Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Digital pianos can never emulate an acoustic - the reason

You may be right, but that's not the same point I'm trying to argue.

Can you reliably play volume 46 and not 47 when you want to?
Can a listener discriminate between 46 and 47, if played A-B?
Is there a difference between the digital and the acoustic in this respect?

Those are all valid questions of fact.

But I am more interested in a different question.

When you hear a difference (hear, not claim to hear), is it in fact there?

Most people have no idea to what extent their brain can produce differences that do not exist.

Very nearly all musicians can hear the difference in sound between a wood flute and a silver one, e.g. This is true in a noisy rehearsal or performance hall where you'd expect it to be difficult, in small quieter settings, etc.; any time they know which is which they can always tell.

As you try to make it a tighter experiment, this gets harder. Make it single blind, make it double blind, make the instruments more carefully identical. With instruments made as carefully identically as possible, the standard test is to play three times, once with one instrument and twice with the other. The task is simple, identify which one is different. The best flute players do no better than chance. But if you tell them which is which, they will "hear" the difference and describe the differences they hear in the usual subjective terms we have for sound. We have yet to find a flute player who can be convinced by these experiments, because their experience is real - they really do hear something.

the point I'm trying to make is that the ability of our brain to fool us is very real, and far stronger than most of us can accept.

There is no way for you to play your Bluthner and not know what you're playing. The touch, the sound, the smell, the reflections off the surface - just can't do a double blind here. You expect 256 levels on it, and you know you only get 128 on your digital, and that's what you're going to here. But it might not really be there.

Interesting points, certainly. However, according to some scientific studies the impact of the key against the key bed plays a genuine role in tone. That's without even including the pedal, to make these sounds reveberate. Digitals don't even put this into the mix- perhaps one of the reasons why those who are accustomed to ease of produing a pleasant tone on a digital sometimes struggle so much on an acoustic.

It's interesting to see how thumping a digital can still result in a certain roughness, overall. But it's highly notable how it doesn't make things anywhere near as ugly as on an acoustic piano. Yet more factors that are readily audible, but which have yet to be simulated.

I like you a lot Ervin...I really do...I think you are a marvelous human being, in spite of your arrogance and lack of knowledge, and who else but the great Ervin to defend the weak and the downtrodden...it's fitting that you have championed the seriously over-rated acoustic piano.

Oh wait...I'm a hypocrite....mmmm...so, did I mean what I just said?

Of course I did...you're a such an amusing little fellow, in spite of your shortcomings.

You and I should have a beer together someday.

Happy trials,

Snazzy

+1 Great post, Tim!

Snazzy

That's a key word, emulate. One of the instruments is being emulated, the other is attempting to emulate.

There are infinite variables with the real thing. That takes the lid off the person playing it. The sky is the limit.

If I could hold down the sustain pedal on a digital piano, and create sound the way I can by lifting the dampers on a great piano, I would buy it today.

Mychal

Important points.

People tend to hear what they want to hear, what they are "supposed" to hear and what they expect to hear.



Well, if you are going to have a beer together, ask the bartender to provide you with several blindly poured glasses of beer into opaque glasses to see if you can identify which beer is what. Even people in the beer business find this a difficult or impossible task.

The question of whether acoustic piano sound can be altered in any way after the hammer returns to the rest position has been kicking around for years. The likely answer is that the preconceived notion in the brain of caressing a note instead of thumping it results in an altered key downstroke even though the player might think the different sound is the result of squeezing the key as it rests on the keybed. Nothing scientific in that statement ; just my own observation standing over grands looking at the action movement while I depress keys with different intents.

In the long run the perspective of the player is more important than the perspective of the listener since the player is usually spending the dough and is best served by an instrument that he is assured will deliver his musical message. The assurance is worth a lot, even if some of it comes from an unscientific basis. The execution depends on the assurance, and the musical product is the result of the execution.

On the thumping issue, IMO nasty is a better word than ugly. To me the clanking metallic sound response of most digitals to a powerful keystroke is an ugly giveaway, as is the character of the low bass at forte, the tonelessness of the uppermost treble notes, and what an industry friend of mine refers to as the inevitable artifacts and quick decay that creep into some notes during the sampling process. The nastiness capabilities of a good Steinway (used as an example...a few other makers have it too) can be accessed anywhere on the keyboard and are useful in a a lot of classical and jazz lit to give some bite or edge to the interpretation. (Side note: Can your vintage Bluthner be nasty? ) I'm not bitching here because I use my digital more than my acoustic, but digital shortcomings are not fantasyland stuff. They are areas for potential improvement.

The sampling layers are a difficult question. If I sit at a digital with high-end headphones (which of course I can't do at all with an acoustic) and play a unison again and again with what I think are different keystrokes, I cannot hear the actual results of the claimed level of sampling. OTOH, I don't think that I can produce with intent anything close to the claimed levels, so it's a moot point. However, in executing a phrase in compositions of my own, I generally feel more expressive potential in my acoustic than I do in any digital. Is that which I feel really there? I'm not sure, but it does improve my level of satisfaction and thereby my level of play. Now, I'm nothing special as a player, and undoubtedly a real performance level pianist can provide on demand and intentionally execute more velocity levels in his keystrokes, but I seriously doubt that even at the highest level of players those levels on demand approach 128, 64, or even 32. What does seem clear though is that consecutive concerts by a top-notch player on the same instrument at the same venue on consecutive nights will be unique, and the uniqueness will come not only from the physical and emotional state of the artist, the tuning refresh of the instrument, and the particular karma of the audience, but from the infinite variables contained within the acoustic instrument to respond to all those variables.

Nyriegyhazi,

Ignore the references to 'son', 'amusing little fellow', 'needing medical help', and 'one-trick-pony-rider'. These are stock references used only to rally the troops into a polarized tug of war. Just sail on acknowledging that there's good and bad in both types of instruments. Some of the discussion you have provoked here from other members is really good stuff, and goes well beyond the usual in these either/or threads.

Well, if you are going to have a beer together, ask the bartender to provide you with several blindly poured glasses of beer into opaque glasses to see if you can identify which beer is what. Even people in the beer business find this a difficult or impossible task.

I'm sure they do, but what about a synthetically produced artificial beer, compared to a real beer? I'd fancy my chances. A beer would certainly be fun though.

Side note: Can your vintage Bluthner be nasty? ) I'm not bitching here because I use my digital more than my acoustic, but digital shortcomings are not fantasyland stuff. They are areas for potential improvement.

I know exactly what you mean. In fact this is something I value greatly in it. If you know how to play it, it can have a highly beautiful tone. On the other hand, it can also sound hideous and the students I teach at home oftne tend to hit it too hard or produce a thin sound. I find it highly useful in this respect, as what I learn from it carries greatly across to grand pianos.

im with you, and I'm on snazzy's side of this whole argument, but the problem with double blinds and these sorts of experiments, is that in art unlike in science or marketing, there is considerable variation in sensibilities, and it only takes a single example for your flute player theory to be disproved. Normally in science if we get a single outlier like this we just ignore and call it statistical noise. But supposed in this whole wide world of 7 billion people there is a single flute player who can hear the difference?

That's the problem right there. Because if you take a look at the wave forms, the instruments can detect a difference and theoretically, someone might be able to detect a difference.

So suppose, we take flute player after flute player, all the best flute players in the world and test them one after another, and none of them can hear the difference, that still doesn't prove it, because maybe there is a young flute player who will be the next maestro.

Music is a funny business. The scientist in me wants to look at these double blind experiments and say "See, you can't hear the difference, it's been proven statistically." But in this sort of thing, you can't really rely on statistics because all it takes is a single human being with extraordinary talent to undermine the tests.

Now here is what I can say. Most people are fooling themselves. Most people think they are the one, like in some matrixy way that they are the single human being that can hear the difference. This is obviously not true for most people, but it's a harmless lie and for musicians in particular this belief often leads to wonderful performances because they think they can hear something they can't. So it does no good to try to disprove someone who insists they are the one that can hear the differences.

Certainly. We haven't proven that nobody can hear the difference - there is always the possibility there is somebody out there with ears an order of magnitude better than ours.

But we have proven a few other things.

We have proven that IF there is a difference, it is astoundingly small, so small it would take that one of a billion humans to detect it.

We have also proven that the 99% of humans who believe they can detect it are wrong.

We have also proven that those 99% (well, most of them) cannot be convinced by the data, because their subjective experience trumps data every time.

And finally, an observation: people who claim wood flutes are different from silver, or red brass trombones different from yellow brass (I know personally more in the latter category) do not believe these differences are small. To the contrary, they describe major life-altering differences.

Is it possible that Mr. N is getting less dynamic change out of his Bluthner than he really is, because he expects to?

Yep.

I'm not saying he definitely is, I'm saying it's possible he's being fooled.

Good points, but it goes further still. The following must also be stressed- supposing you took otbherwise identical cars with automatic transmission and one with manual. If you got a driver who was not used to manual gears, he may well perform better in the automatic, finding the gear changes a hindrance. However, a truly skilled driver would perform substantially better with control over the gears than if it was done for him.

The same applies in these tests. For the differences to be notable, the differences must be properly exploited in the test. I believe that you could take two tests done on identical instruments and have one in which few people can spot the difference and one in which even those who are regarded as having rather ordinary ears can spot the difference. For the test to be worthy, you ought to have a pianist of the calibre of Gilels or Nyiregyhazi playing the instruments. Your average middle of the road concert performer will not do. If the performer does not push the pianos to their absolute limits the test is indequate. Are we striving for digitals that copy some of what pianos can do in the mid-range, or ALL of what pianos can do?

I can say for definite that I could have spotted the fake sound on the film of Katsaris film of the Avant Grand. The characteristics of the false sound are so strikingly distinctive, that there is absolutely no question that I am hearing limitations because I want to. There's an absolutely unmistakable uniformity that is not natural. It's easily distinguished-especially as Katsaris is no mid-range player.

This is a vastly different issue to two acoustic flutes. I'd be skeptical about such differences myself. Regardless, I don't think you'd find many flautists who would be offended at the thought of playing a concert on one of the finest non-platinum instruments in the world. You'd find an awful lot of pianists (myself included) who would not be willing to perform on any electric that money could buy (although I'd gladly play a concert on some far cheaper acoustics). Once more, we're talking about rationally definable differences that have been very widely observed and are well supported by empirical facts, not speculative potential differences in minutiae.

The listening perspective is always interesting, but the player's perspective takes precedence. It is possible that Mr. N. senses more than is physically present in his acoustic. It is equally possible that Mr. Snz. senses something in his digital that is physically not there. It is possible that Mr. X's assorted key attack flourishes add nothing tone-wise to the simple strike velocity and that when the hammer reverts to rest position, it's all empty bravado and nothing more. However, Mr. X, Mr. Snz., and Mr. N.'s beliefs will give them all a comfort level with their instruments of choice that will enhance their actual performance on them.

And please, in terms of the listening perspective, don't ignore dynamic range. Its compression in a digital (along with the character of booming bass if present in the sample) are the easiest ways for an experienced listener to hear the difference.

By the way to go back to the original post ... you know the one where Scriabinghost is asking if a "digital piano correctly emulates an acoustic. I don't believe one ever will"

I was thinking a bit more about Ervin Nyiregyhazi, and I realized of course scriabinghost is right. No digital piano will ever correctly emulate an acoustic piano. I was reading about Nyiregyhazi and one aspect of his play was a super fortissimo, one of the strongest fortissimo among contemporary pianists. The strings would flap around inside the piano. He liked it. Now one reason Nyiregyhazi hated piano rolls is that they were incapable of replicating this particular strength of fortissimo. You have to set an upper limit somewhere for those piano rolls. Likewise you have to set an upper limit for MIDI.

But how about a virtual piano? You can build models that don't have an upper limit, but you need to test these limits in the ways that a pianist might use them for them to work correctly. I think there are several things you can do to an acoustic piano that will never be emulated in a virtual piano. For instance, get a bench, climb on top of the bench and then stand on your acoustic piano. Now jump in the air as high as you can, and when you land, bash with your feet as hard as possible on the piano keys, try to get 500 pounds or more of force on the keys. The strings will flap around in a nearly uncontrollable fashion, going quickly sharp and then coming back into tune as they settle down.

I promise you that this has never been modeled correctly and never will. To make a model perform correctly you have to record many times on an acoustic, and I promise you nobody is going to keep on doing this to a Steinway grand, just to capture a noise that nobody will ever use.

Actually I take that back. Someone might use it. But they will use it on an acoustic pinao, they won't do it on a virtual piano. I have seen some performance art with pianos before where they get crushed and lit on fire. Maybe someone already has a song where you bash the keys like this.

But the point is, there are things you can do to an acoustic piano that will never be emulated properly in a virtual piano because for one reason or another the customer just won't use this particular ability.

I know some jazz pianists like to stand half inside and half outside the piano, and while they play on the keys, they actually pluck the physical strings with their fingers and mix the two.

It's not that it's impossible to model these sorts of things, it's that virtual piano designers won't do it, the customers aren't demanding it, and even if they did model the two above things, there are more things that I haven't even thought about yet, things people haven't even tried yet with an acoustic piano, and you can't model something you haven't thought about yet.

Interesting point. On some of Nyiregyhazi's recordings you actually do really do hear the tuning distorting momentarily, through the sheer explosive force, before returning. I can't think of any other pianist that I've heard this effect so notably from- not even those who employ percussive attacks, instead of close pressure.

Also, the more I think about it, the more I do wonder if one key can be played with 128 notably different strikes (although there are certainly more than the handful of sounds that sampling offers). However, add the percussive force of a heavy attack (that can be varied independently to a good extent) and you add another variable. There is evidence that this can indeed affect what emanates from the soundboard.+ Start adding notes together and they interact with each other. The permutations keep growing and growing. Put down the pedal, and it just starts getting silly to attempt to rationalise every step. It's chaos theory, by this stage. I am now starting to wonder whether 128 different volumes are necessarily distinguishable in themself. However, it's the sheer number of combinations and permutations that are still falling woefully short.

PS. I wonder though, would jumping in the air necessarily provide a louder sound than the manner of acceleration Nyiregyhazi used? It's a complex issue. It may be that a level of acceleration far greater than that of gravity is required at just the right split second, in order to keep add the most energy to the hammer. I think this may be why Nyiregyhazi achieves far more volume than many who fall from a great height with their arms.

Add to that, played by intent, IOW on demand at the wish of the player and proven to be within close variances from performance to performance.



As silly as it might seem to some, breathing is a significant factor in both the controlled execution of the pianissimo and the power of the fortissimo. Control even within the fortissimo is important because mose pianists are not interested in percussive shock value that disrupts the harmonics. They aim for punch without unpleasant side effects resulting from pushing the piano beyond its useful range of expression.

by the way, on my computer I have access to many different kinds of software pianos, model 1923 steinway d, bechstein boesendorfers, steingraeber upright, 1908 clinton, yamaha 7cg.

And my acoustic piano is this HORRIBLE SPINET, oh my goodness it is so horrible, it can't stay in tune for more than a day, the keys have no response, no volume no dynamics, it's a piece of junk horrible, awful giagantic turd. A harmonica has better volume and dynamics than my piano.

I love my software pianos soooo much. And I hate my acoustic piano so much, that if there weren't an ordinance I would set it on fire in my front lawn.

My favorite software piano is just sampled 3 times and not even on every note. It's so beautiful sounding and responsive, and the resonance emulator is so warm.

My real piano is infinite. It's infinite in the ways that it sucks.

No doubt that's true...he certainly pounded much louder than Jerry Lee Lewis.

Jerry still has him beat with the "playing by foot" method.

Snazzy

I won't speculate about what the composers of the past might do with today's technology, suffice to say that many were noted for pushing the envelope in one way or the other. I'd think they would rather embrace it--not for being alike, but for all the ways it's different.

The core of this discussion has been around all the things you can do with an acoustic that a DP cannot (yet) replicate in a live performance. Probably 95% of the pianists out there will never hit that level of technical prowess to take advantage of those differences, but that's neither here nor there. I don't think anyone (even Snazzy) is making the case that DPs are equal to an acoustic on all fronts of live performance.

But that's really only half of the equation. No DP is *just* an approximation of an acoustic. They all--from the most basic level--have additional bells and whistles that give them their own feature set beyond that of which an acoustic is capable. In that regard, they cease to become a simple replacement for an acoustic piano, and become their own unique instrument.

As a musician, I find that flexibility as mind and technique expanding as pushing the envelope to get the most out of an acoustic. Layering, MIDI controls, pitch bend wheels--all feed the creative instinct as much as working the damper pedal and pounding notes on an acoustic. I'd like to think the great composers who came before us would have loved that flexibility as much as I do.

Later,

K

Yep, that makes them handier than a shirt pocket...I particularly love layering piano with string ensemble...I have to play at extra low volume sometimes, as the wimmen in the neighborhood gets all emotional....'spect that kind of sound brings a tear to even old whatsisname...the guy with the old Blunter....Erwin, is it?

Not to forget, K, we don't have to tune the little darlings either.

Going by Lawrence's report on another thread about the CP-1, I think were going to have ourselves a convoy...just to bring in all those pre-ordered Yammies.

Snazzy