Looking for opinions for a project I'm working on.

The basic idea...

I've come up with a lot of synth configurations for emulating vintage synth sounds. But I can't just build one synth design and program it to cover all the sounds.

So what I'm thinking of doing is putting each synth on a layer mixer, and then using the new 1-N and N-1 switches to provide an ability to activate only one synth (layer) at a time. Basically it's a way having push button preset sounds available from a single build, on a single track.

What I like about this idea is for me it's a far more inspirational way for working on song writing in Drambo. It makes looking for inspirational sounds easier than having to load instrument using the preset menu. With a system like this I can load one instrument that contains a bunch of sounds of a particular type, and quickly find one I want to use on a track, and tweak the programing if I want to get is just right.

The part that's holding me up on this project is efficiency. I could build entire complete synths on each layer and select the one I want to use. But that seem like a waist of CPU if all the unused synths have all their Oscillators running in the background doing nothing.

So I'm looking into ways of setting up only those oscillators necessary to run all the synths, and putting then before the Layers. This way I'm essentially using the fewest number of oscillators required, to run all the synths.

But the problem is the only oscillator parameter I can change using a switch is fine tuning and PWM. I should be able to switch octaves using CV pitch module.

This means I have to dedicate oscillators to running different waveforms, unless I want to start using Graphic Shapers on the synths layers, and drive them with external oscillators.

I don't want to devote a lot of time making complex connections.

I plan on putting this up on patch storage once completed, and I think having complete synth builds on each layer is a much better thing in case someone else want's to dig into a particulr synth and see how it works.

Having complete synth builds on each layer also makes it eisier to copy and paste any given synth fron a layer and just paste the one synth onto a track.

How much inefficiency and CPU waist could I expect by building complete synth builds on each layer?

Is it even worth trying to develop a system that runs off of a group of primary oscillators?

I'm tending to think I should just build this as a collection of synths on layers, and not worry about inefficiency.

Any thoughts?

Comments

  • edited September 2020

    You have listed very valid points and what you're basically looking for is a flexible routing and modulation system that can store each combination as a preset.

    This is still the biggest difference between a synth like thor or Sunrizer and modular Drambo. In a modular system, these routings are rather static and with the left-to-right signal flow philosophy, extra care has to be taken when switching between different routings.

    There is no simple solution for this because that would require an additional logical layer inside a Drambo project: Multiple different routings, modulations and module settings that can be stored and recalled independently from p-locks and scenes. Typically, all modules inside one track would remain in place but the interconnections and parameter values could be saved as "synth presets" and be recalled by means of a dedicated preset menu.

    @giku Let me call this a "snapshots" feature: Modules stay in place but routings and parameters can be stored and quickly recalled (using program change messages from outside or step components inside Drambo) so Drambo becomes more efficient for building that custom dream synth without bloating too many tracks. What do you think about this?

  • im not sure if it makes sense in a modular world to try to build one structure to solve it all.

    wouldn't something semi modular with a mod matrix be much more convincing and practical?


    everyone is different, whatever floats your boat.

  • edited September 2020

    I just began constructing a test build to get an idea of how a layered "preset" system will operate. I'm actually liking it a lot.

    Let me clarify so we're all on the same page....

    Step one - Build a synth that contains everything that it needs to work as a completely independent unit.... Next build some other different sounding synths, each one as it's own completely independent unit. Let's say we now have three individual synths.

    Step two - Add an instrument rack to a track.

    Step three - Put a Layer module inside the instrument rack.

    Step four - Configure the Layer module so it now has three layers.

    Step five - Inside the instrument rack, but in front of the Layer module. Add a Number Module set to a value of 1 (one) , then add a 1 -N switch just after the Number Module. Connect the output of the Number Module to the input of the 1 -N switch, and Configure the 1 -N switch so it will have three switch buttons.

    Step six - Copy and paste each of your individual synths, one synth per Layer, into the Layer Module.

    Step seven - At the far right side of each layer, add an AMP Module, turn the amp knob all the way to the left (off), then tap on the Amp knob and connect layer-one Amp to 1-N Switch #1, layer-two Amp to 1-N Switch #2, etc.. Then turn the Mod knobs on each Amp to the right so the Amps are controlled on/off by switch buttons. This will allow only one layer to output sound at a time.

    What "we" just created... Is one instrument rack containing three unique sounding individual synths on one Drambo track, and we have the ability select what sound we want to use at the touch of a single button. :)


    Put TEN synths into the Layer Module. Put TEN buttons on the 1-N Switch. Now we have ONE instrument rack, and we can choose which of the TEN unique sounds we want to play at the touch of one button.

    What I like most about this method is it does not require parameter changing, Just simple button selection on the 1-N Switch.

    BUT. My concern is about efficiency. By having a whole bunch of synths in layers on one track, and only using one at a time. How does that effect the CPU ??

    However... If there was a way to build a new type of Layer Module designed to be used as a synth container that only let one synth play at a time.. Maybe there's a way to put all the non-playing synths into some sort of "dormancy mode" where they remain ready to kick in and play at any moment, but use less CPU while they are dormant ??


    What I think is most cool about the switching idea, is you can have vastly different configurations for each synth on each layer. You can do things that you can't do with only parameter changes.

    With the switching method you can program one track to switch instruments multiple times using track automation. Just by programing 1-N Switch changes.

  • edited September 2020

    it's good that you brought up this topic. Drambo is still at its youth phase and certain aspects need more attention. This one is very important. ( @rs2000 mentioned about this before )

    I need to think about all of this deeper. Some solutions proposed here seems overcomplicated (that probably comes from Drambo architecture)

    For now, I'm sure we need a Switch layer module, where only active layer is processed. This will solve efficiency problem, no matter how many layers you have, cpu level should remain more or less constant.

  • edited September 2020

    Snapshots: it doesn't really differ from presets. Modulations are connections and result in different structure. Drambo approach is different: structure is a first class member, like parameters. Adding a new modulation is not more difficult than e.g. in Zeeon. You make sounds, not constant structures with list of settings / sounds. Structure is a part of it.

    Switch layer: it helps while performing, but its definitely not a solution for trying different sounds / preset browsing. Imagine a switch Layer module called "Bank A" with 100 presets / layers.

    The question is: Whats wrong with presets?

  • nothing.

    we really start over complicating things here.

    presets could be more hm discoverable

    and there aren't enough of them.

  • Thats it! I think that Mod matrix module would be very helpful.

  • Im thinking about adding arrow buttons to Instrument rack in compact mode. For traversing presets in current folder.

  • edited September 2020

    I love matrix mixer.

    if I had eurorack stuff that would be the first thing I would buy.

    because than you don't need to endlessly reconnect things just because you want to have the lfo or envelope on something different ...

  • Elektron style sound locks potential solution maybe? Lock ins rack presets per step and/or modulate through presets from a folder.... I feel like it’d maybe be prone to some gnarly clicks/pops when switching though?

  • edited September 2020

    as I said before modulating the matrix was one of zeeons best tricks (and its so simple)

    and very very musical

  • Switch layers would make this possible. But I think we talk about something different.

  • edited September 2020

    My bad, I just read again and I’d missed the part about being able to switch between the different synth setups with one click. Yeah, switch layers more immediate for this :) Elektron style would keep presets behind browser..

    Tangent - Not sure switch layers would be ideal in use for elektron style soundlocks, maybe though (might even be better). But I feel like the switch layers module would need to react to/feedback what’s on current step? Like if you have 30 drum sounds from 30 different layers, scattered through a 64 step pattern, if you hold a step that has a sound lock on it, the module needs to flip to that layer’s display... though similarly, any kind of sound lock implementation is gonna need this kind of feedback...so maybe it’s easier to make switch layers module react like this than it is to have different ins rack presets display when hold different steps?

  • edited September 2020

    How would you build a mod matrix if the module order can't be freely changed?

    Edit: If it's only about modulation amounts for existing routings (which I believe is what you have in mind) then mod matrix presets could come in useful.

  • edited September 2020

    Modulators on the left, targets on the right, mod matrix in between. Its rather mod mixer, but does the same.

  • Say we have 3x3 matrix of knobs, columns: LFO, EG, Pitch wheel , rows: Osc pitch, Filter pitch, Amp. This could make a lot of situations easier.

  • OK, a matrix of modulatable positive/negative gain knobs with presets that can be easily chosen while playing.

    I like the idea! It would require some proper planning from the patch creator but it's certainly the easiest solution to implement.


  • I think of Switch layers as a type of creative patch storage module, and as a work environment which assists in synth design. I wouldn't want to put 100 patches into it.

    Let's say a user is working on a song and thinking about what type of sound they want to use. With Switch layers you can build/load/paste, a manageable number of sounds of a specific type into one Switch layer module. Now at the touch of a button the user can audit their handful of chosen sounds directly in the one track, really quickly, getting real time feedback for how well the sound works for that particular song.

    Also, because the Switch layers contains the actually synths, the user can edit the synths as they work with the module. The editing can be for the purpose of getting a sound just right to use it in the track.

    But also, Switch layers can be used as a synth development tool. Because it could allow you to begin with an idea for a synth design on layer one. Once layer one sounds pretty good, copy it and paste it onto layer two and keep working on it, odds are you end up with another variety of sound that you also like. Now you have two new sounds. Keep working on layer three, you try out a different idea on layer three, and get a third new sound. In this way you evolve the sound as you work, and each new layer becomes a new opportunity to create a new variation, or add completely new sound design elements.

    Because all your work is available in a single Switch layer Module, you can copy and paste from one layer onto the next, and add modifications as you work. In the end you may like every sound you created. Then you can save it as with a name that reflects the kinds of sounds you made.

    If you were working on pads, now you have a variety of different Pad sounds. If you were working on Bases, now you have a bunch of new Base sounds. All neatly organized in their own module.

    One could even use Switch layers as a way of storing groups of favorite sounds. Rather than hunting down patches, drop your favorite Base Switch layer Module on a track, and use it as a starting point.

    I think a better system of organizing and retrieving patches is also needed for Drambo. But working from a patch retrieval system, IMO isn't comparable to working with Switch layers. Switch layers can be used for rapid sound preview, and also as a design space.

  • I agree.

  • edited September 2020

    +1 :) Switch layers would be dope for banks of favourites. Like the 8 x favourite buttons on op-1. Always useful! Tried similar workflow setups in past in Drambo with the Layers mixer but like others mentioned, CPU etc....

  • I’d like to give a +1 to the idea of a matrix router module. I really like that idea, and think it would fit well in the Drambo workflow paradigm. Fwiw I use the matrix module in miRack all the time, a functional equivalent in D would be fantastic.

  • I think both a mod matrix and a switch layer would make a lot of sense. Which one to use would certainly be a very personal thing.

    For the workflow style that @Horsetrainer suggested, switch layers would be the way to go.

    My personal preference would be to invest more work in a rather "universal" architecture controlled by a mod matrix, which not only allows for seamless switching but also for morphing certain parameters which would be quite beneficial for sound design.

    Both make sense for sure.

  • edited September 2020

    Edit: Using track presets and folders combined with a straightforward preset selection should be sufficient for now.

    Arrow buttons on the track header to cycle through all presets in the same preset folder would be great!

  • Somewhat tangental, but I know having a knob module with 3 knobs stacked would solve most of my overly-complex patch issues.

  • Yep, would like that feature on the knob, slider, trigger and switch trigger modules too. Just like the new N-1 module, if there was a + icon to add additional knobs, sliders, buttons, that would save A LOT of space. I'm working on a processor rack now with 16 knobs, but they take up half the rack in compact mode. Having 2 rows of 8 would look much better.

  • ‘What’s wrong with presets?’

    They don’t save midi mappings. 😜

Sign In or Register to comment.