I've posted on this subject before to Analogue Heaven months ago, but, having spent the entire day patching out a Buchla 200, I thought I'd bring it up on here as well.

Do those of you with both nord modulars and real modular equipment find the presentation of modules affects your creativity differently?

When I work on the nord, I start from a blank canvas. Sometimes I modify other people's patches, but, for the most part, I start with a blank screen and start adding modules.

On the other hand, when working with a modular with a fixed set of modules, I find myself thinking ahead as to how I can implement certain features. While the nord has 110 modules, my ability to use them all is limited by dsp usage. While the Buchla here has 20 modules or so, I have no limitations other than the amount of patch cables, of which, conveniently, I have a ton.

So I find that when I do nord stuff, it's more pointed and specific. Sometimes I just start growing a patch, adding more and more until it sounds good, whereas on the Buchla, I'm just adding as many patch cables as humanly possible until I have some insane modulations going.

I dunno... If anyone cares to share their experiences on the subject, I'd love to hear it.

I don't have an analouge modular system but I work with the NM in the same way you work with your Buchla.

Now I'm in the same situation as with the Buchla; I can't add more modules, so I just start patching the things I've got. This usually becomes

Basically, I think you are talking about the design structure of a synthesizer. Of course this has an influence on your way of working. It's all the 'Interface'.

the real patch cables, like a modular Moog;

prepatched, like the Prophet 5

and no real knobs at all, like the DX7.

Photo's Gemeentemuseum Den Haag

Apart from their sounds, they force you in a way of thinking.

The difference between a piano, a harpsichord and a church organ isn't only the sound, but they demand a different playing technique too.

Another thing to take notice off is 'fashion' or 'time spirit'. Working with with cables seems old fashioned at time the DX7 was very popular. Nowadays I start to sweat when programming that 'chocolate bar', as Roland Kuit use to call it.

This weekend, I'm doing some studying on the DX7 and I was wondering why so many sounds are just slightly different from other sounds.

Here's the structure which 'forces' someone into a way of thinking: if you have to work with dropdown menu's, on such a small screen, it's much better to make a program change then scrolling to the parameter and tweak it with the cursor.

Emile Tobenfeld (a.k.a Dr. T) wrote:

Well, I don't know the sounds you are listening too. I did a lot of theme and variations patches on the DX7 because there is so little

real-time control. If you use some enharmonic ratios or fixed frequency oscillators, you can get away from the cliche DX7 sound pretty quickly.

The NM would be so much better if the modules used only the CPU necessary, not the maximum CPU allotted for a module. For example, if I add the percussion osc module, I might only use 1/4 of its functionality, but it will still take up a set amount of CPU.

> Another thing to take notice off is 'fashion' or 'time spirit'. Working with with cables seems old fashioned at time the DX7 was very popular. Nowadays I start to sweat when programming that 'chocolate bar', as Roland Kuit use to call it.

> The NM would be so much better if the modules used only the CPU necessary, not the maximum CPU allotted for a module.

Good point... I believe papa Pulver and I asked this of the Clavia development team that used to hang out on the list in the olden days, and

their response was something to the effect that if it worked this way, you couldn't modulate parameters in the modules since they would have to be in a static state. Plus, if you built up a patch that had lots of modules but only a portion of them being used and got up to 99%, and you tried to use a part of a module it could potentially cause a crash. I hadn't thought of this.

What might be possible (at least in theory) is to have several "states" of a module, ie, that you can select what features of the module should be able to be used, and use the CPU allotment after that. It could be dome semi-automatically with features being enabled/disabled by plugging in cords, and other parts being enabled with a setting in the module, that obviously is not possible to assign to a button or anything.

This would in practise means that you didn't need as many versions of some modules. So you wouldn't need x LFO modules, just one that could be put into different modes, using different amounts of CPU. But obviously, the amount of CPU used by a module needs to be "static" and can only be changed during edits, and not while playing. Changing the amount of CPU usage of a module will cause the same kind of "hiccup" as adding or removing a module does today.

The benefits would mainly be that there would be less modules to remember the details of, and a bigger granularity in CPU usage. But of course, it also means a complete rewrite of the whole system...

But then you'd be wasting another option in the modular: to make a patch with unconnected modules and wire them live... with your idea it would mean that at every connect and disconnect of a module the DSP's would be recalculated.. not very good... but for home use this could be acceptable, I think... Anyway, I'd like to have 20 times more dsp power!

OK, but this for me is almost the same situation as a NM loaded full with modules and only half wired.. I guess I subconciously need a limit to work up against.

(I think seriously that humans are uncapable of thinking logically about anything without limits) So for me every time I make a patch I'm working on a different modular system.

Things may change but the basis stays. If I want real structural change I start over with a clean screen. Maybe with a real modular system you start thinking about 'what modules do I still have' much earlier in the process?. Now that I think about it. I can see the difference, I think. It's like Lego

About one-line-of-characters screened synth modules... For me they realy change the way I work... first of all they're hardwired, so there is very little room for modulating stuff that modulates stuff that modulates. etc... They are realy small in cocept and very predictable. But with a modular system you need to imagine the whole patch so that you know

(At least roughly) which thing is doing what where. And a big frontpannel helps a lot I think.

Also... You encounter your own work and logic there too, you walk through the modules thinking "where did I put that module that time" and you start remembering the patching process, the way the patch 'works'..

In the past I invariably draw patchschematics on paper before programming a modular synth.

Several reasons, the most important:

I learned to draw those 'patchsheets' long time ago, a simple system with a circle for a VCO or LFO generator, a square for a treatment like a filter, delay or for an envelope generator, and a rightpointing triangle for a VCA or mixer. Just throw some of these on paper and connect them with arrows.

With the NM that has changed as the patch can now be conveniently saved.

This feature has by far been the most important one for me to buy the NM

immediately when it arrived in the shops.

But on many occasions I still draw basic patch schematics on paper before starting to do some serious NM patching. It helps to put things into perspective and give a good guideline. Invariably there will be changes in the actual patch itself, but I like to have that guideline before I start. And on paper you can annotate what you want.

BTW when the NM would work in a way that the DSP only has code for the parts of a module that are used, one can extrapolate this idea in a way that eventually one would have to write the DSP code oneself. Only then can you make the most efficient use of the DSP. But then you really have to know your algoritms! Imho the NM is a pretty good tradeoff between a barebones DSP-box with an assembly language development kit and a synth that is both quite powerful and very easy and quickly to use. After all its a musical instrument, not a computer.

And it is the 'philosophy' of a real modular system: If you have bought a module you have it. The fact that you can repatch cords in real time during playing relies on the fact modules behave as they do now.

Shane Etter wrote:

> Good point... I believe <snip> but only a portion of them being used and got up to 99%, and you tried to use a part of a module it could potentially cause a crash.

What might be possible (at least in theory) is to have several "states" of a module, ie, that you can select what features of the module should be able to be used, and use the CPU allotment after that. It could be dome semi-automatically with features being enabled/disabled by plugging in cords, and other parts being enabled with a setting in the module, that obviously is not possible to assign to a button or anything.

True. A better thing to do would probably to have "modes" on the modules that are manually switchable, ie basically merging different modules into one, and have no automatic selection of modes at all, since as you point out, that would mean you can't repatch things without interruption.

I used to use a badly hand-built modular system, which I eventually tossed for two reasons.

1. I hated having to pick up a soldering iron in the middle of a session.
2. Because I could never get anything back, whenever I had something interesting, I had to use it in every piece it might fit before doing anything else. Indeed, by the end of a take, it could be almost impossible to get back to the state before the take.

I'm finding that the Nord has a related (but not as bad) problem. If I load a really complex patch that i haven't worked on in a couple of weeks, it can take me quite a while to figure out what everything does. This is particularly true with generative patches, which the Nord is really good at.

I'm finding this out the hard way as I'm spending all day trying to make a well documented (so I can use them without computer) set of patches to use for jamming.

Usually, with the Nord, I open a patch (one of my own, or someone else's). After I do this three times, I've almost certainly found one that hooks me and I start tweaking. I think I should start with a blanker slate more often.

Unfortunately, pasting modules with knobs attached doesn't paste the knob assignments.

Emile Tobenfeld said:

> If I load a really complex patch that i haven't worked on in a couple of weeks, it can take me quite a while to figure out what everything does. This is particularly true with generative patches, which the Nord is really good at.

How true. I have tried to solve the problem by labelling the modules correctly and always putting given modules in given locations according to their function. This doesn't always work, though, because the philosophy of saving to prevent loss of valuable work in case a crash occurs at random means that pretty much as soon as you've loaded a module and like it you hit save without naming it properly.

>Something I could do more of, is make a library of building block patches -- very simple things that can be pasted into another patch as needed.

Yes. The old 'macro' approach, where you can gather up all the modules that perform a specific function together and just sort of wrap them up into one kind of customized super-module of your own designing and that then sits in a library that you can either access in a drag-and-drop fashion from your personal library window, or else possibly even drag down from the editor's menu bar.

> Perhaps someone (with more time on their hands than myself) should start a page containing only this category of patches.

That WOULD be nice.

> Unfortunately, pasting modules with knobs attached doesn't paste the knob assignments.

Yes. Another little annoyance. Would be so easy for a little dialogue box to pop up under such circumstances and give us an option about whether or not we want to replace assignments, and if so which ones.

Still... others have said this before. Just repeating them in case Somebody Is Listening and they haven't yet made it into Version 4!!!

To assign a Knob isn't that difficult: (besides the dropdown menu's) just click the parameter, push Edit, grab a Knob on the Synth Panel, press 'Assing' turn Knob and ... voila!

To me it gives me a better overview which Knobs are availlable. I wonder how many Patchers use this way of assigning Knobs? Works only on the Modular, though.

Ehhh, is this still Philosophy? I think so... Editor and Panel working is "dualism"!

But sometimes ... it would be nice to simply copy a module and immediately also have the option of copying over any knob assignments you have made, also. I am just a very lazy man, that's all.

I thought Editor and Panel working is 'mono-lism' or whatever, and that when you have Interface problems then it is 'dualism'.

> To assign a Knob isn't that difficult...

but if you have pasted something that uBes half a dozen, getting them all right can be annoying,