What determines your choice of synth beyond sound?

[Uncle E]

I've made the same argument. It also sounds better than so many of the compromise mass market analogs of the late 80s. I'll take Diva over a Matrix-6/1000, Akai VX/AX series, etc., any day.

Wasn't the VX600 supposed to be quite good?

[ghettosynth]

I've made the same argument. It also sounds better than so many of the compromise mass market analogs of the late 80s. I'll take Diva over a Matrix-6/1000, Akai VX/AX series, etc., any day.

Wasn't the VX600 supposed to be quite good?

Yes, it is the exception. It is a CEM3374/3378 based synth. They're a lot more rare than VX90s and AX60/AX73s.

[Uncle E]

Got it. Yes, I'd choose DIVA over those other Akai's. I usually prefer DIVA over my Juno-106's. I prefer my ATC-1's over DIVA and the sounds from that SE-02 make me think I'll prefer it, too.

[ghettosynth]

Got it. Yes, I'd choose DIVA over those other Akai's. I usually prefer DIVA over my Juno-106's. I prefer my ATC-1's over DIVA and the sounds from that SE-02 make me think I'll prefer it, too.

Exactly. I've argued this point before that budget polysynths from the mid 80s were not sold as the "fantastic instruments" that they are believed to be today. They were cheap keyboards that were targeting the budget market segment. That doesn't mean that they're all bad, however, most of them were full of compromises in every way possible. Don't get me wrong, I still have a bunch of these, I just don't use them much or at all.

So, yes, if you're not going to invest in the really good instruments of the past or present, then save your money and go with soft-synths.

I much prefer the Juno60 over the 106, however, that filter is one of my favorites in my modular. I would generally prefer Diva to the 106 but might pick the 60 if I wanted to record some live knob-twiddling plus arpeggio action. My favorite analog is probably my Octave CAT with 2040 filter, it's all about that filter and filter FM.

[BONES]

Got it. Yes, I'd choose DIVA over those other Akai's. I usually prefer DIVA over my Juno-106's. I prefer my ATC-1's over DIVA and the sounds from that SE-02 make me think I'll prefer it, too.

Yeah, I don't get it at all, I'm afraid. DIVA is just the sort of softsynth I dislike - too complicated. The five pages of stuff in the bottom half of the UI ruin it completely. I may as well just stick with the LCD displays of my Ultranova or Pulse 2. In fact, both those synths are actually really easy to tweak and the performance controls on the Ultranova take it to a ridiculous extreme.

Then there's the fact that it's designed to sound like some old piece of junk, where something like Ultranova is just trying to be the best synth it can be. (Obviously, Pulse 2 isn't like that but at least it's an update from the original manufacturer.)

Looking at Diva's UI reminds me why I was interested in this thread in the first place - usability is ultimately way more important to me than sound. That's why I get way more use from Wasp than I do from Dune or z3ta+ - I can grab Wasp and make a sound in a couple of minutes, whereas doing anything in Dune or z3ta+ is really hard work and takes too long.

[Uncle E]

Yeah, I don't get it at all, I'm afraid. DIVA is just the sort of softsynth I dislike - too complicated.

Oh, sorry, I didn't mean that I don't like DIVA. It's the best softsynth for the kinds of sounds I'm after and I don't find it to be complicated (granted, I mostly ignore everything but the first page). I just meant that there are some areas where real analog hasn't yet been beaten, IMO, if that's the sound you're looking for.

With that said, try Minimonsta and Strobe 2, those provide amazing versatility with seemingly simple interfaces.

[foosnark]

My answers have changed a bit over the past year. I'm not doing tiny desktop synths and toys anymore (*); instead I've gone modular and semi-modular.

So now my real priority is how well will it complement my current gear and preferred workflow?

This covers a lot. Sound, connectivity, controllability, form factor, features.

Breaking that down: to me the point of hardware in this century is to do things software can't or doesn't very well. Audio rate modulation, other unusual modulation, interconnections between voices, building blocks that are rarely or never emulated. How many LPGs are there in VST synths? How about switched-capacitor filters, PLLs, modal resonators that you can feed from other audio sources? There's Reaktor, but I don't think it sounds as good and is nowhere near as satisfying to patch as the real thing.

The flip side is, there's all kinds of hardware I don't need because it's well covered by software. Effects for the most part (unless interesting modulation is in play), sequencing, sampling, some synthesis methods and categories of synth voices. I have no intent of ever ditching software just because I'm addicted to Euro(c)rack.

I have very little interest in fixed-architecture, basic subtractive synths, which are the are the majority of hardware synths and the usual argument for hardware "superiority." I've got a Microbrute -- mostly because it's also a decent, cheap and compact MIDI controller, though I do like its sound overall and it plays well with modular gear -- and far more software VAs than I will ever use. While my modular patches often have some subtractive hybrid elements, the last thing I care about is owning or emulating a Minimoog or MS-20 or whatever.

I find myself only using a few voices in any given track, and some of those are usually software, so my needs aren't great. Space is also a concern. So: no keyboards, and perhaps only one more desktop synth (likely semi-modular; preferably with USB MIDI so I don't need a Thru box). My modular end goal is a 9U rack plus a "lunchbox" or other small case.

I'm much more particular about software now too. A synth has to really stand out to get my attention, and I'm getting that way about effects too, finally.

Basically, I have what I need; all that remains is a few more things I want to experiment with in modular. Anything else is GAS and that's getting easier to resist.

(*) I do still use my Bastl Kastle or Olegtron 4060 on rare occasions, and I'm keeping my Thingamagoop 3000 because I can't part with it.

[aMUSEd]

What primarily determines my interest is something that for me is at least as important as the sound, if not more so (as I can get a wide variety of sounds from most capable synths), and that is performability. I think this is not given enough consideration in synth/effect design and where it is separates the merely interesting from the essential. For example this is by far the main reason for my love of plugins like Alchemy and Kore 2 - they both place just as much emphasis on intelligent and easily accessible performance features as they do on constructing sounds, what is the point of just being able to design sounds if a synth is a pig to perform with? Both Alchemy and Kore have customisable macro knobs and the ability to morph between saved variations of sounds (and in Alchemy's case also between resynthesised sounds) which vastly increases the possibilities in terms of realtime expression. Another synth that got this right from the start was NI Massive, on the other hand others either added such features as an afterthought and consequerently they are vastly under-utilised (Spire for example) or don't even bother (eg Tone2 synths). Some didn't even have a clue about bog standard performance features like sustain, modwheel, pitchbend and aftertouch (eg look at the first versions of Chromaphone or Rapid). Synths and effects need to be designed for performing musicians in mind, not just sound designers.

[rod_zero]

Over the last month I have been considering dumping my Analog keys and get either a System 8 or a Peak.

I really think plug ins are as good as HW sound wise, and I even prefer working with software must of the time: instant recall, multiple instances, more easy preset management, more flexibility, I can save a project and come back to it later without trouble.

The only point of HW is being hands on, more expressive since you have it all in front of you.

unfortunately while I love the sound and sequencer of the AK because of the UI I never program it and I always end up using the presets. This kind of depresses me since I really like it, but for using presets I better use SW. I also struggle a lot to incorporate it to my music as much as I work with its sequencer alone.

[JCJR]

Don't you get bored with it, though? I find that I need new instruments on a fairly regular basis, just to spark my creativity. It is far too easy to get into a rut and working the same way all the time invariably leads to everything getting stale over time.

Not trying to be argumentative for the heck of it, just a flip-side view-- I may be thinking of situations dfferent than yours.

Back when I was gigging, even in the 1970's, noticed in myself and guys I played with-- If we wanted a paycheck at our location in the food chain, we played a goodly share of songs we did not like but the audience wanted to hear. Or songs maybe we liked but did not enjoy so much after playing them nightly for months.

To alleviate boredom we would tend to buy new gadgets to make it more tolerable to play the same old songs over and over again. A kind of psychological escape mechanism. We were not completely miserable or we would have quit playing clubs and found real jobs but any steady gig can get boring after awhile.

Maybe like the nurse daydreaming about what kind of new curtains she wants in the living room while dealing with wall-to-wall gore and misery at her Emergency Room shift.

Even without the necessity of repeatedly playing tired songs for a buck, there was the tendency to buy new gear so that old licks would sound different-- As opposed to inventing new licks to play on the old gear. Perhaps it was simply easy to buy a new gadget but hard to devote enough work to invent new songs or licks.

Some devices are "shallow" with its depths quickly plumbed. Others are "deceptively simple" with few parameters but can take a lot of work to get good at operating it. Maybe in some cases a device would be perceived as "shallow" merely because of one's lack of imagination in finding all of its uses.

Perhaps most acoustic instruments qualify as deceptively simple-- Being as shallow or as deep as the operator has imagination and skill to draw out of the instrument. A nice Martin acoustic geetar can play the Hank Williams Song Book every night for decades. That same Martin can play some of the most sophisticated music imaginable. There's a wide range of licks a talented enough fella can make even with simple finger cymbal or wood blocks.

First time I recall being mind-blown by a "simple device" was reading the programming manual for lowly commodore 64. There were so many cunning details in the humble device, all the memory locations, OS, and 6502 instruction set that a fella could study it for years after the C64 is completely obsolete without completely plumbing its depths.

Some synths are merely "deceptively simple" like maybe an Arp Odyssey or MiniMoog. And some of them are so deep one could spend years in the numerous internal programming details. That is just from the sound design perspective. After one has burned out programming sounds, there is still the near-limitless different kinds of music that can be played on the axe.

I usually liked to "properly learn" a synth before getting a new one. Seemed wasteful to buy a new synth if an old one's depths have not been plumbed enough to at least "scratch the surface". There are many deep synths. The last one I got was a Roland FA-06. I only program it once in awhile. There is so much stuff crammed in there that at my current usage rate it will be years before I've explored everything in there. If one uses a computer patch editor, even the old lowly Sound Canvas SC55 is deep enough to hide a heck of a lot of sounds in there. Probably most of the big workstations are lots deeper than deceptively simple, and even deceptively simple can take years to plumb.

[BONES]

My answers have changed a bit over the past year. I'm not doing tiny desktop synths and toys anymore (*); instead I've gone modular and semi-modular

This tells me you are far more interested in the process than the result. I am the almost the complete opposite. I don't want to have to stuff around patching signal paths, I want that path to be completely obvious from the get-go and easy to work with. Even mod matrices annoy me because I have to do too many things to make them work. e.g. I want to look at the LFO and see exactly what it's doing and to what or, from the other end, look at the filter and see exactly what modulation is being applied.

You mention things in your modular that you never see in software but the reality is that you never see them anywhere because they aren't very useful. The last thing I want to be doing is sifting through a bunch of useless things to find the useful ones, which is why building my own VSTi in SynthEdit was just about the best thing I ever did for my workflow. I used a separate process to create the perfect instruments to make my workflow as slick and fast as possible. Each is relatively simple and does one or two things really well so I know which one to grab for any particular situation.

What is the point of just being able to design sounds if a synth is a pig to perform with?

That, of course, will depend very much on what style of performance you are looking for. The stuff you describe is of little interest to me. As long as Pitch Bend works and I can assign the Mod Wheel and After Touch to sensible parameters, which is usually more about the host software than the instrument itself, I'm golden for live performance.

The only point of HW is being hands on, more expressive since you have it all in front of you.

That can be at least as true of software. If I set up my KeyStep in front of my laptop, with the mouse in my right hand, I get every bit as much of a "hands-on" experience as I do from something like my Minilogue with it's (mostly) one knob per parameter. In fact, because of the way the preset management works in most modern hardware, my softsynths actually feel more hands-on, simply because the knobs on screen represent the current setting, whereas the knobs on the front of my Minilogue don't. In that respect, my Ultranova often feels more hands-on than the Minilogue or Monologue do.

I usually liked to "properly learn" a synth before getting a new one. Seemed wasteful to buy a new synth if an old one's depths have not been plumbed enough to at least "scratch the surface"...
... Probably most of the big workstations are lots deeper than deceptively simple, and even deceptively simple can take years to plumb.

That is completely true, which is why it was, on average, about five years between upgrades in those days. I think I got my SQD-1 in 1985 and replaced that with an ESQ-M in 1987. As technology was still evolving quickly, I got a KORG M1 in 1989, then an 01R/W in '92 or '93 (mostly because it had way more pattern/song memory), and a Trinity around 2000 or 2001.

I would be surprised if there was anyone on the planet you knew the 01/W as well as I did. I really liked some of the sounds so I got really deep into it's synthesis method. The wave-shaping it had was incredibly powerful and could be really expressive, too. I think it remains the most underrated synth of all time. Still, eventually I had plumbed its depths and run out of inspiration, so I moved on to the Trinity, which was more V/A like and, therefore, more easily understood.

The most obvious example for me was when I bought a KORG microX. I picked it up in Hong Kong, then went straight to the airport to fly home. In the 90 minutes I had in the Cathay Pacific lounge, I plugged it in and started working through the presets with headphones. Before I had to board my flight, I'd stumbled on a nice drum rhythm and a cool lead/pad, put them in a split, changed the effects to suit my style (distortion on the drums) and written the first song I'd done in a couple of years (eXterminate, on I Hate God). It was a combination of simplicity and flexibility, even for something as seemingly locked-down as a microX, that just sparked my creativity when I'd been in a ruts for ages.

Orion had a similar effect, only it lasted probably a decade before it started to level off. But in the first six months I had it, I did more than I'd done in four years before it.

It happened again recently when I started buying hardware. The Minilogue and Roland Boutique synths I bought seemed really good in the shop, but ultimately did nothing for me, but when I got the Monologue it switched that part of my brain back on like nothing had since the microX. Then the Rocket supercharged it and I've been on something of a roll for the past three months, when I had decided I'd probably give it up after we played in Germany. That's why I never underestimate the power of something new to get me going. You just need to find the right new thing.

I just realised as I was re-reading this post that the Monologue and Rocket also fit the mould of what I look for - simplicity. I'm already done with the Monologue, I'll probably sell it soon, but the Rocket is definitely a keeper.

[foosnark]

This tells me you are far more interested in the process than the result.

The process is a big deal to me. Patching is part of composition. But my results are better, and more prolific, than they were before. I am cranking out an average of 5.6 tracks per week, and about 3 of them entertain me later on. It's not like patching is slowing me down any; it's inspiring me to do more and has given me a lot of practical knowledge.

You mention things in your modular that you never see in software but the reality is that you never see them anywhere because they aren't very useful.

What's "not very useful" to one musician is essential to another.

The real story is that most synths, hard or soft, follow the "East Coast" architecture: VCOs (1-3) -> VCFs (1 or 2) -> VCA with a couple of ADSRs, and minor variations on that.

I'm mostly not a fan of the random bullshit that a lot of West Coast synth types embrace, but the synthesis methods are solid and criminally under-represented in modern hardware or software, outside of modular. VCO -> wavefolder -> LPG (or VCO -> LPG -> wavefolder) with a couple of slope generators is not rocket science, even if you throw FM or a saturation stage into it too. Triggering that slope generator at audio rates with another VCO for sync/waveshaping/octave division isn't hard either once you play with it a little.

There's kind of a standard bass/lead sound I often find myself using. Triangle with a touch of wavefolding, saturation and an LPG, with a sharp attack and exponential decay. It takes basically no effort to set up on my 0-Coast. I can patch it on my modular in about 30 seconds with 3-4 modules.

With East Coast synthesis, you can fake an LPG by using two slightly different envelopes (if the response curve is right, anyway) and a separate filter and VCA. I can get pretty close to my 0-Coast bass with Microbrute and Mini Slew, with nice saturation (but no modulation of the saturation) but no wavefolding. I can also fake it in Serum, with more effort. Most of my software VAs can't really come close though.

[HunterKiller]

...

Whether my girlfriend likes the thickness and texture of it. She's a great judge of character and taste.

[IncarnateX]

I have come to the point where new gear is not the answer to keep the inspiration going. I have a family to feed, so budget is tight. However, for the time being I find great inspiration in changing my ways with what I already got.

Not long ago, my studie set up looked like this:

Now it looks like this:



The rest is stashed under the bed.
I wanted to change from a slider-n-knobs jam oriented studio to a simplified production oriented studio and since I own the best workstation ever made imo, namely MC909, I thought: Why not just use this? With all music made internally on MC909, total recall, more hands on controls than any other workstation, and no needs to hook up other gear, things are as easy as working ITB with a PC daw. JD-Xi will be used for singing and vocoding only. I have even hidden my minilogue under the bed: I do not really need my analog style patches to be made with an analog synth. With my MC909's expanded 1513 waveforms of which many are from classical analog gear and a fairly deep 4 tone structured synth engine, I have more options to do analog style patches than I will ever have a chance to use. There are patches that my other synths can do better than MC909 for sure, but not to any extent that make them indispensible for my needs. When I have finished some tracks, I may dig up my synths and replace some voices before recording to audio, but for the time being everything is going to be made within this wonderful groovebox and be ready to be performed using this alone. It fulfills all my preferences stated in the OP, not a least because I prefer its size compared to workstations with keys. It is made in solid metal and aluminum, not hard plastic like modern Workstation. It has lots of real time controls, and up to 4 times as many recordable midi events than even the most expensive modern workstation. Thus you can make whole album on it just using its internal memory and have another ready on Smart Media Card if needed.

My neighbor which is a party-musician and a joiner/carpenter is willing to make my beloved box even more exiting. I have hired him to make a wood case for it with a build in Arturia keystep controller in minimoog style, that is, the MC909 will be standing up front instead of laying down and keystep, which is about the same length as MC909, will be its permanent keyboard.

[BONES]

That was always the lure of workstation synths for me. By the time I got a Trinity, I could do absolutely everything in one machine. In fact, the first thing NOVAkILL ever recorded was done entirely within the Trinity, down to recording my vocals on the multi-track recorder card I'd bought for it. Similarly, for the last 17 or 18 years I've been doing everything within Orion without the need for any external gear, except as ornaments for live performance. It was the obvious next step and one I am very reluctant to step back from.

The process is a big deal to me. Patching is part of composition.

I agree completely but it's a part that needs to be as fluid as the rest of it and often it is anything but.

But my results are better, and more prolific, than they were before. I am cranking out an average of 5.6 tracks per week, and about 3 of them entertain me later on.

That's another huge difference. There has never been a time in my life that I have pumped out more than about 5.6 tracks per year. And I never call them "tracks", they are either "songs" or they are just sonic doodles that may one day become a song.

It's not like patching is slowing me down any; it's inspiring me to do more and has given me a lot of practical knowledge.

As I said, you are into the process. I hate the process, it is nothing more than a means to an end and if I could press a button and have the song in my head finished and ready to perform live, I would pay a lot of money for that button.

What's "not very useful" to one musician is essential to another.

But if they were useful to the majority, they'd be in every synth.

The real story is that most synths, hard or soft, follow the "East Coast" architecture: VCOs (1-3) -> VCFs (1 or 2) -> VCA with a couple of ADSRs, and minor variations on that.

Yes, because that gets the job done. In the 80 and 90s hardware vendors tried all kinds of things but eventually they all basically fell back to that architecture. And they did it for a reason - it works.

VCO -> wavefolder -> LPG (or VCO -> LPG -> wavefolder) with a couple of slope generators is not rocket science, even if you throw FM or a saturation stage into it too.

An LPG is just a filter with modulation on it, you don't need a specific device to create the effect. From what I've been able to hear of wavefolding, it seems like a good way to add movement to your sound but I can't see it being any more useful than other forms of cross-modulation. It seems like a different way of doing the same kinds of things.

I can patch it on my modular in about 30 seconds with 3-4 modules.

But then you have to find the sound, so there's another 5-10 minutes. Whereas I can recall a patch I made six months ago, but couldn't find a good use for at the time, in three seconds. For me, instant recall is pretty much non-negotiable. The Rocket is an exception because it is so thoroughly well thought-out that changing sounds quickly is a breeze.

Most of my software VAs can't really come close though.

Give us a few audio demos and I'll make something functionally similar in a few minutes. It doesn't seem to do anything particularly magical to me.