Eurorack Modules You Want in VCV Rack

Plugins & Modules
82

hmm … thanks, ok will try it out.

Frankly, from technical standpoint, aliasing is critical issue and can be resolved by oversampling. Nevertheless, from a musical point of view aliasing is just an avoidable digital artifact, like analog VCO drifting / S&H drooping etc… As long as it sounds musical, who cares. Moreover, digital eurorack modules done by Mutable Instrument / Noise Engineering / Endophin.es etc., they all aliases in certain extent if you use a complex modulation source. The key is to embrace it, like Noise Engineering’s oscillator IC has a variable sample rate, so that the aliasing artifact align with the harmonics of the tone. This simple idea can already be an interesting VCV Rack module. If it is used along with ADC->DAC + Walsh Function Module, for instance

83

All those things are true, but… I don’t think any of those Eurorack modules alias like crazy in normal operation. Yes, they can be driven there…

Most people will not be surprised that I have analyzed many, many, many wave-shapers in VCV. Many of them are quite good and have some sort of alias mitigation. Others are awful and have no alias mitigation at all. That would be fine with me it they told the user “this is a special ultra dirty wave-shaper”, but sadly they usually do not.

Also, of the wave-shapers that can output both even and odd harmonic, many will output DC on top of the waveform, which can be a recipe for pops and clicks.

My “Shaper” wave-shaper gives user control of oversampling to reduce the aliasing to almost nothing. And it has and optional HPF to remove the DC for audio applications, but still allow it for CV applications.

A quick look at BZ mapper indicates it seems to have not alias mitigation or DC removal. I suspect it’s mainly designed for CV applications (The author is certainly a DSP expert).

84

I see, I think that is usually the case - a module designer has a vision to their module and implemented it for a specific purpose, but the end-user end up mis-using or re-appropriating it to something else. I guess that why modular is so fun.

Hmmm … I think all waveshapers can output even and odd harmonics emphasized waveform if you mix the input waveform with a DC offset before sending it to waveshaper (wavefolder symmetry). Interesting to know about a lot of them will output DC on top of the waveform and that will contribute to pops and clicks. May be I will try messing up some of them with a slow LFO and see how they response.

I thought digital modules are less prone to that. Seemingly it is perhaps best practice to use a steep slope HPF to filter out DC voltage at the end-of-chain, like most modules in Eurorack are DC coupled, except the AC-coupled end-of-chain audio out module.

85
2 Likes
86

Yes, you are right that you can get even harmonics that way. It may require an external module if the waves-haper don’t already have a DC offset.

Didital module are less prone to random offsets, or little offsets that creep in due to component tolerance, etc… That is for sure true. But they are of course susceptible to the laws of math. To take a trivial case, absolute value (rectification) is a common thing. And obviously if the output is always either 0 or above it will have DC. But that’s true in less contrived wave shapers, too. They need to either have a filter built in, or have internal math to counteract the DC tendencies of the non-linear functions.

And, yeah, for sure it’s a best practice to not output DC if you don’t have to. But - spoiler alert - look at the pulse wave outputs of almost any VCO in VCV - you will find a huge amount of DC, often more DC that signal.

2 Likes
87

It’s being worked on as we speak :wink: .

4 Likes
88

So I was reading my digital copy of this month’s Sound on Sound and one of the modules they spotlighted was this:

It sounded interesting and I wonder what it sounded like, so here is a quick video of my take on it:

Still working on cleaning up the UI, and I am wondering if there is anything I can do to take it “next level” beyond what the original offers.

Anyway, It does seem like a pretty interesting sounding VCO - pretty easy to get vowel/formant sounds without a filter. I’m sure other folks like Omri could do something more interesting than I did with it.

Should be in library soon :slight_smile:

10 Likes
89

That module is pretty easy to make with 2 oscillators and a switch. edit: unless I’m totally misunderstanding what it does.

90

If you read the description - it is pretty easy to make with 3 oscillators and a switch, until you start adding some of its other features. Thanks for the feedback tho

1 Like
91

Thanks once again, for creating a new ‘Recombination’ Osc engine, it is appreciated. Looking at the price of the actual module reminds me why I can’t afford to go down the Eurorack wormhole! The Formants option do sound an interesting, how would you further enhance this or any other option?

93

still pondering. I did make the pw skew bi-directional, but now waiting for inspiration

94

Sure the muse will strike soon and you’ll come up with something odd and wonderful.

95

I love matrices.

Are there something similar to this other than VCV Octal Router?

I also think of an altered application of sequential matrix for tuning the carrier:modulator ratio for TZFM oscillator, are there something similar to that? You use CV to modulate the x-y direction of the matrix, the pure intervals lock the C:M ratio to pure-intervals (i.e. 2:1 / 3:2/ 4:3) only for a more musical outcome. A Just intonation / 12-TET switch and a knob for different scales (to skip certain intervals) might be great too. A button might be needed for 1-12 to reduce the range of frequency permitted.

matrix
matrix1920×1316 142 KB

1 Like
96
VCV Library - Bogaudio SWITCH44
VCV Library - Bogaudio SWITCH88
VCV Library - Bogaudio SWITCH1616
2 Likes
97

Just got it from the library, and I love the wide range of interesting tones that I can quickly dial in, without knowing what the hell I am doing. Looking forward to studying it in detail.

1 Like
98

I did some more experiments, and though your module definitely creates some great sounds, it is missing some significant features from the original. I think the biggest difference is your DNA VCAs always apply scissor switching, whereas the original starts biasing the switching at 12 o’clock such that switching is completely off at full open (5 o’clock). There are a lot of interesting sounds achieved with this feature.

Also I think the PW skew effect is very subtle in your version compared to the original. Then there is the missing Glue section and your DNA outputs are much limited, so fewer internal modulation sources for things like AM and FM.

I get it, you weren’t trying to create a replica. But the sonic range of the OSC2 is staggering, maybe worth emulating a bit closer. Either way, you provide great modules for all of us for free, and maybe you have done all you want with this module.

I created a full on emulation using a bunch of free modules, and it is awfully fun to experiment with. I think it could be the basis of some incredible patches. I’d be curious as to your opinion. Having the full functionality in a single module would be so much better!

FSS OSC2 Emulator.PNG
FSS OSC2 Emulator.PNG1920×1040 285 KB

1 Like
99

I don’t own the module so I don’t have a comparison, but I’ll see if I can make it even more comparable

I just saw the description in Sound on Sound :slight_smile:

100

In my memory my attempt to clone the Recombination Engine was inspired by an interview FSS did with Molten Modular but, now that I check, I think it was this interview with DivKid:

edit to add that I mean that I was directly inspired by a diagram in this video. If I hadn’t seen that I wouldn’t have bothered.

101

So, I think I am pretty close. I really wish FSS had a manual online and not said just “watch the video”.

I noticed in your patch you are using the Mix control to balance between positive and negative DNA oscs, but I don’t think that is what the module does. Unfortunately, I can’t really tell what it does do :slight_smile:

But I do have the 12oclock/5oclock VCA controls working (with cv controlled bias).

There was indeed a bug with PW Skew, which has been fixed, but I am going to now make the skew amount a CV controllable parameter, which the original can’t do

If anyone knows what Mix and Mix In, do, that would be awesome! Otherwise there is going to be a knob there that will be like the Close Door button in an elevator :smiley:

1 Like
102

Good, I’m glad you looked at my emulator.

Yes, the DNA mix and GLUE section was the hardest for me to understand. I spend a long time studying the DivKid video to get an understanding. I’m pretty sure the Pos DNA signals are initially simply summed to provide the Pos mix. At some point it likely needs to be attenuated, as that can exceed 10 volts. The same for the Neg. I don’t know if the attenuation occurs in the DNA mix circuit, or the GLUE, or both?. However it is done, the OSC2 seems to have a fixed algorithm. I use the poly VCV Sum to mix the 3 DNA signals (sin, tri, ramp), and thought, "Why not give the user the ability to control the attenuation and mix ratio of the Pos and Neg DNAs.

So in the GLUE, the incoming Pos and Neg mixes are mixed at a fixed ratio, and then the “Mix In” gives you the opportunity of mixing in a 3rd signal. The Mix Level attenuates the “Mix In” before it is mixed with the Pos/Neg mix. The final mix is then amplified by the GLUE VCA , which has a VCA CV input with a depth control (attenuator), plus a VCA bias. I use an attenuverter instead of an attenuator for the CV depth control.

I’m pretty sure the PW skew preserves the waveform within the pulse window as it is stretched or shrunk. It took a while, but I finally figured out the Log2 math. Assuming I have bipolar -5 to 5 volt CV controlling the PW, call it x, and the PW duty cycle varies all the way from 0 to 100%, then the Pos V/Oct delta is -(Log2((x+5)/10) + 1), and for Neg it is -(Log2((x-5)/-10) + 1). But for the Bogaudio VCO the PW ranges from 3 to 97%. So the formula is adjusted to -(Log2((x+5.3)/10.6) + 1) and -(Log2((x-5.3)/-10.6) + 1). I’ve got a modified version of the emulator that preserves the waveforms perfectly as they are expanded or shrunk. Assuming PW goes from 0 to 100, the formula exactly halves the frequency (doubles the wavelength) of the stretched DNA signal. Conversely the shrunk DNA frequency goes down 1 octave at 25%, 2 octaves at 12.5%, 3 octaves at 6.25%, etc. And then just as the width goes to 0, the frequency becomes undefined (infinity) which makes sense.

The other thing I struggled with is the exact configuration of the DNA waves. The FSS Molton Modular video confirmed for me that the DNA sine is actually fully rectified. So assuming the Pos and Neg are at the same frequency, then the rectified sines are mirror images of eachother, but effectively at double the frequency of the base frequency. My understanding is the triangle waves are also mirror images, but not doubled. When I scale and offset the triangle sources to get the Pos and Neg, it messes with the phase, so my triangle starts at 2.5 volts instead of 0. I discovered I can get the Pos and Neg triangles with the correct phase by rectifying a ramp instead, but that introduces a horrible high frequency buzz at the triangle peaks that is an artifact of the anti-aliasing signal at the source ramp discontinuity. So I am sticking with the out of phase triangle for now in my emulator. Based on the graphics on the OSC2, and also from what I can see in the videos, the Pos and Neg ramps are Not mirror images, but rather true ramps that move in parallel.