.
Now blend the two together and you get something like this:
http://ratcliffe.co.za/music/octave/octave3.mp3
In the third example I've used a couple of tricks to blend the two nicely:
- Highpass (AKA low cut) filter on the guitar sound and lowpass (high cut) on the octave at similar cutoff frequencies. This prevents any overlapping frequencies to muddy up the sound. If you're playing into a desk with only highpass filters, just turn the highpass on for the natural sound and cut the highs on the octave with the shelving EQ.
- The octave sound is compressed a little (4:1 ratio), but the natural sound is not. This keeps the bass at a more even level, so it still comes through clearly when playing softer and does not overpower when I'm beating the hell out of the strings (last and second last repetitions of the sequence respectively).
Finally, another take of the same thing, but using the guitar modelling from my GR-55 and using the "12-string" setting to lower the pitch of the three bass strings only, while leaving the trebles untouched for more distinct separation between the two and for more natural sound on single note runs on the treble strings:
http://ratcliffe.co.za/music/octave/octave4.mp3
More advanced examples with GR-55 modelling tricks soon...
[EDIT: 17.05.14]
Taking the last example a step further, using the 12-string shifting, here I shift the bottom three strings down an octave and shift the first three strings up an octave. Using an acoustic steel patch, you get something that resembles a mandolin on the high strings while still keeping the low, bass voice:
http://ratcliffe.co.za/music/octave/megaguitar.mp3
Using the 12 string shifting is a way around the GR-55's single modelling voice. It's a bit of a compromise as both "instruments" share the same signal path, so I can't be as creative with models, filters, panning or FX. The VG series modellers can do multiple voices, so it's possible with them (I've considered adding in a VG-99 or a second GR-55, but then things might get a bit silly[-er]).
