Octaver question

[bilo01]

Hi all, i'm looking for a vero layout off a clean octaver, not the fuzz realated kind like on tagboards. Does anyone have a suggestion?

[deadastronaut]

there is no such thing imo...(unless you want half hearted 12th fret up type lame faux octave)

for good  octaves , as a diy 'er i'm ashamed to say i'd just buy one...

or learn dsp programming etc...

[bilo01]

I'll take your word for it deadastronaut. I guess it's shopping time again

[deadastronaut]

yup!..for instant proper octaves i would...saying that

the awesomely clever and knowledgeable dudes on here like ice-9 mick and ian -slacker mess with the spin semi FV-1 and gget phenomenol results.

and they would no doubt point you in the right direction diy wise..

but its way above my head....

[Mark Hammer]

I don't think any of us here have experimented with it, but theoretically "clean" doubling can be achieved by feeding an input signal into a VCA, and modulating that VCA by the same signal.  Aphex uses this method to produce the added harmonic content in one of their "exciters".

[samhay]

Mark - from memory, you still get problems with intermodulation distortion that way, so it is only clean for single note work.

[Mark Hammer]

That may be why Aphex uses it in conjunction with a steep highpass filter that removes the fundamental! 

[tommycataus]

I've looked around for one and the best design I've seen is the pearl octaver. Hopefully madbean still has PCBs for sale on his site as I'm planning on buying one. I doubt you'll find one on Vero that will be of a moderate size so this is the next best thing for a DIYer imo, short of buying one ready made.

[Digital Larry]

I don't think any of us here have experimented with it, but theoretically "clean" doubling can be achieved by feeding an input signal into a VCA, and modulating that VCA by the same signal.  Aphex uses this method to produce the added harmonic content in one of their "exciters".

I tried this last night on my DSP system and it sounds anything but clean.  Multiplying a signal by itself (squaring) does something similar to rectification.  I didn't do any filtering.

[thehallofshields]

I tried this last night on my DSP system and it sounds anything but clean.  Multiplying a signal by itself (squaring) does something similar to rectification.  I didn't do any filtering.

I don't mean to be a smartass, I'm clueless with DSP, but you should be doubling the Q for an Octave, not squaring it.

Example: 440hz = A3,  880hz = A4, 440 ^2 = 193,000hz

[vigilante397]

I don't know as much as I would like to about DSP, which is why I think the POG2 is a killer deal on something I can't build myself.

[Digital Larry]

I tried this last night on my DSP system and it sounds anything but clean.  Multiplying a signal by itself (squaring) does something similar to rectification.  I didn't do any filtering.

I don't mean to be a smartass, I'm clueless with DSP, but you should be doubling the Q for an Octave, not squaring it.

Example: 440hz = A3,  880hz = A4, 440 ^2 = 193,000hz

Mark was talking about modulating the control signal of a VCA with the signal itself, which is also going to the VCA's input.  Since I don't have a VCA handy, I used the "multiply" (actually a volume control) function of my DSP just to see what it sounded like.  There's a trigonometric identity:

sin^2(x) = ½[1 – cos(2x)]

Which says, if you "square" a sine wave, the result consists of a DC component and a signal at twice the frequency - not the square of the frequency.  So at one mathematical level, it should work for sine waves.   I'm not sure why it sounded so crappy; might be worth further investigation.  But if it did work in general, seems people would be using it lots more than DSP and raspy sounding octave fuzzes.

Well OK I think I have it after a bit of thought.  It WILL work for sine waves - though you get some DC in the output.  

But, if you consider the case of two sine waves added together like this:

sin(a) + sin(b)

If you square this, you get:

sin^2(a) + 2 sin(a)sin(b) + sin^2(b)

It's the 2 sin(a)sin(b) stuff that is all the extra hash.  If you could get rid of that you'd be pretty happy about it, I'd wager!

[merlinb]

I'm surprised no one's mentioned RG's Mu Doubler:
http://www.geofex.com/Article_Folders/jfetdoub/mudoubler.htm

[Eddododo]

I'm surprised no one's mentioned RG's Mu Doubler:
http://www.geofex.com/Article_Folders/jfetdoub/mudoubler.htm

I never got mine to work , though the error is beyond doubt my own

[Digital Larry]

I did a test last night, comparing the sound of the signal "squarer" with a raw rectification stage.  They are actually quite similar - but the rectifier "octave fuzz" has a harsher, buzzier sound.  They sound much better on single note lines than chords.