Serge thread

[robbbiecloset]

Hey folks, debugging a not-fully-working R*S Sequencer / Programmer module. The symptom is that I’m not getting any CV out from output A.

I am very new to debugging something like this, but from the look of other banana jacks on the board that are functioning, it looks like the jack for A could be the culprit:

I don’t think it’s fully separate (although it could be?) but it’s certainly not as well connected as others.

Just looking for someone with a bit more experience to say if re-doing this connection might be a good place to start?

Thanks!

[kbit]

Absolutely a good place to start, looks potentially problematic and is easy to just reflow / reinforce that joint. Was that a DIY build?

[robbbiecloset]

My understanding is that it was a Patch Point build. Thanks for the tip! I will give it a try.

[pelang]

Continuation of the modulation series of serge drones.
Slow modulations and subtle shaping with the resEQ.
(play at low volume)

[tytytytyty]

Glad you’ve enjoyed those walkthroughs, @muncky & @gnome666. NCOM has been on my short list to write about for a while. I have a bunch of notes (many courtesy of @cebec) from long-gone blog sources with lots of tips and uses for NCOM. And I still haven’t had a chance to work my way through @ht73’s excellent comparator post from last summer (there’s just not enough time to do all of the things!).

In any case, thanks for wishing aloud for what you want - it helps to know!

[robbbiecloset]

So a month and a half ish after my original post here, I have a recording of my first patch on this system.

I’m having a ton of fun so far. I think the only thing I’d tell my past self is: you’ll miss having dedicated VCAs… but! it will inspire you to patch a little differently.

For example, most of the movement in this patch is achieved by being fairly aggressive about modulating filter frequency to reduce cutoff. Obviously a simple thing to do, but something that I’d found myself doing only minimally before, if at all. It had me thinking more about texture.

This had a downside as well. As I mention in the linked thread where I posted the video, I took about a week trying to build something else on top of this sequence (before adding all the effects) and really struggled to find room for anything else. In the end this is also very ok. But there was a little frustration I had to work through.

I’m not sure if I will leave this boat as it is. I have the filters module up for sale so I could potentially fund a filter / stereo mixer module, which is now available at Patch Point. While I’m having fun with the texture focus, I think I’d like to be able to experience a Serge system with some more VCAs for more traditional sounds.

Similarly, I feel like I want to try a dedicated oscillator at some point. While I love the ideology on display here that allows for the possibility of modules to satisfy a variety of different purposes, I admit I miss the “fine” knob a dedicated oscillator would have

Also, I find myself missing more complex envelopes. When I did experiment with using the DSG & Carnivore as a LPG (& VCA), I struggled to dial in AD envelopes that I really enjoyed. I can’t say that I ever see myself making room for the Extended ADSR module, so I guess I’ll stick to textures & plucks in Serge-land

[GrantB]

Are you feeding the slope generators back into themselves to control the rise and fall functions? Joining the two sections? A DUSG can cover a lot of shapes.

Random*Source Serge DUSG 2/7: Envelope Generator (LMS EE)

[robbbiecloset]

This is awesome, thank you!

[gnome666]

yea, that whole DUSG series is the most in-depth well explained Serge module tutorial i think I’ve ever watched

[ht73]

@robbbiecloset

I can’t say that I ever see myself making room for the Extended ADSR module, so I guess I’ll stick to textures & plucks in Serge-land

I wouldn’t bother with the ADSR, it’s rather slow regardless of setting or exponential shape/feedback and I don’t miss it. The DSG can be very fast, capable of very acoustic-style plucks and strikes.

To expand the range of envelope shapes, there are three techniques I find useful:

1. Feed a very thin pulse into the USG input instead of triggering it via gate in. This way you can get very long decays without skipping triggers, like a standard AR or AD. The other USG can be used to synthesize the thin pulse by setting the rise and fall times to minimum and triggering it via gate in.

2. Use feedback as mentioned – output into rate control. you can get exponential shapes with positive feedback, or inverse exponential with negative.

3. Full-wave rectification, if possible – The '73 CV mixer/processor naturally full-wave rectifies (takes the absolute value of) its intended output, so negative CV’s get inverted and become positive. So you can process a percussive envelope by gradually introducing a negative offset. The shape will start ‘welling up’ at the end, also with reduced attack. The rectification trick allows you to morph an envelope into its inverse without zeroing it out, like a scaling control would do. Unfortunately I think the rectification was considered a bug and optimized out of newer designs, but if you can figure out some other way to patch this on a modern system, it can be helpful in terms of modifying envelope shapes in useful ways.

These tricks don’t get you the full functionality of an ADSR but they can significantly expand what you’re able to do.

In general for pluck sounds I’d use linear shape to control the VCA and exponential for any other kind of timbre control – and always use the thin pulse trick. Triggered use cases are better for slow LFO’s that are synchronized to the main time grid.

[swannodette]

@robbbiecloset

Similarly, I feel like I want to try a dedicated oscillator at some point. While I love the ideology on display here that allows for the possibility of modules to satisfy a variety of different purposes, I admit I miss the “fine” knob a dedicated oscillator would have

You can actually use the the rise and fall CV input knobs for fine tuning control, it works well.

[robbbiecloset]

You just blew my mind.

I did a very quick test here (perk of WFH days) and this seems to be true for the for the VC input knob on the oscillator side (left) of the DSG on Carnivore, but maybe not on the right side? I will try with more detail later on tonight.

Either way, maybe I just need to switch to using one of the DUSG mkii as the second oscillator. Thanks so much for the tip!

[markeats]

Yes I’ve found that too with those ‘mini DUSGs’. DUSG XL makes for a good oscillator though.
Another useful not-enough-VCAs hack: If you’re using a unipolar sound source then you can use the lower slope gen as your envelope and pass the sound through the Trough section.

[robbbiecloset]

@markeats

Another useful not-enough-VCAs hack: If you’re using a unipolar sound source then you can use the lower slope gen as your envelope and pass the sound through the Trough section.

Yeah, I saw this up thread in another comment and it was a life saver. (Also, super cool to think about why this works. I feel like these modules better encourage me to try to understand what is going on for real.)

Another VCA question / observation: using one half of DUSG XL as a VCA (trough approach or directly) sounds great. Using the smooth section of the Carnivore I find that it’s super quiet. Anyone else have a Carnivore and use it this way experience the same?

[mzero]

Over the last few weeks I’ve made this little beast:

\

Which is a small microcontroller with some add on circuitry:

To get 4 CV outs and 4 triggers out to my Serge.

The system uses the same PWM technique that @lsky was using, but with proper filtering of the PWM, buffering of the signal though opamps – and some low level tweaks in software to get the PWM to work well.

Here’s it being a sine LFO:

Here’s a short Instagram video snippit of my performance with it last week: https://www.instagram.com/p/CEcoaPXB3Sv/

I’m running a MIDI looper on it - where I play on MIDI drum pads into it … then loop that out as CV and triggers to the Serge. But the same hardware, and the basic code that drives the outputs could be put to any use you’d like…

If a few folks here were interesting in make these, then I’d be into starting a new thread for it - putting together my schematics and code - and helping people through building and coding them.

[wheelersounds]

This is superb! I was planning to build something similar with a teensy, but this looks pretty special, and I’d love to learn how you applied filtering.
It looks like a Feather M0, an OLED “wing,” and dacs? Are you maybe using Arduino, CircuitPython, or C?
(Also, excited to maybe use this on my sidrax!)

[mzero]

Workshop idea initiated in this thread: Build a Microcontroller C.V./Trig Thing Workshop

[vals]

I’ve just received this BOG panel and I’d like to rewire the molex connector to a Neutrik NC4 MX 4-pin XLR for power supply.
Can you give me an advice on how to wire them?
(I also realized that there are only 3 cable instead the usual 4. Is this something to concider?)

PS. Sorry for the bad pictures

[emenel]

Rather then re-wiring, I would make an XLR > molex cable and plug that into the panel.

The 4th wire is a second ground connection I believe, so maybe it’s ok without it?

[vals]

R*S suggest in this document (pdf) the following:

As far as understand from that doc, If I make an XLR > MTA-156 cable, I just wire pins 1-4 to both connectors?