Soft Synths, Analog Emulations, Noise Machines, Modular Synths – many choices of tools these days. I was humbled when a friend pointed out that my article Analog vs. Digital Synthesizers – My Take on the Old Debate made it to the very top of Google searches for “analog vs digital synths” (a Bob Moog interview was second – not a bad company to be in…). Sure enough, I have received tons of emails asking for my opinion on synth choices, creative synthesis approaches and about Studio CS gear during the past years. The article was cited in forums, in education, re-published on various websites, and visitors shared their great points and stories in the comments. Now we’re here in 2019, about six years later; a lot has changed on the electronic musical instrument landscape – or has it? The evolution of electronic musical instruments isn’t an exception from Moore’s Law from a technical aspect, but how about the evolution of their musicality? Certainly not doubling every two years; nevertheless, a topic interesting enough to explore in a sequel article. So, let the synthesis begin!

Even the emulation plugins are now sounding good – really good! But…

Thank to the increase of the average household computer’s processing power, to the greater efficiency of code and the hard work of dedicated R&D teams (fueled by a much-increased competition), we have reached the point where analog modelling plugins can actually sound damn good. Not only they provide high-quality sounds (like static sample-based rompler plugins already did 15 years ago), but now they act well, too: finally, some of the filter designs have gotten so convincing that it’s hard to distinguish a “sweep of the saw” happening in the box or via voltage in the transistor or diode. It’s impressive how close their sonic action has gotten to the classics they are often modeled after; just recently, I was beta-testing and creating presets for Brainworx’ brand new bx_oberhausen plugin (an Oberheim SEM emulation with expanded-functionality), and it really is super hard, if not impossible, to distinguish its sound and modulation behavior from my classic analog Oberheim synths. It was quite easy, and (for a plugin), unusually enjoyable to create uber-convincing presets for it – just listen to the character of the arp and the warm filter-action in this sample that I made exclusively from three of my own oberhausen presets (aside from reverb, no additional effects were used):

More and more software companies realize that the analog character is as much in the behavior, as it is in the emulation of the sound itself. Meanwhile, new plugins that are not meant to imitate a classic sound come out every day, and are just as good and exciting as they have ever been (I can’t wait to see what Magnus Lidstrom is cooking up next). So far so good!

However, a recently (seemingly) popular idea of software companies is to release emulations of early digital synths. Aside from price, I don’t see the value – why would you buy a basically sample-library-rompler-copy of a digital synth whose character mainly came from the patented (i.e. uncopiable) way you programmed it, interacted with it, and how its (often unrefined) components shaped those compressed waveforms through their efficient algorithms, (usually) cheap reverbs and early D/A converters? The classic Casio-s, the D50, M1…even the Fizmo now come in unofficial plugin forms with fake names, matching colors and font types (reminding me of the Recebok shoes sold in Chinese flee markets for quarter of the price of real Reeboks…).

People don’t seem to understand that it’s not the static samples of a synth that make it special, it’s the way you interact with the instrument. Click on image for more.

What I’m surprised about is that most major companies are still largely ignoring the interface through which we interact with our instruments (read more about the great importance of this here). The default way the majority of people trigger plugins is the mouse (sadly, still alive and well in 2019), or maybe a run-of-the-mill feel-less keyboard, on a better day. Programming music by drawing static notes and linear automation lines in the MIDI editor is still the favorite way of most lazy “producers” – while the point of any musical instrument is to provide the performer with a seamless performance tool through which to express, and sometimes even expand, a musical idea and feel.

There have been some pioneers pushing the boundaries; some even too far off without making enough sense or doing market research (roli blocks, anyone?). By now, I was expecting the larger players of the industry to follow the late Robert Moog’s vision with his theremin: music-making and instrument design to gravitate toward performance. That is, instruments that offer a flexible human interface and sound generator in one, integrating form and function into the same self-contained instrument – or at least connect to a dedicated software (like the Continuum, which has been successfully doing that since 2002!).

At least, there is more hope with controllers; I’m referring to those with usable interfaces, like the various wind controllers, the Seaboard or Enhancia’s Neova Ring (soon we’ll see how naturally it tracks finger movement) – but not referring to gimmicky toys like the AlphaSphere and the wi-fi MIDI gloves that rather belong in the arcade or circus than in the studio (sorry Imogen, you’re great but those are toys, indeed). Interestingly enough, most of the promising controllers and devices transmitting refined musical expressions come from small companies and startups. Sadly Yamaha gave up on developing advanced breath-controllers and I guess Roland had dropped its We design the future slogan for a reason…

An interesting question to debate is whether we need controllers that offer “new ways” to control our sounds, or rather new controllers that translate our well-defined and well-practiced expressions more accurately. While the former has the advantage of inspiring new expressions (therefore new kind of performances, sound design, and potentially even influencing the composition process itself), the latter could get us closer to finally expressing our musical intention and concept the way we naturally imagine it to being with. Are we going to see a split where developers either aim to teach new gestures to sonic experimenters, or learn acoustic instrument playing techniques themselves to create products for the seekers of organic performance gestures?

Either way, when it comes to alternate MIDI controllers, another limitation is the language (protocol) itself that is trying to describe complex and precise gestures: MIDI has changed little since 1982, and while it has been an amazing standard with a major impact on music creation (good and bad alike), its limitations now really get in the way of transferring high-resolution expressions between controller and host / generator. Control voltage came back in the 2000s, but requires the use of physical cables. Therefore, most new controllers opt for wireless (sometimes proprietary) protocols, which then limits their connectivity to the other components of the studio (not counting the option to down-convert to MIDI’s 128-value resolution). I love MIDI but it badly needs to evolve.

The good news is that, even though MIDI 2.0 has been talked about (and rejected) for many years, it seems to be becoming reality during the next year… or two, according to new announcements at NAMM. The new specs look promising (increased resolution, more standardized expressions, system recognition, full backward compatibility, etc.), but so far surprisingly few companies are showing support in form of product development.

Minimalist technology and toys: can simple sound great?

Hack, yeah. The growing circuit bending trends of the 2005-2015 decade (usually with questionably useful results) has been replaced by the “returning to the retro cheap stuff” trend of the recent years: finding new use for old, usually simple gear. Whether it’s a toy synth or a low-bit processor or a super-basic old instrument with a distinctly “bad” sound (i.e. not bad, just grainy, grungy, dull, thin, whatever… character), reaching back to simplicity is often the answer to rewarding new approaches and sometimes even new intent (let this sink for a bit… limitations of old technology driving new compositional approaches today… brilliant!). That simple old Casio sound processed with modern effects, or even just left raw but placed into a new musical context, helped many forgotten models to gain the appreciation of early re-adopters (and that of smart eBay sellers). This isn’t really a new idea; think of the intentional misuse of technology for creative results since the ’40s, or the re-emergence of the TB-303 or the SID chip in the ‘90s. But now the story of rebirth (pun intended) isn’t tied to a particular genre or style, rather to an “it’s so bad that it’s good” aesthetic – which is not a surprise, given the all-too shiny and overly “perfect” sound of today’s productions. What better way to add contrast than by throwing a simple and unrefined sound into the arrangement?

The Junk, the Stupid and the Badass. Sometimes it’s hard to tell the difference by looks. Click on image to compare.

However, there is another trend gaining popularity, infecting musicians who are often trying too hard to be different for the sake of being perceived different: it’s the “make any noise and convince yourself that it’s cool” disease. Smart companies feed the snobs and bored bedroom sound designers with literally anything that looks unusual and makes any noise – sometimes literally just distorted noise that’s coming from a $300 box (which cost $10 to make). The victims believe that these boxes are ‘freaking-awesome cutting edge stuff’, thank to their clever marketing and nerdy presentation. In reality, these are just boxes with a couple of transistors and pots with esoteric sounding parameter names, and with knobs that actually do nothing special but add distortion, bit crush, white noise. Sometimes I read their specs, written in a pseudo-engineering language, only to amuse myself. As expected, their primary target audience is modular synth fans – but more about this later…

Affordable, educational and badass little synths

Now these developments are amazing. Today’s technology has made hardware synths so affordable to manufacture (while a competitive market has influenced business models to aim for lower profit margins), that the first time in history, anyone can get a POWERFUL sounding synth for under $300 – brand new! This includes true analogs, half-modular synths, classic synth reissues, new concepts and some really great-looking models full of knobs! It’s just astonishing what sort of sounds you can generate with these very capable little instruments. Kudos to Arturia for starting the trend on the large scale. The only drawback in this price range is versatility, but I expect that the new synthesizers of 2019 and 2020 will address that as well.

Used in education, sound synthesis can be finally taught without major investment for the students or the school, and at least with some preservation of performance expression (as opposed to just clicking around with a mouse). Being also a teaching professor of electronic music production myself, I always look forward to hearing the work of students who are completely unexposed to electronic music and suddenly get to interact with real, physical electronic instruments the first time (well, I kind of force them to drop their mouse…:).

The modern analogs keep coming

The comeback of analogs started in the early 2000’s and it has only accelerated since the Moog Voyager. As the owner of several vintage synths, it’s a bittersweet feeling to witness how the old classics are reborn in retro-looking bodies, sometimes with original, sometimes with almost perfectly identical sound (and sometimes not… oh, Behringer…). Often the new cheap versions even de-value the vintage models while diverting the attention from the classics’ real value: their imperfections, which are rarely replicated.

These new imitations are not the same as the vintage models, but the reasons to spend time, money and frustration with keeping the old favorites alive (like my Yamaha CS-60 which is still unmatched by any modern replicant) are getting harder to defend. Dave Smith still keeps putting the magical character into his synths today (and included a wonderful tuning-fluctuation feature that I had been asking for for ages) – his synths have a unique vintage flavor but with the option to sound “modern and clean”.  So many companies have gotten into the Modern Analog game recently, that we actually have more new analog synth models to choose from today than ever before in history(!).

For my taste, most modern analogs are still a bit too clean (see why, here), but can sound more powerful than any other synth. As always, there are bad examples, too, riding the waves of reputation, but in fact being cheap reproductions that only hurt the name of the original. My respect goes to the inventors of new and cool little analogs, less so to the businessmen who reverse-engineer and release cheap copies of classic analogs and market them as identical. Why would anyone prefer to cheaply imitate when they could rather create the new (analog) sound of the future?
However, it’s only natural to expect the same character and features of a classic instrument from its modern re-issue, but actually, it’s pointless to compare. Who cares if the OB6 sounds exactly like the OB-XA or if the new and old MS-20s and Odysseys are perfectly identical? Unless you’re in a Journey cover band, you shouldn’t. If it sounds good without a nameplate, just play it.

Modular synths – new tools or just an overhyped trend?

A polarizing topic that’s hard to ignore – but it often makes me think and adjust my views. On one hand, modulars can be seen as cool, tangible variations of csound, pure data and the like, with patch chords and actual knobs taking the role of code, nodes and the mouse in a flexible environment. On the other hand, I’m getting sick of tech snobs who try to convince unsuspecting musicians and coolness-factor & trend-driven sound designers, that modular synths are a brand new way of making music that’s superior in some way… which of course is completely false. What also makes me cringe is when self-proclaimed “synth wizards” talk about “modular synthesis“. Someone please explain to them that such term is nonsense; the choice between assembling a setup from modular elements as opposed to using an already assembled hardwired synth will not effect the type of synthesis (subtractive, additive, fm,  etc.) the machine employs. If anything, “modular synthesis” should refer to the practice of using a signal processing chain in which your audio and control voltage can be flexibly routed through blocks of function components – which is a similar idea to what you can do in any Digital Audio Workstation with a few plugins: route with total flexibility.

Are modular synths functionally inferior to non-modulars? Of course, not. They are just different machines, and definitely not meant for everyone. They are (mostly) flexible tools, and when it comes to routing, many consider that flexbility a source of freedom that allows the user to break traditions, expand the sonic possibilities. They enable the user to custom-build their setup, which can look and feel very tech-y and flashy (although often quite gimmicky – a popular choice of aesthetic these days for pseudo-deep live performances). If the setup is in the right hands, I’d agree with most of the benefits, though there are way fewer expert hands than eurorack modules around.

Modular analog synths are far from being a new concept; the first modulars originated technically in the ’50s, and become available to a somewhat wider audience from the ’60s on (a’la Buchla, Modular Moog). Doepfer has been building euroracks for decades… this miniaturized standard has been around since the mid-nineties! The modular synth-building concept itself isn’t very unique either; actually teenagers have been using “modular synthesis:p” in a software form without knowing it: just think about the good old Reason, or its grown-up versions like Reaktor, Nord Modular, VCV rack, etc. These days they’re available for free or cheap (although the Nord G2 is still $2G for other good reasons). These systems allow a completely modular approach to synthesis in the digital domain. Or, you can take the other end of the spectrum with CSound, Pure Data, the Kyma Systems, or the EigenMatrix for pretty insane flexibility – for those with lots of time and patience. Either way, modularity is a neat but old concept. So then what’s my issue with the modular snobbism of today and why had I decided to get out of the modular game just in time before it exploded a few years ago?

Modular synth? Not a new concept, but enjoys a new wave of marketing for sure. Click on the image for more.

The answer is simple: for me, a typical modular setup’s limitations exceed its advantages – when used to create the type of result I consider humane-expressive music. Today’s re-emerged (usually eurorack) analog modulars are somewhat limited when it comes to creating a full musical arrangement: once you have assembled your purest oscillators, special-order filters and trendy-as-of-last-week modulation and effect modules, you can successfully generate one (and often only one) part of your music. You want to hear what the second part sounds like (or perform it live)? Buy a second set of oscillators or signal splitters, filters, and pretty much everything else to play part #2. Have a complex arrangement? You will need deep pockets.
(Those of you who are making the clever argument that modules can share multiple input signals and split control voltages… or speak of the creativity in committing to the sound and recording each part without audible context or the ability to modify: I am for limitations, but wait a second… wasn’t flexibility the main point of connecting modules with patchcords to begin with?)

Now, introducing the term music opens up discussions about modular synths to major controversy and debate, as it can be interpreted in as many ways as many people there are on Earth. So let me clarify: I’m referring to the type of electronic music that goes beyond just featuring cool sounds, but also communicates ideas, feelings, thoughts, emotions between the composer and a (hopefully) musically receptive (and typically non-technical) human audience via its deliberately composed notes, rhythms, potentially harmonies, dynamics, structure and performance expressions. Don’t get me wrong, I enjoy listening to some musique concrete (and even teach it in my electronic music class), and, I do like the work of several modular synth performers. However, what I’m referring to here is the modular fanboi, creating an interesting sound, then thinking “hm, this sounds like a bird with a diarrhea, so let’s name this composition diabird“. Without any concept, emotions, message, only by recording some random sounds and giving it a weird title, the outcome won’t necessarily gain aesthetic value as music. Here is a typical (even curated!) example of some beautifully emotional, musical modular performance (not!):

If you think this is awesome, you might want to start exploring some actual electronic music.

Even a small modular system can be suitable for creating some beautiful tracks or even emotionally rich pieces. However, while limitations are great for creativity, too essential types of limitations can also seriously limit the potentials and scope of the expression of an idea (such as the structure of musical arrangement, musical nuances of a performance, etc.). If I want to express a feeling with a 5-part harmony, but I only have 4 voices, or if I suddenly feel the urge to add vibrato to a note I’m holding but vibrato isn’t patched in, then technology is not just keeping my work minimalist but is also compromising my creative ideas and expression.

Modular Tinkerers or Composers?

I may sound like someone who’s against modular synths; I’m certainly not. But I am against the unfounded hype that they get these days. Where I see a cool benefit of having a modular setup is a studio that incorporates some modules into a more flexible and capable setup. Controllers will likely expand their musicality in the future, and stand-alone modular performance systems will shatter their current limitations.

Funnily, many users of modular synths produce close to nothing complete with their equipment. Based on my own observations and experience with modular synths, countless conversations with performing colleagues and friends alike (and after having been the co-organizer of a synth meetup that ended up turning into a modular nerdfest), I have found that the combined output of about seven (7) hard-core “modular only” users matches the musical output of one (1) typical electronic music composer / producer. This is not the case because modular work takes more time, rather because exclusive modularists are often rather tinkering enthusiasts without a message, than musicians with something to say. Another major aspect that’s usually missing is performance; not the knob-turning action (the same kind of action that modularists criticize DJs for), but the intimate connection with each and every note of an instrument that you play. With that said, nothing is wrong with having an amazing hobby, like cooking up interesting sounds, tweaking knobs for hours and drive your neighbors insane, showing your patches to like-minded individuals and keep expanding your setup with new toys until it looks like a misplaced Christmas tree. Seriously, modularism is as entertaining of a tech-hobby as assembling computers – but just like putting together computer parts doesn’t make someone a programmer, connecting synth modules with patch cords won’t turn most users into a composer either. There are some notable counter-examples, though, creators who have managed to find their sound without the need for a concept or without much of an expressive interface at all.

Speaking of an “experimental mindset of modular performers” is also a grossly broad generalization. I remember the virtual analogue snobs from my early days of beta-testing new synths and demoing them at trade shows: they swore that their Nords and Waldorf VAs sounded more analogue, warmer, more powerful and had more modulation options than the Roland and Korg romplers of those days. And they were right. However, they were completely wrong about those synths being any better tools for music composition or sound design – they were just different tools. (Back in the mid-’90s I worked with Roland, and watched EDM-loving audiences deeply touched by performances and the lush preset(!) sounds coming out of the JD, MC and XP synths at some major shows, while the Waldorf exhibitors only attracted some nerds and JDs with their thick bass lines that kept looping for hours without any change but a filter sweep. (For the record, I am not taking any stabs at the Waldorf brand; I like Waldorf synths quite a lot.) The same is true for modular synths, which are just another neat tool; however, I believe that due to the less musical, more technical approach to their use, and their limitations in polyphony & multitimbrality, they are not any more inspiring sound synthesis tools than a complex synth, and are actually less capable tools for creating complete pieces of musical arrangements than the average multitimbral hardwired or software synth today. For others, they can be quite inspirational, though.

I can appreciate the freedom you get when picking and choosing the components of your modular setup and keeping the whole thing in a relatively compact form. It’s interesting to note, however,  that 9 out of 10 modularists have only basic subtractive synthesis components in their eurorack: VCOs, LFOs, a couple of EGs, VCFs, and a sequencer. Hardcore modularists also use the sound design argument, namely, that you can do more with modular gear than with “conventional” synths. This is another statement that may be true – or may be completely false, depending on the units you’re comparing. While it is true that experimenting with a wall-sized super-modular system can be liberating when I comes to designing new sounds, there are also non-modular synths (both in hardware and software form) that allow you way more flexibility than a base eurorack loaded with conventional synthesis modules (think of the Doepfer A100 vs the Waldorf Quantum – or the Hartman Neuron, for an extreme example).

By the way, connect the output of any synth (or computer) to any guitar effect pedal… congratulations, you have just made your first modular synth! You can get some really neat stomp boxes for $25 on the used market. If you want to keep things flexible and unique, why stop at guitar pedals? Plug your household electronics into your studio (impedance matching advised:) for some truly unique sonics. The idea “the studio is your instrument” comes from the ’60s and still rings true. Heck, turn your entire house into a big modular system!

This forum post shows that some people turn the art of creating music into a tool-collecting competition – how much more sad can culture get?

Missing the Point

There are many examples of over-hyped technology and product development trends that largely miss the point. The worst offenders on the top of my list are specialty mixing headphones (nice try), audiophile cables (waste of money over $10/m/channel), automated mastering algorithms (they should all disappear) and “intelligent” music mixing software (end of creativity)… and for the average user, modular synths. I see a place for the trendy and ever-more-complex tools in the sonic wizardry scene (which, as a sound designer and beta-tester myself, I respect and have been an active part of for decades), though I believe that these tools actually put many users off-track, musically speaking. They are sold as inspiring electronic musical instruments, but most non-professionals end up using them for the enjoyment of a technical process. We do need hobbyists and software engineers who enjoy spending their time with comparing waveforms on oscilloscopes (like those pixel-peeping photographers who take more DSLR sensor test images than actual photos) – even if their work takes place with no musical goal in mind. But for creators and performers, getting wrapped up in the technicalities of the process can easily lead to missing the big picture: the photo, the music, the story.

Technology, including all type of synthesizers, should be the tools not the purpose in the creation of music, sound design, sonic experiences. Cool sounds, textures can be extremely inspiring, and sometimes the more unique, the better. They can become small elements or robust building blocks, and even the main focus of music. But, creating sonic elements that never end up in a piece of music or a concept communicable via audio, is not writing and not performance. In my opinion, sonic experimentation is crucial in today’s electronic music, at the same time, experimentation only, without any intent or feel, remains no more than an act of toying around.

So, before we talk shop about oscillators, synths and music technology, let’s try to answer the most important question: Where is the Music?

Read the original & 2020 followup articles:

Part 1: Analog vs. Digital Synthesizers – My Take on the Old Debate

Part 3: Analog vs. Digital Synthesizers – Part 3: The New Wave of Wavetables

“Analog, digital, virtual analog, software synths… which one is the best option? Which ones sound the best? This is one of the most frequent questions I get from my listeners, students and even friends from the music industry. So I thought I’d post my thoughts about this in a very condensed and non-technical form. Obviously, I won’t be comparing specific models or talk about the exceptions in this article, rather discuss the benefits/disadvantages of the four major categories in general (as for professional or advanced amateur studio use). So let’s go from the most fundamental to the more recent concepts.

Can a digital synth sound as good as an analog one?

Define good. A digital synth can sound horrible or it can sound way more interesting than an analog, depending on what type of sound and character you’re after. They both have strengths and weaknesses, and it simply doesn’t make sense to compare them without talking about some specific characters. For example, for sounds with a truly vintage character, for soft-saturation-rich sounds, for naturally unstable tones or somewhat uncontrollably fluctuating parameters, analog instruments will have an advantage, while most acoustic instrument imitations will sound quiet pathetic coming from analog electronic instruments. Besides actual samples of acoustic instruments, many digital synths will give you access to non-traditional parameters, diverse control options including MIDI, other than subtractive synthesis methods such as stable FM, PM, PD, additive, wavetable, granular, etc., and modeled analog component designs (algorithms), which might be unfeasible or practically impossible to build in the analog domain. So, get that old digital synth down from the attic.

Virtual analogs never sound as good as real analogs, right?

Wrong. Again, define good. The individual character of a virtual analog will likely come from different elements/attributes of the sound it produces than those of real analogs, and if that’s what you’re after, they might very well sound bigger, fatter, harsher, punchier than their analog relatives. Built-in effects may be available as part of the sound design process (as opposed to just system effects) in many of the virtual analogs (think of the Virus or the Supernova), and can often be controlled and routed in unconventional ways and interact with the more traditional attributes of sound. While virtual analogs will certainly have their own character, they will let you create sounds with a different character quite easily, as virtual analogs typically offer a large number of editable parameters, the non-traditional combinations of these, and highly flexible routing and control capabilities. Classic analogs have their individual character, which is fantastic, but might be somewhat more limiting when you’re trying to create sounds that don’t necessarily feature the typical character of a given model.

If you are using a VA and going for a classic”imperfection” (of steady pitch, amplitude, control and dynamic behavior) of a true analog sound, an often highly desirable character, you will have to imitate this imperfection by randomizing the value of several parameters– still, it just won’t be the same as the real thing. While the “warmth” of an analog synth can be described with an algorhythm and added to a virtual analog sound, a monotonously consistent (sampled) analog character or a digitally pre-programmed change of that character just won’t give you the same sonic quality and feel as the dynamically (and often randomly-and-dynamically) changing warmth and imperfection of a classic instrument. These exciting imperfections result from minor, unpredictable voltage changes, unrelated parameters’ coincidental effect on each other, use of a less consistently acting keyboard, difference in design (often no consistency there with hand-built models) and degree of stability between the oscillators, temperature changes, etc. Not to mention the unpredictable effect of a true analog instrument’s mood at a given time, i.e. what kind of day it’s having…

Do the new versions of vintage analog synths, those with fully analog signal paths (like the Voyager, MS-20 mini,  OB 2-voice, etc.) sound the same as their original versions?

Comparison test (including my own) show that they never sound the same. First of all, most units of classic analogs don’t sound the same to begin with, while there is very minimal difference, if any, between the mass-manufactured units of today’s analog synths (with a couple of exceptions). These synths can be really great as their analog signal paths and components give you the raw, warm, forgiving, “pushable”, powerful sound you might like, but with character stability, digital control and convenient features (flexible routing, performance controls, MIDI, software interface, patch memory, etc.).

Yet, sounding different doesn’t necessarily mean that they sound better or worse. More stable, yes. The question is whether you like a more or a less stable tuning, and a more or a less consistent reaction (amplitude, pitch, filter amounts, dynamics, etc.) to your otherwise consistent performance. Whether you like a cleaner or a “dirtier” analog output? In other words, when recording it, do you prefer more of a “the same notes played repeatedly never sound the same” behavior, or a more predictable, consistent sound? If the latter, recently (post-2002) built analogs (or most VAs) will be your best choices.

Another factor that might greatly infuence your choice is the synth’s ability to store your patches or settings. Today’s analogs often offer this functionality – for most, this is a clear advantage (more about the rest of us at another time). Lastly, besides the differences in sonic character and features, classic analog, new analog, and virtual analog instruments feel entirely different as a whole; the materials they are made of, the feel of their keyboard and controls… and of course, we all have our individual preferences between numeric displays, hi-res lcd screens with lots of details – or no screens at all.

Softsynths are just like digital synthesizers, but thank to the host computer they run on, they have more memory, handle larger samples, have more polyphony, offer higher multitimbrality, feature better interfaces… overall they are just better.

As far as the specs: yes. Better overall: definitely no! It is true that the numbers make software synthesizers seem like clear winners (especially if you don’t plan to show them off on stage), but I would argue that they have at least two major disadvantages, even when compared to digital synths.

The smaller disadvantage is that the sound they create either stays in the digital domain, or gets converted to analog signal by the D/A converter of your computer’s sound card or connected audio interface. This might not be a disadvantage for one or two sounds, but when 10-12 different sounds come from 6-8 different plug-ins or software, and they all go through the exact same host application and the same hardware’s D/A conversion, they might (and often they do) get a bit processed, “homogenized” the same way. Think of this as an extra spice, an extra character, which becomes part of every one of your sounds, making them sound a little more similar to each other than they did before conversion. On the contrary, if you keep them in the digital domain (the sounds never leaving your DAW), or if you have a very high quality interface with highly transparent D/A converters, you will end up with no added character in any of your sounds, at least in theory. This might be exactly what you want, but personally, I prefer to get that little bit of an extra sonic diversity, as long as it comes from different D/A converters, pre-amplifiers and other components for each digitally generated sound. I have blind-A-B tested the digital (adat, s/pdif) and analog outputs of several of my digital and virtual analog synths, and in 80% of the time I preferred the analog signal. (Surprisingly, the 20% of these synths where I opted to use the digital output, were virtual analogs!)

This is not to say that the advantage of an analog signal is that it’s warmer, less sterile, or in any way better than the digital signal – in fact, in some cases it’s less “clean” and less dynamic.  But, the analog signal is usually a bit more exciting, as its character isn’t as consistent as the digital signal’s more “always perfect” character. Letting favorable accidents (like analog distortion, signal degradation) happen can lead to unexpected (good or bad) character in your sound – and now we are talking about a creative element of sound shaping, as opposed to just sonic quality!

The other disadvantage of software synths is even more interesting, partly because it’s actually a difference between software synths and all types of hardware synths (not only digital), in other words, in-the-box vs outboard gear. It’s a less obvious yet HUGE difference that only a few uesrs think about:

The consequences of a different interface

Based on my experience as a professional electronic music composer and synthesist, and as an owner of dozens of hardware synths and a long list of softsynths, I can assure you that the differences in the way we control software and hardware synths differentiate them more from one another, than their sound or features ever could. While hardware synths might feature (ideally) a large number of switches, buttons, faders, sliders, dials, encoders, joysticks, software synth applications and plug-ins typically offer a streamlined graphical interface, showing the image of buttons, sliders, selectable parameter windows, drop-down menus and value fields.

Our physical connection with an electronic instrument plays an important role in musical sound design: being “one” with a familiar model often yields more diverse and interesting results. While “mousing around” on the screen and trying to access certain parameters in menus and sub-menus of a software synth can make the sound sculpting process way too streamlined, unintuitive and even frustrating, the literally hands-on operation of the hardware synths makes the user able to hear or imagine a sound or tonal character first, at the same time reach for a dedicated knob and take the sound to the just imagined direction instantaneously. No ideas lost or “textures in your mind” gone before your ears get to hear the changes in the sound – the mind is ready to take the instantaneously heard sound to a new direction in real time by directing the hands. This sensation of literally touching a parameter, this continuous two-way feedback between the two hands and the ear/mind ensures a highly creative and idea-inducing sound shaping process, which is largely or completely missing when one is using an indirect controller such as a mouse, is entering values, scrolling through menus. Even with a controller keyboard, you are limited to one particular interface, the controls of which were not designed with your particular soft synth in mind. We use the mouse for way too many things in our lives today, why make the way we control all of our instruments so uniform, too?

Although some awkward menus of small-display digital hardware synths might slow you down in the sound design process, they will still offer a direct hands-on experience, and a menu system featuring a different structure and logic for every instrument – less likely that you will follow the usual left-to-right structure of soft synths, and less chance that you will end up at the same place when you’re trying to create a brand new sonic texture or behavior.

The significance of the difference between what tools (like a mouse vs physical buttons) and what senses (such as seeing vs touching) we use to communicate with an instrument is greatly underestimated by most, yet they might allow users to unlock the real creative and unique potentials of their hardware instruments, and more importantly, expand and realize the user’s sonic ideas.

Don’t get me wrong – I like many software synthesizers and actually use several in Studio CS. So where do I see their advantage? Some of them will feature unique parameters or functions not found anywhere else, and using these in conjunction with the more traditional ones can lead you to new sonic territories. Unfortunately, these instruments are quiet rare, I could name only a handful that give you truly new and actually useful options (and they aren’t the really popular ones). If you want a totally sterile, noise- and unintentionally distortion-free sound, they are a great choice, especially when you’re keeping your entire production inside of your DAW. Then, there are the obvious practical benefits, such as saving physical space (they take up none), and saving money (they usually cost fraction of the hardware instruments, or are even free) for the user. They usually have a faster learning curve, they offer a simple-to-understand, large graphical interface, many preset sounds and the ease of saving user sounds with a single click. They won’t increase your electric bill, you can’t drop them by accident, you won’t ever need replacement parts, and they don’t even need cleaning – softsynths are definitely the most convenient option when it comes to synthesizers.

The question is, however: is it convenience that you want, when it comes to your sound?

Read the 2019/2020 followup articles:

Part 2: Analog vs. Digital Synthesizers – Part 2: Soft Synths, Modular Synths, Emulations, Noise Machines

Part 3: Analog vs. Digital Synthesizers – Part 3: The New Wave of Wavetables