What's new

Real Vs Sampled Piano, is the difference worth the effort?

Instead of sampled, look into modeled piano plugins
They are even much less convincing to my ears, unfortunately. :)
(there were some discussions here, and i don't feel the urge to refresh them...)
The physics behind some things is much more complex than one might think.

Resonance buildups in a piano can have a very interesting (almost chaotic) behaviour.
That's one reason why it's extremely difficult to cut different takes of a piano recording:
You might play the phrase almost exactly in the same way, with the same pedalling, yet you have completely different overtone amplifications. The funny thing is, sometimes a cut might seem perfect but then you suddenly hear where the overtone content "jumps". :)
 
Only extremely auto-tuned and super processed vocals are "winners" and all the rest is for boomers.
Not at all. Vocals of course sit on a side of the aforementioned spectrum furtherest from drums or pianos. You could have extrapolated that from the previous post. 😉
 
They are even much less convincing to my ears, unfortunately. :)
(there were some discussions here, and i don't feel the urge to refresh them...)
The physics behind some things is much more complex than one might think.

Resonance buildups in a piano can have a very interesting (almost chaotic) behaviour.
That's one reason why it's extremely difficult to cut different takes of a piano recording:
You might play the phrase almost exactly in the same way, with the same pedalling, yet you have completely different overtone amplifications. The funny thing is, sometimes a cut might seem perfect but then you suddenly hear where the overtone content "jumps". :)
To be fair, I think resonance effects depend on piano lid position (open, quarter open, closed, completely closed) and mic position as well. There are recordings with inside piano perspective (really wet) and there are those with tiny studio ambience (really dry). And piano not in solo context changes the game altogether.
 
To be fair, I think resonance effects depends on piano lid position (open, quarter open, closed, completely closed) and mic position as well. There are recordings with inside piano perspective (really wet) and there are those with tiny studio ambience (really dry). And piano not in solo context changes the game altogether.

Yes i know, however i was referring to situations where the same positions of the pedal are reproduced...
As said, these things are complex...
 
. . .
I guess to get really more convincing emulations of the resonances and also the spectral behaviour of fast repeated notes there is much more physical modelling necessary than it's actually possible.
The differences are not even subtle. They are super-obvious.
Recreating the aural phenomenon of a hammer/s repeatedly hitting already moving strings is a tough one for sampled pianos. Undoubtedly it's not the same as triggering samples which were captured with the hammer hitting the strings in a resting state, regardless of how many velocity layers, round robins or sympathetic sample overlays there are.

The old East West Quantum Leap pianos were sampled pedal up, pedal down and with multiple repetition samples in an attempt to capture this sound. It works to some degree, but there are issues & problems with these sampled pianos which have been talked about previously. Still, kudos to East West for addressing this aspect of piano performance.
 
😂😂 this fuckin' place... lol. Dude starts barking around about repeated notes, resonances, and playing Liszt's B minor sonata... in response an excerpt from freakin' La Campanella is learned and recorded in half an hour and it's seen as sterile with a need for a more discerning ear.

That's peak VI-C. Nice work! 🤣🤣🤣🤣 Maybe if I used more pedal next time!
 
I'm less concerned with the "difference" and more with "fit for purpose". For inspiration and encouraging a minimalist approach, the nuances of some sampled pianos are priceless (and I do find the use of the sustain pedal gets intense focus). Just as much as a real full concert Steinway gives you goosebumps pounding out some of the classics in a way that no other alternative can. Unfortunatelty modelled pianos do neither for me as yet.
 
To me, rapidly repeated notes, and the resulting amplification of certain resonances, aren’t even anywhere near the biggest obstacle standing in the way of a sampled instrument attempting to pose as a real one. That’s more of a “Ah, yes, that too, of course” sort of problem, I find.

It’s the near-complete absence of living colour, living timbral detail, delicacy, dynamic sculpability and overal dimensionality of the sound that bothers me most. All of this in immense contradistinction with a real piano and the way it deals with the chaotic energy that enters its complex being via the hands of the player, and then responds accordingly. That, and the way a real instrument’s sound takes to, and claims the room. Quite simply: unsampleable.

That absence of chaos and complexity (and these two being replaced with the tedious poverty and predictability of dead sound) are, I feel, the main reasons why virtual pianos — all of them — are but superficial, flat and extremely tiresome substitutes for the real thing.

I can’t stand 5 minutes listening to, let alone playing, a Synchron piano (or any other virtual piano for that matter). It bores me immensely. Playing a virtual instrument, you’ve explored its territory completely within half an hour, you know its idiosyncracies, limitations and its strengths (if there are any) down to the tiniest details, and once all that is established — like I said, usually within half an hour, often much sooner — the instrument will never surprise you again. It can’t, as it’s dead.

A few posts ago it was suggested that the Synchrons are ready to tackle the entire classical repertoire with totally convincing results. How anyone, let alone an accomplished player, can say such a thing is more than I can understand, but I guess we all listen very differently to music, to the way it is performed and to the sounds that carry it into our brains. (And that bizarre, almost religious fanaticism of VSL adepts, the way they are convinced of the infallibility of just about anything that VSL releases, and turn a deaf and acrimonious ear to everything that contradicts it, plays its part in this matter as well, I have a feeling.)

Anyway, below are 10 examples of real piano recordings — and I could easily post days of non-stop music like this — which, in my opinion, no virtual instrument is entirely ready for. I focused with these examples more on the delicate side of the piano, to my ears the side that will instantly reveal how inept virtual instruments still are.
(Synchron pianos, at least the two I own, can be quite satisfactory when rendering the big, bravoura side of a piano. My Roland V-Piano also sounded best everytime I got “medieval on its ass”. A piece like, say, the “Russian Dance” from “Petrushka” is something I imagine a Synchron piano could do quite well.)

A virtual instrument compares to a real one the way I, as a piano player, compare to Martha Argerich. I can play a scale as well as she can. But that's no proof that I’m in Argerich’s league, is it? And it’s the same thing with virtual instruments: a good sampled piano is capable of suggesting a tiny portion of all that a real instrument can be, and do so quite convincingly, sure, but that’s where it ends. Beyond that ‘tiny portion’, there is still an infinite vista of musical life, timbral expression and ultra-precise, context-aware definition and articulation, all of which is the exclusive province of a real instrument and thus totally inaccessible to a virtual instrument.

https://users.telenet.be/re-peat/Pianos/Ravel_Prelude.mp3 (<b>Ravel: Prélude</b>) (Georges Pludermacher)
https://users.telenet.be/re-peat/Pianos/Stravinsky_LHistoire_Fragment.mp3 (<b>Stravinsky: L&#039;Histoire du Soldat</b>) (fragment) (Christopher O'Riley)
https://users.telenet.be/re-peat/Pianos/Dett_Song.mp3 (<b>R. Nathaniel Dett: Song</b>) (Clipper Erickson)
https://users.telenet.be/re-peat/Pianos/Prokofiev_Cinderella.m4a (<b>Prokofiev/Pletnev: Cinderella</b>) (Martha Argerich & Mikhail Pletnev)
https://users.telenet.be/re-peat/Pianos/Debussy_LaDanseDePuck.mp3 (<b>Debussy: La Danse de Puck</b>) (Jean-Efflam Bavouzet)
https://users.telenet.be/re-peat/Pianos/Mendelssohn_LiederOhneWorte_AndanteUnPocoAgitato.mp3 (<b>Mendelssohn: Lieder Ohne Worte, Andante Un Poco Agitato</b>) (Javier Perianes)
https://users.telenet.be/re-peat/Pianos/Beethoven_PianoSonata14Csharpminor_op27_Allegretto.m4a (<b>Beethoven: Piano Sonata Op. 27 nr. 4, Allegretto</b>) (Maria Kodama)
https://users.telenet.be/re-peat/Pianos/Liszt_GrandesEtudesPaganini_Nr4.mp3 (<b>Liszt : Grandes Etudes de Paganini Nr.4 in E major</b>) (Daniil Trifonov)
https://users.telenet.be/re-peat/Pianos/JSBach_CapriccioBVW992_Andante.mp3 (<b>J.S. Bach: Capriccio BWV992, Andante</b>) (Claire-Marie Le Guay)
https://users.telenet.be/re-peat/Pianos/Haydn_PianoSonataEflatMajor_Adagio.mp3 (<b>J. Haydn: Piano Sonata in E-flat major Hob. XVI:52, Adagio</b>) (Rafaël Blechacz)

And then we haven't even discussed jazz piano yet, a field where virtual pianos sound at least as embarrassingly flawed as they do in most classical music.

_
 
Last edited:
Peat, do you remember this exchange?

https://vi-control.net/community/th...is-your-favourite-and-why.105578/post-4761887

My suggestion would be to practice more with your Synchron pianos… and not just for the feel of the instrument, but practice sculpting the sound and with the various key-by-key parameters and of course, mic positions.

Sampled pianos require different adaptations for each specific individual instrument, even if made by the same developer. Spending more than 30mins before dropping them into a template is advisable!

I have done this with the VSL Steinway, but not the 280VC yet. Going key by key and utilizing the per-key EQ and volume, plus automating the room mics, body, key noise, pedal noise, or sympathetic resonance, then in the mixing phase upping the polyphony to 1,024 notes will give you want you want.
 
I was going to bring up Mozart/Beethoven and the fortepiano, but I'm glad someone mentioned Bach! At the risk of venturing into the philosophical, perhaps it's worth exploring the nature of music: to me, music seems to have dual nature, music = design, but music = sound as well. The discussion hitherto focuses on the sound nature of music, but in the context of classical music such matter is beyond the scope of the design. In other words, the instrument for which the composer wrote was vastly different from today's concert grand that all this discussion about timbre, resonance, dimension, nuances etc. is a reflection of your personal preference and not necessarily the composer's intention. I feel it's important to make this distinction.

Another way to look at it: there is a school of thought that 'authentic' music should be played on instruments contemporary to that period, e.g. Bach on the harpsichord, Mozart/Beethoven on the pianoforte, Chopin on Erard piano. So if you play Bach on a Steinway it would be disqualifying anyway, making any discussion about VI moot.
 
Last edited:
I'm a multi-instrumentalist (primarily a pianist). While I often think of all these virtual instruments and libraries as a "necessary evil" and more convenient and affordable alternatives to a real thing, the practical benefit of working with these is evident, despite all the shortcomings.

From my experience, when dealing with any kind of acoustic instrument, particularly in your own environment, something is always lacking. While you can buy audio interfaces and microphones that can give you "professional quality" even in the budget range (at least, in theory), only a few of us are fortunate enough to also have a perfectly balanced recording space and conditions, or a meticulously maintained piano. Of course, capturing all the imperfections of a certain instrument might just be what you're after.

Speaking of character, thanks to the vast library of virtual pianos out there, you can more easily find the one that best fits your particular needs, as each brand, type, age, etc is different. Sometimes you don't even need a library that is "technically perfect". Numerous examples at Pianobook are living proof that you can capture the sonic essence of a particular piano with less than stellar conditions and equipment.
 
Look, I'm a massive Synchron fanboy and wouldn't even be here if I didn't love them as they are. But it's good to be realistic about what may be possible and what has been shown to be possible. It may be that experienced engineering talent could find the glow and resonance of that Olafsson within a Synchron piano. But until someone demonstrates this it doesn't make sense to assume it's possible. Upping the complexity of modelling (with Synchron) seems a more likely pathway towards that particular place.

Absolutely nothing but respect for your 2 minute mock-up. At the same time if anyone finds that to be in the same aural universe that's a testament to what re-peat suggested about how we all listen in our own individual way.
 
these arguments about sonic purity are so boring because they're just philosophical masturbation about nothing important. Everyone knows a VI is not a real instrument equivalent. We all understand there are limitations and trade-offs. These arguments always end up at the same conclusion. most people who listen, don't know the difference and don't care....and the very hypocritical argument for this VI-focused group persists. VI's are nothing like the real thing....but most of us use VI's for our work and still create emotionally moving pieces of work despite the limitations of the technology. So it "feels" different, who cares? There are plenty of ways to adjust VI's to draw more out of them, but if we're talking about accessibility, and practicality in today's markets, VI's are a good thing.

Hell, I could have recorded with a real piano for my music. I had the budget for it, but I wanted flexibility. If I was to do a piano solo, I would likely not use VI's because, yeah, a piano feels different when it's by itself, exposed. Still, that entirely depends on the music, and its needs. Not everything needs to be such a pompous exercise in snobbish musical standards. If your argument to others is "get your ears checked! it feels sterile", the only relevant rebuttal is, most people don't give two shits about how good you think your ears are. They like what they like, and a VI is a suitable substitute for most people.
 
I think we can all agree on two complementary propositions:

1 Sampled pianos do not sound exactly like very well-recorded expensive pianos: the difference can be more or less obvious to us depending on repertoire.

But and also:

2 99.999% of people who will ever hear our music are not the kind of cork-sniffers who would be able to tell the difference or care about it if they could. They will react emotionally to the musical content, and it's moot whether it was recorded on a real piano or a good VI (assuming a certain baseline of good taste and technical ability).
 
I can't agree with your second point @SupremeFist but only because I do agree with your first, that "the difference can be more or less obvious to us depending on repertoire."

If the score leans heavily on the piano as a solo instrument, with exposed arpeggios or "Liszt-like" passages, I don't think samples will sell it adequately. There are plenty of arpeggios and exposure, for example, in Marinelli's score to a (fairly) recent "Pride and Prejudice," for example.

Even a layman can hear it. Often people who are not musicians will tell me my mockups are "just fine" -- until they hear a real orchestra play it. Then I think anyone can tell the difference, really (unless it's all buried in percussion or sound design or synths).
 
I can't agree with your second point @SupremeFist but only because I do agree with your first, that "the difference can be more or less obvious to us depending on repertoire."

If the score leans heavily on the piano as a solo instrument, with exposed arpeggios or "Liszt-like" passages, I don't think samples will sell it adequately. There are plenty of arpeggios and exposure, for example, in Marinelli's score to a (fairly) recent "Pride and Prejudice," for example.

Even a layman can hear it. Often people who are not musicians will tell me my mockups are "just fine" -- until they hear a real orchestra play it. Then I think anyone can tell the difference, really (unless it's all buried in percussion or sound design or synths).
And yet a lot of people also loved the piano-led score to The Queen's Gambit, which iirc was Noire layered with Alicia's Keys (!). I think sampled pianos in general are more convincing to the general public than whole orchestras as yet.
 
Top Bottom