The elephant in the... analogue vs. digital synth debates

There have been decades of debates (and even pontifications) about analogue synthesizers being "better" than digital ones. In the rational cases the comparisons were made within the confines of some specific context of behavioural aspects or particular sonic identity for particular sub-genres of (electronic or other) music. However, in most cases cases a central aspect was ignored or side-stepped. 

After decades of sound synthesis evolution with myriad different methods opening many different creative avenues and even creating new workflows, it is remarkable that one still has to point at the tacit, ginormous, and blindingly fluorescent elephant in the room.

Let's first quickly jolt down what is not worth digging into when one wants to introduce this elephant, because the fallacy is far more fundamental than the following technical details. So no, this will not dig into details of:

• The evolution of computational power, thus mathematical abilities, too, which rendered decades-old arguments about accuracy of signal representation, oscillator behaviours, characteristics of digital filters, and effect chains factually out of date and/or blatantly unaware of current realities. The fundamental problem of these debates is not in the area of who can hear what analogue imperfection, warmth, charming instability, smoothness etc. in true analogue gear when compared to analogue modeling synths. 
• Coupled with the above, the evolution of physical modeling (of circuits and electroacoustic phenomena) arrived at a point where one could specify in even the age of a circuit - and produce remarkable emulations of the original analogue version. Again, any measurable or perceivable differences are again unrelated to the central problem. 
• Subjective preferences in work flows and nostalgia (e.g. about the ephemerality that comes from those work flows). Again, these have nothing to do with the particular, very large, elephant in the room even when there are valid rational aspects (e.g. pros and cons of instabilities, the need or lack of need for instant recall etc.)

Now then... about that elephant... 

Well, to make its presence disturbingly obvious, let's temporarily take a seemingly bizarre step, even leap, far outside the world of sound synthesis. Let us venture a bit into the history of painting - and, within that, the history of pigments and painting techniques. 

Although the use of oil paints dates back to ancient times, the phenomenal revolution that changed everything came only in the 15th century - not only because oils replaced eggs as a binding agent, but techniques perfected by Jan van Eyck and Leonardo da Vinci for the use of these oil-based paints enabled the creation of unprecedented details, smooth transitions, subtle layering of pigments, optical effects, hence realism. Certain materials allowed artists to invent and use new techniques that have led to entirely new, extended, artistic palette (pun intended).

Leonardo's use of the sfumato technique (just think of the Mona Lisa) was enabled by this type of paint that could be applied in thin layers to produce unprecedented subtle transitions. Jan Van Eyck's somewhat creepy Arnolfini Portrait is another fabulous example of his perfected oil painting technique - so astonishing, that later on (the anyway only partially trustworthy) Vasari even attributed, wrongly, the invention of oil painting to him - and this was parroted by many for way too long... even by Google art history pages at present time...

In addition, the expansion of trade routes gave painters access to materials that had been either entirely new or just horrendously expensive until then. A red dye extracted from an exotic insect gave painters a vibrant crimson never seen before. Chemists, from the 18th century onward, created new pigments or discovered synthetic, hence affordable, versions of extremely pricey ones - let's just think of ultramarine, which in the Renaissance era could literally bankrupt a painter. 

With the invention of myriad synthetic pigments, industrial paints, and the new painting techniques that emerged with the use of these, artists created never before imagined nor seen tones and textures. To mention just one milestone: acrylic paints suddenly gave painters water-based, quick-drying, and cheap paints that had vibrant colours. Painters could achieve watercolour effects whilst exploiting how the canvas absorbed the water-based pigments - and so on and on.

Rothko is just one example of artists who used all kinds of pigments, materials, and techniques, from watercolour to acrylic dyes to oil to ink. After all, instead of ideologising and/or eulogising one type of material or technique, some embraced whatever was available - because those pigments and techniques were tools of expression. 

They did not define themselves by the tools they used, somethint that would have totally turn upside down the relationship between the artist and the instrument(s)... Does the latter sound familiar from somewhere?...

So let's imagine a painter in the 21st century stating that using primed panels and egg tempera, the type of paint that used egg yolk as binding agent (think of the Renaissance greats), is "better" than all of the countless dyes and techniques, hence ranges of new colours, optical effects, transitions etc. that got added to the world of painting from the 15th century onward.

Such painters can only avoid sounding ludicrous or nonsensical if they frame their claim within the context of a personal methodology, preference, and/or artistic aim that needs those particular pigments and techniques whilst disregarding myriad others that deliver their own unique artistic effects.

Some do justify their self-limiting choice within such well-defined context, with the choice and the results being highly respected - think of James Lynch's landscapes where he returns to the toolbox of Medieval masters because he is after specific effects. 

The point is not that Lynch adopts old pigments and techniques. The hopefully very obvious point is that some artists adopt a certain well-defined subset (!) of currently available means of artistic expression in the service of particular and specific end goals.

OK, so... back to synths then - and that elephant in the analogue vs. digital synth debates. 

If not terribly obvious by now, the central aspect that renders such these debates vacuous is what analogue sound synthesis occupies in the galaxy of sound synthesis methods. 

Most analogue synthesizers are based on so-called subtractive synthesis, Other analogue synthesis methods can add to pre-existing frequency-domain content - for example, ring modulation adds frequency components that make highly characteristic sounds. Frequency modulation, when done via analogue means, can also result in complex spectra - but this is lightyears behind complex multi-operator setups for numerous practical reasons.

For example, frequency modulation-based analogue synthesis cannot create digital multi-operator FM synthesis in the vein of John Chowning's revolutionary method, because the precise control of myriad parameters (and an inherent mathematical stability essential for the reproducibility of a certain sound) cannot be achieved via analogue oscillators and their analogue controls. Tiny instability or change in parameters can produce vastly different sounds, therefore these parameters need precise and stable means of control, not to mention the programmable evolution in real time of myriad parameters. 

Without going down the sound synthesis rabbit hole, the obvious fact is that analogue synths and all their vast, world-changing, and still highly fascinating sonic capabilities occupy a mere subset of sound synthesis methods. 

The list can go on and one. As just one extra example, any objective person with any command of sound synthesis, related physics, and signal processing knows that no analogue synthesis method can produce the timbres created by Wolfgang Palm's particular type of wavetable synthesis, nor those of granular synthesis... 

But again, the key point is not the capabilities, strengths, and limitations of certain synthesis methods and the instruments that employ them. 

The "A is better than B" debate, when it comes to analogue "vs." digital synthesizers, only holds any water whatsoever when it is framed, exactly as in the case of painting, by personal particulars. 

The problem with a nebulous "better" is the irrational generalisation of a narrow and  personal frame of context, i.e. the elevation of one specific subset of sonic capabilities above the entire range of capabilities. 

If someone is only and only after what subtractive analogue synthesis can produce in terms of sonic palette then one can eulogize and even ideologize it - as long as the confines of that sonic world are honestly acknowledged. 

If one is only using subtractive synthesis in its pure analogue form for musical genres, styles, 'moods' and sonic colours that are characterised or even defined by that synthesis method, then fine. Be rational about it and see that a subset of colours are being used from a vast palette. Even if that subset is fiercely said to be inimitable and unique (this is a whole other debate on analogue modeling accuracy), it is still a subset. 

The arguments centred on a resounding "synthesis method X is better" are as fallacious as an "egg tempera yellow is better than all other vast ranges of pigments and painting techniques" statement would be in the sphere of visual arts.

Irrational, even deplorably desperate attempts to square the circle did and do exist. Some had set out to demonstrate that everything that came after analogue (subtractive) synthesis can be created via the latter. It is deplorable not only due to the immense waste of creative and engineering efforts that could have been used for using the right tools for the right job (and focusing on the artistic goals, if any) - but also because it shows the cognitive dissonance at work. Attempts to make rudimentary and impractical (this, unusable) analogue 'replicas' of some digital synthesis methods reveal the psychology of obstinate denial of fundamental physics, acoustics, and sensory perception laws.

For example, the spectral movements caused by Palm's revolutionary wavetable synthesis (not to be confused with the generic uses of the term 'wavetable') cannot be achieved via analogue means - due to its core, defining, principle. Internet examples of euphoric analogue fetishists demonstrating the morphing from one simple waveform into another one via immense effort and purely analogue means are as commendable as they are tragicomic. 

Those attempts demonstrate not only the fundamental limitation that is hit immediately by them - but also the lack of understanding of what a PPG Wave synthesizer and all its famed successors did to create their unique sound world.

Similarly, attempts to demonstrate multiple analogue oscillators frequency modulating each other demonstrate the inherent absurdity of the task: beyond a two-oscillator setup, the tiny differences in the oscillator circuits (from instance to instance), instabilities in their parameters, and the inability of controlling precisely the notoriously vast number of often rapidly evolving parameters make the sounds impossible to control and reproduce.

If one takes physical modeling synthesis, for example the Karplus Strong algorithm of a string model, and dials in precisely controlled (even physically impossible!) parameters that cannot occur in the real world... or excites the string with complex audio waveforms instead of e.g. plucking... it is simply pointless to even discuss the analogue emulation of this. 

One could go on and on. 

In both worlds, painting and sound synthesis-employing music, analogue sound synthesis methods are a subset of methods that have produce their eminently different sonic worlds - a simple fact, even if in our post-factual society facts have been reduced to mere subjective opinions. 

Please, whoever resorts to the sweepingly general "better" rooted in a false dichotomy: add an acceptable reasoning that is rooted in your personal artistic aim - and acknowledge the boundaries of the island on which you chose to move to. 

That island has unique beauty and buried treasures to uncover over time - but there are vast, very different continents out there. Ignoring them is a personal and artistic choice, so is traveling around the world and embracing all the distinct marvels of all other islands and continents. Some choose and even need the latter for their sonic ventures, like all artists throughout history who embraced all creative means and put them in the service of their art - instead of defining themselves via a tiny subset of them.

Misynformation: synth reviews in the post-truth world

The term post-truth (or post-fact) was coined by Steve Tesich in the early 1990s in his essay The Nation. It has been widely used in political sphere, denoting a situation where not the factual details matter. Instead, one appeals to emotions, and there is, as a key characteristic, rejection of factual comebacks.

One would not expect to see the same shift from facts to a post-factual discourse in something as technological and inherently objective as the world of engineering specifications for electronics and within that, electronic instruments.

However, a recent trend is that not only manufacturers, but also reputable synth technology-related media have shifted to a surprisingly post-factual style.

Marketing was always about overstatements, let's face it. Whether that comes from the manufacturer or secondary forums that have a vested interest in promoting / selling the products, is not a key distinction.  They are information sources for prospective buyers / users, therefore they all have quite a responsibility.

However, one could see the intentional eroding of even fundamental instrument categories, mechanically repeated by thought-to-be technically informed and reputable media. One could see sensationalist headlines along the lines of "is this the future of...?" when they reviewed products with features running more than a decade behind the times.

It seems in synth world, too, the increasing  hype and sometimes desperate overcompensation from some manufacturers, in an attempt to hide the lack of actual ideas, is being parroted by even specialist media without even a quick comparative look at specifications.

Countless social media debates then promptly denaturate into typical post-factual discourse: if some come with an objective point based on glaring technical facts in the product specification or its real-life use, such comments are attracting the "fake news" type instant dismissal.

It is rather interesting to see this phenomenon in the sphere of electronic music technology.

Examples abound... but a note in advance: many terms here are considered to be known based on long-standing, even historic, definitions. To use analogy based on the Blackadder comedy series of yesteryear, Baldrick's hilarious "cat = not a dog" definition does not detail what a dog is, if latter is well-defined and well-knownm and 2 seconds in Google can provide it :)... But back to a (slightly) more serious look of sometimes hilarious synth reviews...

Before going to glaring and monstrous examples of media mis-stating fundamental aspects of new products, a more subtle (and in need of erudition) case is that of the recently released and exciting Mostro FM synth.

Even SynthAnatomy chose a headline that is shockingly unaware of signal processing fundamentals and of actual synthesizer history, seeing more than 2 FM operators as "DX7 backstory" that the reviewed product has none of.

Why can any DSP person say this is shocking lack of awareness of signal processing fundamentals? Well, the "DX7 backstory" (or lack of, in this case) is not actually a DX7 backstory.

It is fundamental mathematical reason stemming from Dr. Chowning's revolutionary paper on FM synthesis. DX7 did not chose to have more than 2 operators on a whim, but because it is necessary for sufficiently complex audio spectra.

Not having more than 2 operators in an FM synth is not a plus... for very fundamental reasons. Again, the more than 2 operator FM synthesis is not a "DX7 back story", it is fundamental need for complex spectral changes in FM synthesis.

One can ask, how can such reputable synth media make such headlines and completely side-step not only mathematical but also synth history facts?

Regardless of subjective preferences for brands, a notable misinformation case was that of Yamaha Montage. It was accurately marketed as a synthesizer, not as a workstation, by Yamaha. However, countless very reputable and usually serious/informed media wrote about it as "flagship workstation"  or "best workstation synth", to quote just a few examples.

This got even more tragicomical when its cut-down repackaged version, the MODX, was written about even by SynthAnatomy as "workstation".

It was and remains factually and fundamentally incorrect to categorise it as a workstation, kudos for even Yamaha accurately stating the correct category for these instruments.

One of the very few serious synth and studio technology reviewers that emphasised from the start the key difference was, as usual, Sound On Sound, who have re-stated several times the distinction.

The post-factual furore was at full swing in social media, from YouTube to Facebook groups and so on. Correcting this huge misnomer attracted endless subjective furore, exactly as certain factually wrong or self-contradicting political tweets or articles do...

This particular example is merely about overcompensation for long-established pre-existing features lacking in these products, whilst those features are actually central to the instrument category definitions themselves...

However, similar eloquent cases can be found when manufacturer over-uses the word "new" - and even reputable media mechanically repeats this, proving the fact that even simple comparative look at specifications, informed by instrument and technology history, has not been done.

Such case is the very recent launch of Roland's Fantom workstation.

The fact, that the manufacturer hyped the product, is understandable and forgivable to an extent. However, in the era of post-factual media, it is more important to note how automatically the factually incorrect claims have been repeated by even serious synth review sites and retailers.

Andertons are asking: is this the future of workstations?...  A simple look at the specs, even before getting hands on with the new product, would have eminently told one that this is the past of workstations, if one considers e.g. synth engines and sampler parts. Why not be honest about the novel UI and the performance capabilities, instead of putting "new" where there is none - and even old feature is more limited.

Others even labeled it "the ultimate workstation", again forgetting the simple fact that it has series of missing features and considerably more limited or lower-performance features compared to long before existed workstations out there. Others introduce it as "all new" in their first sentence, whilst is has many merely repackaged and long before existed elements and subsystems.

Not only it repackages long-existed synth engines (e.g. from XV family of synths), but it actually lacks key features that existed for long time in other synth workstations with several times higher specifications.

For example, the sampling ability is shockingly limited both in functionality and capability, if one does a very quick comparison with e.g. the long-existed Kronos. Even OASYS had several times more synth engines integrated with a single user interface front-end.

The "seamless transitions" do not exist for Fantom programs and effects, even the review video has clear and not seamless transitions... They only and only work for the so-called scenes, whilst other long-existed workstations can make seamless transitions between programs, combinations of programs, with entire effect chain transitions.

The problem is not that manufacturer, repackaging pre-existing elements into a new product, the real problem is the overstating the "novelty" element.

The bigger problem is how end users can be subjected to many synth reviews and demos that, without basic specification checks, repeat falsehoods or hype the product without realising that long before existed products had the same, and better performing features.

In the sphere of social psychology, it would be quite an interesting topic to dig deeper into this trend, where post-factual rhetoric is permeating even formerly technical facts-based discourse.

Within the world of synths, as end users or technology aficionados, one has to wonder how we can actually end up here.

There seems to be a strong correlation between endlessly repackaged pre-existing technology and the hype overstating novelty, even when it is glaringly missing in specifications - let alone in terms of factual synthesizer history. 

This may exacerbate over time, and social media with typical post-factual treatment of objective comments is making it increasingly easier to drown out factual discourse.

As the late Umberto Eco postulated, in the hyper-real world fakes can seem more real than the real thing. And that seems to go for synth review claims, too.