# Frequency shifter unit

Been looking at trying to come up with a frequency shifter patch in the ER301 based on some of the patches in this website:

https://www.cim.mcgill.ca/~clark/nordmodularbook/nm_spectrum_shift.html

In particular figure 13.2 looks the most straight forward to implement (see below):

However to do this patch I’d need to be able to create two sine waves that have a phase difference of 90 degrees. Anyone got any ideas how I could do this patch or alternatively make a frequency shifter unit

3 Likes

Cool!

See the TZFM thread for info on how to use a fixed delay to create phase shifting of wave forms - seriously insane stuff can happen if you feed a phased sine back on itself

In short, create your sine wave however you want to do that, add a fixed delay after it make it 100% wet and adjust the delay time, you could do the maths and work it out, but I never bother with that stuff, just do it by ear!

Also think about the Linear VCA and how you can use that to invert any signal - loads of fun to be had!

Yeah, I’d thought of the fixed delay but that would only work if the oscillator frequency is fixed. As soon as you change the pitch you’d have to change the fixed delay time. Thanks though!

Ahh yes!!

No idea how this will pan out in reality as it will depend on the ratios, but what about mapping the pitch to the delay time? Might be tricky to get right!

I tried to do it with a clock delay, i take the first sin, with a limiter i transform it in a square and i use it for a clock for the delay. After that i search for the good ratio for num and div. unfortunatly it acts strangely and what is working for let’s say 200hz is not working for 205 hz. but i think it could work with a little trick.

Another way to do it would be to use the single cycle waveforms in a sample player and shift the start point, I don’t think this will work with the current firmware because of the need to have triggers as soon as you have a slice, but you could manually edit the wave form elsewhere to get the phase!

Yeah, I think this is the best approach combined with a sync trigger to make sure they start at the beginning of each sample file. Going to try this tomorrow morning, I’ll report back!!

3 Likes

I’ve put together a quicksave based on that same diagram and was about to start a new thread before I came across this one. The diagram only shows the “sum” output, which is the positive frequency shift. To get the “difference” (shift down) output, you can just add a linear bipolar VCA at the end of the sub chain of one of the mixer channels with a gain of -1. EDIT: This doesn’t seem to work. Looking into it more. This means that the Morton Subotnick trick below won’t work…yet
To use this quicksave

1. load whatever sample you want into the “Sample” global chain (alternately, set the inputs of mixer channel 1+2 in the ‘Out 1’ to whatever other signal source [as long as both mixer channels are set to the same source])
2. Go to the G.Clock2 global chain, and increase the bias to the frequency of the Aliasing Saw in that chain, or map it to some positive-only modulation source. 10000Hz is 0Hz frequency shift, relative to the primary oscillator. Frequency shift is therefore (frequency of G.Clock2) - 10000Hz.
3. BONUS to get some Uncle Mort magic going a lá Silver Apples of the Moon, copy the chain in Out1 to Out2, add the aforementioned inverting VCA to one of the mixer channels to get the “difference out” and pan them left-right with a snappy envelope on the G.Clock2 modulation to get some wild psychoacoustics.

(Updated the quicksave, had the phase values of the cosine oscillators off by .25. I’m still getting a weird extra sideband when shifting up a sinewave, but it still works well enough. Also added a fourth global chain “Ctrl.” This offset object controls osc 2 as well as the LPFs in between the first and second modulators all at the same time.)
Freq Shift Quicksave v2.zip (11.1 KB)

Thanks @Joe for the awesome video on making a ring mod with the VCAs for helping me wrap my head around the concepts needed to make this work!

Here’s a little video of me demonstrating it on a breakbeat sample (the weird aliasing you hear is because I forgot to change one of the G.Clock oscillators from Sine to Saw. The attached quick save sounds much better).

5 Likes

Looking forward to checking out what you’ve built here. I’ve made a few attempts at a frequency shifter and never quite got to anything that worked to my satisfaction.

1 Like