ModularGrid Rack

Hey folks,

I want to share some ideas with you that I figured out in the last year, building this Rack/Patch.
I'm dying to hear your ideas for optimizing or pushing it further in any interesting kind of way.

I've limited myself to this 4 rows of 104 TE (MDLR-Case, Performer Series) and want to achieve:

a Techno-Live-Gig-Rack without external gear, but with both: sound and visual output.

The rack started in march '21 and the setup is now in version 3 - I'm nearly where I wanted to be with it. The Setup could (can) run the Kick-drum and most of the voices in a technically "danceable" way for 60 min. on its own (NerdSeq, of course).

I want to focus on mostly changing the modulations, give completely new Input (e.g. microphone) or interrupt / change the whole schedule / procedure.... live... for the sound and for the visuals.

I think of it in "blocks":

**Control block:
- NerdSeq
- 2x CV16 expander
- Trigger 16 expander
- More Video I/O expander
- QSlider
- PM Mutes
- Genki Wavefront + Ring (on my right hand)
- special DIY microphone (connected to X1L3 Shard Input via "t.bone Sync1")

NerdSeq is controlling nearly all the voices via outputs and expanders. NerdSeq channels programmed:
- on channel 1 is the Trigger 16, running drums and some specials
- channels 2 - 5 are for all the controlled voices
- channel 6 is for special modulations and video CV-controlling
- channel 7 runs samples from NerdSeq directly into PMixer
- channel 8 is the video programming channel, where mostly Text and full picture colours / strobes are going out through the More Video I/O => CVBS

**Kickdrum block:
- ModBase 09 MkII
- Ghost
- Kick + rumble; Ghost as send fx (100% wet), mixed again in PMixer. For some tracks I layer a "small" Queen of Pentacles Kick (from the separated output) additionally into PMixers Kick-channel... Presets of ModBase are controlled by NerdSeq CV.

**Voice blocks:
- BIA => Pico VCF1 => PMixer
- Manis Iteritas => PMixer
- Loquelic Iteritas => Quad VCA => PMixer
- NanoRings => PMixer
- 2hp Play => PMixer
- Queen of Pentacles => PMixer
- Bitbox Micro => PMixer
Everything obvious. Bitbox Micro is playing an intro (triggered by NerdSeq via CV-Thing) and some fx-samples. Otherwise I shoot some fx-samples via touch display or (not live) record stuff from the ambient-voice going through Bitbox.

**Ambient block:
- X1L3 Shard => WASP Filter => Electus Versio => Quad VCA => Bitbox Micro (only tru) => PMixer
The Quad VCA recieves a Kick-sidechaining-signal from Roti Pola: Envelopes from Messor + Ghost are mixed. Ambient block can "pump".

**Mix/Master block:
- PMixer main out => Prism => Messor => HN Out (main input)
- PMixer cue out => FX Aid as a "big room" Reverb => back to PMixer
- PMixer aux 1 out => Imitor Versio (running Electus Versio firmware) => HN Out (2nd input)
Prism acts like a Master-HiPass-Filter with special short-delay-FX, controlled by the Q-Slider. Pressing the button on Genki-Ring opens signal path coming from PMixer cue out via Quad VCA, so selecting the send-elements works with cue buttons. Via PMixer aux 1 I send a small amount of elements to create a general atmosphere in the background.

**Modulation block:
- Quadrax
- Zadar
- CV Thing
- Genki Wavefront + Ring
- Muta Jovis

Quadrax for envelopes and modulations, Zadar only for modulations. CV ThingUSB out => Structure USB in; this makes presets of structure available to be controlled via CV from NerdSeq. CV Things TRS-Midi-out is connected to Bitbox Micro, so theres another option to control Bitbox via CV. I use three channels of CV Thing to only display the current value of Genki Wavefront's outputs. This way I've got the WASP-filter for ambient-sweeps and/or e. g. main voice's envelope length "in one finger". The ring can of course temporary be de/activated and Muta Jovis can "kill" some modulation connections.

**Visual block:
- NerdSeq more Video I/O via CVBS => ET Structure => CVBS to HDMI => ext. display

The saved presets of Structure are controlled by USB-Midi-Signals, coming from CV-Thing.

Any opinions, suggestions or questions are welcome!

You might find more compelling results by focusing on getting more out of one or two voices rather than ... looks like 9? I've found it instructional to compare LIFKA, Lady Starlight, Lucas Marchal, and Blawan's sets on the techno side, and Helen Vogelsinger and Caterina Barbieri on the melodic side. To me, the simpler setup typically produces the best results and allows the performer to be directly involved and more spontaneous.

It seems you could be approaching from the mindset of covering all the bases of a typical commercial techno track rather than really thinking and feeling through how to extract the most and do something unique with every component of the track. I've found this produces generic results. As sad as it may be, running part of your set as samples coming from an Octatrack is something nearly every performer lands on eventually—you might find it more efficient to 'fill out the track' that way than rather than building an Eurorack orchestra (I'm barely willing to admit this to myself).

In my own setup—which is perhaps extremely minimal—I use a single oscillator, kick, and hi-hat. Everything else—melody, other percussion, rumble, etc.—comes from running the mix through an LPG, filters, delay, and reverb. I've been surprised by how well this works.

From another perspective, I think your system seems totally fine as is, it's just overpowered and finding the sweet spots may take a lot of experimentation. It may also be perfect for you in ways I can't imagine!

Finally, I could give much better feedback if this post were accompanied by a recording.

Finally, I could give much better feedback if this post were accompanied by a recording.
-- alerion

Hey alerion,

thx for your honest feedback!
You brought some good points I really have to think about, especially for the number of voices (maybe more modulation/Utility instead?) and of course the missing recording ;)

If you like to dive deeper into my idea, here some further information to make my intention a bit more clear:
The rack is not primarily built for improvisation or finding a sound "per accident" - therefore I'm using a smaller Rack with other modules: Steppy, Mimetic Digitalis, Kickain, etc...

I think of this big rack more like a kind of programmed machine where Improvisation comes in (or on top) while it is running a programmed schedule; kind of like the the song-mode on the octatrack / analog rytm.
Maybe 80% of the patch are still patched from 2021 and come from my experimentation and experience with the source-modules. Thats why some modules are combined in "blocks" (e. g. Shard + WASP-Filter + Electus Versio: love it) and other modules I like to use directly into Mixer (e. g. Manis).

The NerdSeq captures a few "basic-produced" tracks (all possible 176 patterns are used in the current file). No track uses all voices at the same time (except one I think), but every voice is playing a key-role for any of the tracks.
For example on BIA and Manis all CV Inputs are patched from the CV16 expanders and get controlled by Nerdseq => this way I can fully restore every ("sweet-spot") sound for a single track automatically and use voices only how I exactly want them. If I want to freak out, I use the knobs or pull/switch cables, but the modulations already kind of freak out, how I programmed them.
For backing tracks I could use Bitbox, but except the intro I don't need backing tracks at the moment. Other external gear like Octatrack oder Beatstep Pro is banished from this project (decision).

Some aspects I'm currently thinking about:
- using a CV-recorder instead of the PM Mutes (skimpy used because active elements are controlled by the patterns)
- changing a voice, e. g. throwing out BIA for another interesting thing with at least 6 CV inputs (doesn't necessarily need to be a voice, one of the tracks would need a new "hook" from one of the other voices)
- in terms of pushing further, starting a new file and make a completely new "schedule": with or without changing the patch?