Hello! I’m teaching some courses on synthesis this semester at the Griffith Conservatorium of Music. It’s a great gig and affords me access to some gear I would likely never purchase myself including two Doepfers which I’ve been forced to “test” as part of my “research”. Below are the results and beyond the jump so more thoughts about synthesis.
So with my impending tenure approaching I have been furrowing my brow at the latest developments and having sometimes fruitful but mostly frustrating discussions with other musicians about gear and software. Much of this discussion has stemmed from a post I shared on Facebook, originally made by Tom Ellard on his blog, regarding the re-release of Korg’s MS20 in miniature format. To summarise, Tom has been there, owned the originals and thinks it’s an uncreative waste of time. Much to no-one’s surprise, most of the world disagree with him. The analog modular craze it would seem is well and truly back with bearded guys all around the world throwing cash they might have previously put towards cars instead towards expensive boutique synth modules. I can’t say that this doesn’t please me but i’m in two minds about this Korg re-issue.
OK firstly, Korg are seemingly a very successful company in times of financial crisis. Given the popularity of the Monotrons and Kaoss variations it seems apparent that re-releasing a miniature version of an old semi-modular is actually not much of a risk at all. The fact that a noisy buzzing retro monosynth, will likely be one of the best-selling electronic instruments this year is pleasing to me as hopefully it might encourage more companies to take similar “risks” with modular synths. (Nord I’m glaring ominously at you!) However, something Tom is saying that people are not really listening to is the fact that they are paying essentially more, for less. I won’t reproduce his Mininova VS MS20 chart because frankly it doesn’t really sell his argument as both synths have completely different purposes. The Mininova is a cheap all-rounder VirtualAnalog Polysynth whereas the MS20 mini is a not quite as cheap or portable Monosynth that is likely to be used as predominantly as a filter or in noise-bands.
The point he is making that people are missing is, why does buying an 2012 miniature replica of an MS20 in 2013 require accepting the limitations of a 1978 synth? Some people will argue about the curse of the preset (I have before) but the reality is the companies like Buchla make a bunch of excellent analogue modules that allow for things like saving presets, using MIDI controller values etc. What you are getting for your $700 is an “analogue experience” complete with the original patch books (that actually came out with the VST version also). Being analogue the circuits are effects by actual climatic conditions rather than programmed oscillator drift. The question that each synthesist drooling over this new device should ask themselves is “what will this do for me?”
I can absolutely see why bearded modular guys will go for the MS20. For the cost of one or two regular modules they get a standalone filter with some great cross-modulation capabilities. I’m not buying one because i’m more interested in flexibility. I can use Korg’s VST which is sonically excellent despite being from a time when computers had much smaller screen resolution. After having played around with two Doepfer Dark Energy IIs I can honestly say I still prefer Reaktor where I can build passable emulations of the Dark Energy and the MS20 in Reaktor. Of course why would I want to do that? I want something flexible that I can create my own routing for, a synth I can design that isn’t just like every other synth. Short of a Nord Modular G1 or G2 falling in my lap I doubt i’m going to be adopting hardware synthesis as my primary vocation for a some time.
What I do know is that CPUs are getting fast enough to allow for oversampling, which helps deal with the problem of digital aliasing in the oscillators. My Ultralite Mk3 remains a sonic powerhouse and with the extra power I can potentially output 192k/24bit and higher. The Mk3 will even send Control Voltage if required. And there are some fantastic virtual modular systems out there and frankly if i’m going to make music I want as much control over the sound as possible – even if the end result sounds shit – it’s my shit! So my money this year is going towards a new computer so I can better run my virtual modular systems. These are my favourite VMs – all are creative and offer plenty of creative possibilities. This is the result of about 3 months worth of searching for the right tools – your mileage may vary.
For Sound Quality:
So stand-alone Vaz Modular 3 is the real winner. It has a diverse array of modelled oscillators and filters and a nicely saturated sound for software. Lots of extra modulators and math functions to play with here. If you want the most “analogue” experience and sound, forget Arturia and go for this one. You can even host other VSTs in it. One catch – Windows only. So either break out Bootcamp or load the VST in VFX.
If you are interested in more diverse physical-modelling synthesis and have some cash to throw at expansion modules I’d suggest Tassman 4 which has a more flexible sound than VAZ. Builder and Player sections are separated which is a source of controversy among users – many seem to hate it – I kinda like it but admit it is inconvenient to have to tab in and out to audition stuff. It is also much more difficult to get good results out of as the filters and amps seem to digitally distort quite easily. The Harm-Visser set is about the most beautiful sounding stuff i’ve heard from a computer and the Himalaya set is great value with lovely electronic-piano sounds and a useful drum machine.
For Ease of Use:
There is a reason why the KarmaFX is always featured in Computer Music magazine synthesis tutorials. It is among the easiest to use with separate modules connected together with easy click click routing and a pleasingly modern sound palette. There aren’t too many modules that you will get lost but enough to get creative. Has a drawable oscillator that Vaz and Tassman lack and some very nice FM and Additive modules.
Possibly usurped by:
On Mac and iPad 2++ and still in active development, Audulus is equally straightforward with an entirely visual, node-based approach familiar to anyone who has ever looked at flow diagrams. Modules are a little limited at the moment, however you can add your own Audio Units which makes up somewhat. I eavesdropped on a KVR Forum conversation where it seemed like the developer was hinting at users being able to build their own modules… now that would be nice!
If you ask me to suggest only one software VM – this is the one. It took me a while to get how powerful Reaktor is. Beyond the huge user library and the excellent paid ensembles that NI sell you will almost never need to click the EDIT button. But when you do – prepare for a rabbit hole. Reading Chapter 7 of the Getting Started pdf is your gateway. Only one or two steps down the Rabbit Hole you have realised that each of these ensembles is made up of instruments and each instrument is made up of macros. You can put these macros together yourself – build your own instruments. These macros are the same tools available in Vaz, Tassman, and even the Nord Modular. But there are 100s. Maybe thousands including those in the user-library. Here is a picture of my ridiculous Moog Modular rip off, made using the Modular X library.
Don’t misinterpret my use of the term “Novelty” as being frivolous or silly. I use it in reference to Aalto being unlike any other synth out there. It is perhaps the most like some of the DIY synths coming out from the MAKE and Dorkbot communities. It features oscillator and filter quirks coded into each registered version so each sounds distinctive. It isn’t quite fully modular, more a semi-modular like some of the U-He synths but the fact that stuff gets pushed through a “complex oscillator” makes it sound much more like a fully patched up MS20. It also has the best manual of any synth i’ve ever come across.
From the makers of Zebra which doesn’t quite make this list because I think Aalto is more interesting. Bazille is probably the best free multi-platform modular available. It free as it is part of an extended beta – however it works really well, sounds almost as good as Vaz, is very easy to use and surprisingly deep. You can get the Beta from the U-He KVR forum.
I’ve never been entirely clear why this exists. From what I gather it is a single voice from the G2 with all the programmability minus some multi-voice specific modules. There are some very deep synthesis workshops that actual use this demo as the basis for demonstration. No VST version exists and Clavia have repeatedly stated they are interested in hardware, not software. As a demo for Nord Modular it is astonishingly effective in my eyes. Yet they are discontinued. 😦 For what it is though, it is amazing and if you use Jack on Windows or Soundflower on OSX you can pipe the output into a DAW and record it! (It doesn’t work on OSX 10.7 and beyond though because it needs Rosetta)
Bonus Category – Modular DAEs:
So all these creative modular systems are all fine but if you are just using Ableton Live or Pro-Tools, are you really getting the most out of your music making. Digital Audio Environments buck the linearity of the studio inheritance and offer new tools and new approaches to performance, mixing and recording.
I feel like I really started making music with Audiomulch. The ability to route modules and plugins together and experiment with signal flow was really the thing that got me into making electroacoustic music. Forget tracks, choruses, arrangements – just plug a bunch of things together and record what it does! I feel like in those first 5-6 years before I started using Max/MSP, Mulch was indispensible but I feel like I outgrew it. As a step away from the standard way of doing things I would highly recommend it. Also there is nothing really like the Metasurface – a key that really makes Mulch one of the better live processing environments.
When I want to patch stuff together quickly of late i’ve been mostly looking to Bidule. It has an ugly GUI but is very versatile, has some nice algorithmic midi and spectral stuff built in and it runs as both a plugin host and a plugin. It’s a good way of hosting VSTs or AUs in hosts that don’t like one or the other and anytime I need to audition a VST or just create a simple patch it is the preferred option. Still… it isn’t a full music creation solution and aside from Max/MSP I really thought there was nothing else until I discovered Sensomusic Usine.
Basically Usine has been off the radar because i’ve been predominantly Mac-based for the last 5 years. One of the reasons I initially went over to the “dark side” to begin with was that I could use both OSX and Windows on Apple computers. With Vaz Modular being a windows app i’ve been using my Bootcamp partition for more than just gaming. When I read about Usine coming to OSX I thought I’d check out the demo in Bootcamp. Boy was I smitten!
To my mind Usine is the missing link between Audiomulch and Max/MSP. It features a mixture of DSP, Modulation and Mathematic modules that can be organised together in racks or grids. As with Reaktor and Max/MSP the idea is to build an interface supportive of the context be it performative or otherwise. Each module is a patch with signal and data nodes that can be modified and rerouted as you like. Plugins can be easily dropped onto the workspace and patched up as well but where Audiomulch and Bidule give you access only to the signal sends, Usine makes all the controller variables accessible as patch nodes. This is at times insane but soooo awesome. There is even a programming language that i’m staying away from at the moment. However there are already thousands of add-ons covering everything from synthesis, surround modulation and filtering to video playback and networking. It is fully setup to be used with a touch device… something that i’ve been trialling with Splashtop and the iPad and it works well but a bigger screen would definately make it a killer live tool. OSC, MIDI and other interfacing options are already present – the next edition promises the ability to use networking abilities so you can co-op over a lan.
So I have a new test for music software. I call it the “Open-Mic” test kinda like how comedians are tested by how well they can just get up and improvise. The test involves using the software and the internal laptop mic. If I can make a moderately interesting track just with the built-in mic then the software has potential. This is what I made the other night with Usine, the internal mic and some Melda plugins.
This is why i’m not buying a fucking MS20!