How to convert guitar signal into sine wave ala Ambarchi?

[orangettecoleman]

I've been listening to Oren Ambarchi a lot and I'm trying to figure out how it is he processes his guitar. I know he uses a lot of delays and ring modulation... is ring modulation the way to get that bell/sinewave tone out of a guitar signal? I know Keith Fullerton Whitman did it with a laptop, but that can't be the only way because Ambarchi doesn't use computers at all. Here are samples of what I am talking about:
http://www.last.fm/music/Oren+Ambarchi/ ... ?autostart
http://www.last.fm/music/Oren+Ambarchi/ ... ?autostart

What could I use to get this sound? Is it a ring modulator, or some sort of voltage controlled filter? I am a guitarist and I want to find ways to get a bigger variety of sounds to work with, I feel too tied to recognizable guitar tones.

[flies]

ring mod certainly won't do it, it will add partials. it's bell-like because the partials it adds are not in the harmonic series of the fundamental (generally), like with bells.

i might venture to guess that he's doing it by pitch-shifting the notes down 2 octaves (since i've seen him with a whammy pedal).

if you can find a musical device with a phase-locked loop (PLL) oscillator, this might get you the sound you're after (i thought the moogerfooger freqbox had one but it looks like i was wrong: it has oscillator sync, which, if you're not familiar with it, sounds like it should be what you want but isn't). i also know he uses a ring stinger, which i thought had a pitch following oscillator.

you could just use a guitar synth. sine waves are one thing that even the crappiest digital synths can do really, really well.

[orangettecoleman]

What would be a decent guitar synth to look for if I'm on a budget? really I just would need something that would be able to do sinewaves and different waveforms, none of that sampling nonsense. one of the really old roland gr300 synths would do the trick but those are now kind of expensive secondhand because people want vintage stuff these days....

[Moon]

I think I know how he makes those bell sounds, though it's not a very interesting answer:

Use the pickup selector or a kill switch to mute the guitar, play a note and un-mute the guitar. Use slight overdrive or whatever device to get the sound "warm". Helps to have a little bit of reverb as well, but not too much.

From what I've heard, he uses the ringstinger on the noise washes or as a tremolo, very seldom on the melodies.

Here's an example, skip to the end for louder guitar loop.

Edit: worky link

[orangettecoleman]

I think I know how he makes those bell sounds, though it's not a very interesting answer:

Use the pickup selector or a kill switch to mute the guitar, play a note and un-mute the guitar. Use slight overdrive or whatever device to get the sound "warm". Helps to have a little bit of reverb as well, but not too much.

From what I've heard, he uses the ringstinger on the noise washes or as a tremolo, very seldom on the melodies.

Here's an example, skip to the end for louder guitar loop.

That's a cheap answer. that's my favorite kind of answer. I can put a kill switch pedal together easily, I'll give it a try. Thanks.

[jkudler]

technically speaking, "warming it up" with distortion *adds* harmonic content, whereas a sine wave has no harmonics/overtones. that said, i am not convinced what the original poster wants is actual sine waves. i associate ambarchi's sound with ring mod, definitely. as flies mentioned, "bell-like" usually equals ring mod. why not just use a sine-tone generator to get sine tones? for instance, that ring stinger looks to have an oscillator out.

by the way, none of the audio linked above works, though the aquarius clips to me definitely sound like a pickup switch used to bring the sound in after the initial attack. i also hear a bunch of loop/delay and some pitch-shifting. he's probably using ebow at times too.

-j.

[cheapmachines]

...yep, delay plus pitch shifting is principally what I'm hearing here.

[Moon]

It might be worth to notice that he uses a pickup on the neck of his guitar as well. I've read/heard him explain that it's there to bring out more harmonics. Experiment with that instead of what I said, perhaps?

If you're going to play around with what suggested, make sure to use overdrive and not distortion. The differences in odd and even harmonics makes the difference between a bell and... a distorted bell..

[Red]

Replace the .m3u extensions with .mp3 to get the links to work.

I agree with the others -- I don't think it's very sine wavy, but it does have a prominent fundamental with rolled-off harmonics.

I hear it bell-like in the warm, slightly overdriven, early Fender Rhodes Suitcase sense rather than like a church bell.

!) The attack is very interesting and not very guitar-like. Possibly the kill switch idea would work.

2) The pitch is very stable.

3) The combination of pickup selection, subtle overdrive, and EQ could help to get close to the timbre.

[orangettecoleman]

There should be a somewhat easy way of filtering out the harmonics aside from the fundamental, right? I mean, a circuit that just senses which harmonic is loudest and filters out everything else. The neck pickup oren has is presumably used the same way that fred frith uses his, to amplify the notes on the "wrong" side of the fretted note. I'm not sure if that would affect the harmonic content of the notes played on the "right" side of the fretted note. could be wrong though. I think killswitch/overdrive/pitch shift is probably the right diagnosis. I think i have an extra switch, I can put it on my guitar in like 10 minutes when i get time, and i have a boss super overdrive pedal. I have a danelectro octave pedal as well, but that's just octave down and two octaves down. maybe i can figure out a cheap and dirty way to pitch shift up, because I hate whammy pedals. maybe one of those octave fuzz pedals with the fuzz turned down. Danelectro makes a cheap one..... I wonder if I can circuit bend an octave pedal to do "up" shifting instead of "down" shifting? having a pedal that would give me two octaves of upwards pitch shift would be pretty excellent.

[jkudler]

i still don't get why you want sine waves out of the guitar when you could just use a keyboard or something, but i'm sure you can do this with a computer. i also thought the moogerfooger freqbox did this, but flies says no.

the pickup on the neck should pick up sympathetic vibrations on the left side of your hand fretting the note, in the same way that picking behind the bridge resonates the strings over your other pickups. it'll give you some added related harmonics, or just the harmonic vibrations if the other pickups are off.

-j.

[orangettecoleman]

i still don't get why you want sine waves out of the guitar when you could just use a keyboard or something, but i'm sure you can do this with a computer. i also thought the moogerfooger freqbox did this, but flies says no.

the pickup on the neck should pick up sympathetic vibrations on the left side of your hand fretting the note, in the same way that picking behind the bridge resonates the strings over your other pickups. it'll give you some added related harmonics, or just the harmonic vibrations if the other pickups are off.

-j.

You are probably right, sine waves aren't exactly what I'm looking for. Something with a really strong fundamental would do the trick, "bell-like" in the Fender Rhodes sense like the above poster said. That's exactly what I'm looking for. I could use an ebow and a kill switch to control the attack. I guess I just want to figure out a way to de-emphasize the non-fundamental harmonics so I can experiment with note combinations and the resultant overtones, beating effects and what-not without having inherent overtones from the guitar itself messing everything up. I can't really play keyboards to the extent that I know what the note will sound like before I play it the way I can with guitar, it makes it a lot easier to be intuitive with your note choice when you know an instrument well. I think a lot the ideas everyone here brought up are definitely worth trying, and don't involve a lot of expense, which helps. all your suggestions have been really helpful, thanks everyone. I'll make sure to post some stuff under the demos section when I have it sorted out.

[Moon]

I think the FreqBox can't produce sine waves because it's using hard-sync. There is an oscillator output on it, but it bypasses the input when you plug something into it, like on the ring modulator.

Read up on the Ring Stinger now and it has a lot of features that would help in moulding the tone in the way he does, including octaves down. I almost forgot that it wasn't just a normal ring modulator.

[flies]

There should be a somewhat easy way of filtering out the harmonics aside from the fundamental, right? I mean, a circuit that just senses which harmonic is loudest and filters out everything else.

this would be practically impossible for an analog pedal (it might involve something like 1000 parallel bandpass filters). detecting seperate harmonics is tricky. one could create an algorithm to 'filter out all but the loudest harmonic' easily with fourier analysis, but it would be glitchy. once you're in the digital domain, you might as well do a more refined pitch detection.

I wonder if I can circuit bend an octave pedal to do "up" shifting instead of "down" shifting?

based on my understanding of how octave down pedals work, i don't think this is possible either. unless i'm mistaken, these pedals detect zero crossings. they create a rectangle/square wave output that switches polarity every two zero crossings. for a simple input, like a sine wave or something, this creates one square wave period for every two periods of the input, halving the frequency, hence octave down. it wouldn't work the other way.

there are 'octave up' pedals (often called "octave up fuzz", they were popular in the 70s). typically, they are 'full wave rectifiers', and in principle this is a very good approximation of doubling the frequency (perfect full wave rectification is almost like ring modding a signal with itself, so that the difference frequencies are 2f and 0). In practice, the doubling is clearly audible but surrounded by fuzz due to nonlinearities in the rectifier circuit. these pedals will considerably brighten the tone of your guitar, so i'm not sure they'd be helpful for getting the soft bell-tones you're after.

fwiw, i think moon was right on when s/he suggested just killing the attack with a switch or something. i would start right there and then see if you want an overdrive in addition.

you could also try an envelope filter or something to further muffle the treble. This would only work in conjunction with moon's suggestion, since the envelope filter (aka autowah) produces that quacking/wah sound when used normally. hmm i guess this might only work if you mute the output of the filter rather than the input...

[Red]

There should be a somewhat easy way of filtering out the harmonics aside from the fundamental, right? I mean, a circuit that just senses which harmonic is loudest and filters out everything else.

As flies notes, this would be ridiculously complex to do in hardware. Assuming the fundamentals are in the same general register, just use an EQ with high frequencies rolled off, or use a low-pass filter.

[Moon]

I propose playing the guitar and a vocoder keyboard being fed the guitar at the same time in the same pitch.
*drum solo*

Seriously though, I think I actually saw a video with Battles where it looked like they were doing it. Good laughs.

[d.at]

if you do ever hear sine tones from ambarchi it is because he has a test tone generator he also uses.

http://www.youtube.com/watch?v=I3pEaTBaWp4

52 seconds into this video.

[nicenick]

way late to the discussion and not a cheap answer, but there are also max/msp and pd externals that read audio input and track the pitch - spits out a number that can then be using to generate sinewaves (plus some other attack/decay info). been a while since i looked at it, but i believe they're called fiddle~ and pitch~, and there are probably others. no idea if this is what ambarchi does or not though.

[kp*]

if you didn't care about maintaining you original pitch then you could do this easily with an envelope follower. I have done it is software, with hardware i guess you'd have to have the env. follower send out CV to an oscil.