|
Welcome to the Piano World Piano Forums Over 3 million posts about pianos, digital pianos, and all types of keyboard instruments. Over 100,000 members from around the world.
Join the World's Largest Community of Piano Lovers
(it's free)
It's Fun to Play the Piano ... Please Pass It On!
|
|
66 members (bobrunyan, anotherscott, AaronSF, apianostudent, beeboss, brdwyguy, benkeys, 15 invisible),
2,196
guests, and
389
robots. |
Key:
Admin,
Global Mod,
Mod
|
|
|
Joined: Oct 2007
Posts: 3,836
3000 Post Club Member
|
3000 Post Club Member
Joined: Oct 2007
Posts: 3,836 |
But if the main factor indeed is hammer weight (and corresponding counterweight in keys so that the overall static force is about 50g to push down a key), can digital pianos be modified so that more mass is added both to hammers and keys so it feels like a truck, too? ^_^ Don't worry. The action of the Yamaha piano you just ordered is considered "heavy" by most people. Unlike Kawai actions the static weight is pretty high, which is typical for Yamaha actions, while the dynamic weight is the same as for every digital piano. So you can look forward to your "truck". Lol. I love truck-ey pianos, as I'm used to them, it seems like I can control even-ness of phrasing and dynamics better. But I'm sure if was used more to light actions I would be more comfortable with them. I admire pianists who have so much experience that they can switch between very different actions and don't have any problems. It's almost like giving someone with a 9-5 office job suddenly re-stocking duties, lifting heavy boxes...not a true analogy, since piano is not supposed to be a "labour", but some amount of force is required (and that amount depends on the heaviness of the action), one cannot really cheat physics. (just my opinion, don't kill me. also, switching from a heavy action to a lighter but "real" (acoustic, upright perhaps) action seems like fun initially, everything is super easy and quick, but after some time, switching back to a heavy action feels laborious) I know what you are saying. I prefer heavy actions as well. Early on in my own DP research process many were advising me against the VPC-1 over an MP11SE which had a slightly lighter action especially when playing towards the fall board. I'm so used to a heavy action that I went with the VPC-1. I've never really felt a digital piano action (I'm not referring to some hybrids) that felt exactly like the real thing, so as long as you are not expecting perfection, you will be happy. My opinion on the VPC-1 is that it's action is slightly mushy and bouncier than the millennium 3 action on my RX-2 and SK-2 acoustics (which honestly felt the same to me). But this does not necessarily make the VPC-1's action bad at all. In fact it reminded me a lot of an Estonia 168/190's Renner action I tried out in Sarasota a few years ago. That "mushy" kind of action that you can easily adapt to. It's a fine action. I'm hoping you still give the VPC-1 and Pianoteq a tryout before you make a final decision. I could care less if one less VPC-1 is sold or if you don't take my advice. But I honestly think you are doing yourself a disservice by not at least trying to experience for yourself what the fuss is all about. Those two videos I posted by Hugh Sung (which honestly I never saw before yesterday) explain exactly what I've been trying to say. It's all about control and your ability to shape a phrase that makes Pianoteq such a compelling product, and the pianos sound phenomenal in the latest version. I guess you could get any DP and just use Pianoteq with it, but with the VPC-1 you are giving self an advantage right off the bat because of that collaboration Kawai had with Moddart in the development phase of the VPC-1. Velocity curves are very important and in the VPC-1 it is fine tuned by professional sound engineers specifically for Pianoteq. If the VPC-1 had a bad action I would not advise it as a choice and as you have seen yourself, it is pretty good. You might want to think of it this way. Pianoteq gives you the ability to strike the piano with one finger in at any moment in time 11176 (the maximum allowed by MIDI) different ways when spread across 88 keys whereas with the use of sampled sounds whether they be onboard or via VST's you only a palette of around 880 possible ways to strike the piano with one finger. If you were to use all your fingers just multiply those numbers by a factor of 10. This does not include all the other variables that Pianoteq takes into account like pedaling techniques, resonsances etc.. Compared to an acoustic, even Pianoteq is by definition limiting because based on MIDI technology today, there's just so much you can do with those 1's and 0's, but why hamper your creativity right off the bat? Pianoteq gives you the best chance to express yourself more accurately than sampled sounds. In my opinion, like Hugh Sung says, it's too limiting that you feel like you are handcuffed and that's exactly how I experience it.
Last edited by Jethro; 06/15/19 12:35 PM.
|
|
|
|
Joined: Feb 2019
Posts: 2,345
2000 Post Club Member
|
2000 Post Club Member
Joined: Feb 2019
Posts: 2,345 |
Well I'm off to play on a concert grade Yamaha grand piano for the first time in my life. I have a Casio AP-470 and I suspect there will be a world of difference in the actions.
All this discussion I have been following will be made real to some small extent by being able to do a physical comparison.
Definitely leave a feedback after you try it! And good luck (if it's for your recital) The difference is huge, and I think I get some idea and about what dynamic weight is when it comes to playing. That dynamic weight means there is a much greater physical connection between the effort you are making and the loudness. With the digital piano it is so easy to get a different velocity I think it is hard to control. With the acoustic grand there is actual physical effort required to throw that hammer and I felt I was able to regulate what I was playing far more. I want loud I've got to put effort into it, with my DP, I want loud I've just got to move my finger faster and there is little in the way of inertia to overcome.
|
|
|
|
Joined: Sep 2009
Posts: 14,439
Yikes! 10000 Post Club Member
|
Yikes! 10000 Post Club Member
Joined: Sep 2009
Posts: 14,439 |
I think you've become lost in the numbers. Pianoteq gives you the ability to strike the piano with one finger in at any moment in time 11176 different ways when spread across 88 keys whereas with the use of sampled sounds ... you only a palette of around 880 possible ways to strike the piano with one finger. Woe is me! I have only 880 ways to play a note. I'm missing out on the other 10295 ways. I'm doomed!
|
|
|
|
Joined: Oct 2007
Posts: 3,836
3000 Post Club Member
|
3000 Post Club Member
Joined: Oct 2007
Posts: 3,836 |
I think you've become lost in the numbers. Pianoteq gives you the ability to strike the piano with one finger in at any moment in time 11176 different ways when spread across 88 keys whereas with the use of sampled sounds ... you only a palette of around 880 possible ways to strike the piano with one finger. Woe is me! I have only 880 ways to play a note. I'm missing out on the other 10295 ways. I'm doomed! You realize when you play an acoustic that those levels raise to infinity. How do shape a phrase and maintain a beautiful line if you are only playing with a fraction of the palette sounds available? It's like when my little brother couldn't understand why my lego homes always looked better than his. I was playing with regular bricks and I always gave him the Duplos. He never figured it out. LOL
Last edited by Jethro; 06/15/19 02:49 PM.
|
|
|
|
Joined: Dec 2016
Posts: 326
Full Member
|
OP
Full Member
Joined: Dec 2016
Posts: 326 |
I'm hoping you still give the VPC-1 and Pianoteq a tryout before you make a final decision. I could care less if one less VPC-1 is sold or if you don't take my advice. But I honestly think you are doing yourself a disservice by not at least trying to experience for yourself what the fuss is all about. Those two videos I posted by Hugh Sung (which honestly I never saw before yesterday) explain exactly what I've been trying to say. It's all about control and your ability to shape a phrase that makes Pianoteq such a compelling product, and the pianos sound phenomenal in the latest version. I guess you could get any DP and just use Pianoteq with it, but with the VPC-1 you are giving self an advantage right off the bat because of that collaboration Kawai had with Moddart in the development phase of the VPC-1. Velocity curves are very important and in the VPC-1 it is fine tuned by professional sound engineers specifically for Pianoteq. If the VPC-1 had a bad action I would not advise it as a choice and as you have seen yourself, it is pretty good.
Did it ever occur to you that you sound like an exclusive Pianoteq+VPC1 salesman...? Anyway, it won't be so easy. To try VPC1, first, I have to return Yamaha P-515 (which might or might not happen, depending on whether I like it or not), then I have to buy VPC1 (unfortunately, -10% sale I found on <acoustic prolongation of a sound>.com ends on June 22nd, so I'd have to buy it for a full price). Then I have to buy Pianoteq and start tweaking and playing until I feel I "tried" it enough. I'm afraid at that point I'd be unable to return neither VPC1 nor Pianoteq. On the other hand, now that I'm waiting for Yamaha P-515 I'm weirdly sad for not getting VPC1. I really like the design and even though it's action is not super realistic, it's better than anything else I tried (maybe except Juno DS88, which felt amazingly realistic, but a bit too light. I wish there was a DP with Juno DS88 feeling, but heavier). Sometimes, I base my decisions a lot on my emotions so I might end up getting VPC1 just because of this. Plus I'm a computer nerd so I feel like I would enjoy playing around with curves just to analyze the interaction between player and instrument.
Last edited by Chopin Acolyte; 06/15/19 02:57 PM.
|
|
|
|
Joined: Apr 2007
Posts: 7,268
7000 Post Club Member
|
7000 Post Club Member
Joined: Apr 2007
Posts: 7,268 |
Pianoteq gives you the ability to strike the piano with one finger in at any moment in time 11176 different ways when spread across 88 keys whereas with the use of sampled sounds ... you only a palette of around 880 possible ways to strike the piano with one finger. No, that’s not correct. Sample based pianos morph between the layers. Which is why you have 127 different ways for each of their 88 keys and so it’s the same number as on Pianoteq. You’re spreading a lot of misinformation based on your lack of understanding how things work. Which kind of speaks for itself about the credibility of your other statements
Last edited by CyberGene; 06/15/19 03:04 PM.
|
|
|
|
Joined: Dec 2016
Posts: 326
Full Member
|
OP
Full Member
Joined: Dec 2016
Posts: 326 |
No, that’s not correct. Sample based pianos morph between the layers. Which is why you have 127 different ways for each of their 88 keys and so it’s the same number as on Pianoteq. You’re spreading a lot of misinformation based on your lack of understanding how things work. Which kind of speaks for itself about the credibility of your other statements My guess is, that as far as volume is concerned, 127 should be well enough (can you hear difference between 61 and 62?) However, it's the sampling of timbre that worries me. In my original post I mentioned the Steinway in the local music school and how beautifully changes timbre from soft to a brighter tone while never sounding unpleasantly, even if one puts tremendous amount of force into it...it always have some new shade of timbre to surprise us with. So the question is: how sparsely is timbre sampled on digital pianos? If the timbre never changes (one sample), it's not going to be realistic - recording a piano played softly and just playing it louder is not the same as recording the piano at higher dynamic markings...
|
|
|
|
Joined: Apr 2007
Posts: 7,268
7000 Post Club Member
|
7000 Post Club Member
Joined: Apr 2007
Posts: 7,268 |
Digital pianos around 2000 used to have three layers, probably a pp, mp, ff. 20 years later it’s not officially listed in the specifications for hardware digital pianos anymore, so we can only guess. And for software ones it varies but is probably around 10. One can argue that having 10 layers covering everything from ppp to fff and blending/morphing would effectively give you the same result as sampling 127 velocities.
I think real pianos sound, well, real, because there are so many tiny inconsistencies and even stochastic factors. Neither a sample based piano nor Pianoteq can give you that. In my entirely biased opinion Pianoteq to my own ears sounds even more monotonous than sampled pianos. Jethro and others won’t agree. But you can always download Pianoteq and try it. A few black keys are disabled but other than that it’s fully functional. With 6.5 there’s been considerable improvement indeed. I liked it to some degree although I’m usually very sensitive to the typical modeled artifacts. It still has some road to go but is a good piano considering price, performance, size, configurability.
Last edited by CyberGene; 06/15/19 03:24 PM.
|
|
|
|
Joined: Oct 2013
Posts: 3,868
3000 Post Club Member
|
3000 Post Club Member
Joined: Oct 2013
Posts: 3,868 |
@Jethro : You said : “Pianoteq gives you the ability to strike the piano with one finger in at any moment in time 11176 (the maximum allowed by MIDI) different ways when spread across 88 keys â€.
The actual number is 127 velocities with plain MIDI, multiply by 128 if you add HiRes Velocity (16256). Pianoteq does support this extension. And some digital piano does send extended velocities (some Casio). However, the Piano Phoenix of Adele H is designed from confirmed pianist needs and the number of useful velocity seems (according to the maker) around 1300 velocities. The Steinway Spirio is around the same numbers (1020). No needs of 16256.
But I have an hard time to hit multiple notes and not change the velocity between 2 hits. The 127 limit won’t impact my play.
|
|
|
|
Joined: Oct 2007
Posts: 3,836
3000 Post Club Member
|
3000 Post Club Member
Joined: Oct 2007
Posts: 3,836 |
@Jethro : You said : “Pianoteq gives you the ability to strike the piano with one finger in at any moment in time 11176 (the maximum allowed by MIDI) different ways when spread across 88 keys â€.
The actual number is 127 velocities with plain MIDI, multiply by 128 if you add HiRes Velocity (16256). Pianoteq does support this extension. And some digital piano does send extended velocities (some Casio). However, the Piano Phoenix of Adele H is designed from confirmed pianist needs and the number of useful velocity seems (according to the maker) around 1300 velocities. The Steinway Spirio is around the same numbers (1020). No needs of 16256.
But I have an hard time to hit multiple notes and not change the velocity between 2 hits. The 127 limit won’t impact my play. Interesting. It's even more than I thought. Explain this HiRes Velocity to me please.
|
|
|
|
Joined: Feb 2010
Posts: 6,730
6000 Post Club Member
|
6000 Post Club Member
Joined: Feb 2010
Posts: 6,730 |
But I have an hard time to hit multiple notes and not change the velocity between 2 hits. The 127 limit won’t impact my play. The value of having a certain number of velocity levels has nothing to do with one's ability to generate or repeat a particular value on demand.
|
|
|
|
Joined: Dec 2016
Posts: 326
Full Member
|
OP
Full Member
Joined: Dec 2016
Posts: 326 |
The value of having a certain number of velocity levels has nothing to do with one's ability to generate or repeat a particular value on demand.
IMHO it does. If best pianists in the world cannot repeat the same velocity level (e.g.) #4950 out of 16256 and finest ears in the world cannot tell the difference between #4950 and #4951, heck even between #4900 and #5000, what are all those levels for, then? Even with a well-regulated acoustic piano, if you take a machine that strikes the key with precisely calculated force won't produce the very same sound twice in a row, why strive for that amount of precision, digitally? What I'm saying is that we should aim for some reasonable balance. One/two/three timbre levels are perhaps too few, what's enough? 10? 20? 50? I don't know. Velocity levels? I think 127 is enough, at that point one can't tell the difference anymore, but if you think you can tell the difference, we should definitely do a test. Both playing and listening and see what's the typical margin at which one deems the level being as same. Make a pianist play same note/chord at some dynamic marking when he thinks he plays it at the same velocity and see in a software what the distribution looks like (gaussian?), i.e. standard deviation as a % of the mean loudness. If it's too little, we might have a problem. If it's spread out a little and people would still say "yeah he played all chords mf, very even and nice", then making levels more fine might be pointless. Is my logic wrong here? I think it's obvious the number of needed levels relates somehow to what people can reasonable produce and hear in practice.
Last edited by Chopin Acolyte; 06/15/19 06:07 PM.
|
|
|
|
Joined: Feb 2010
Posts: 6,730
6000 Post Club Member
|
6000 Post Club Member
Joined: Feb 2010
Posts: 6,730 |
For one thing, let's say you can only reliably repeat a note within 10 digits out of the standard MIDI 127 (i.e. plus or minus 5, which would still be pretty good). Repeating numbers 50, 51, 52, 53, 54, etc. would all be different, because 50 would repeat as something between 45 and 55, 51 would be something between 46 and 56, and so on.
And if you were to use the "+/- 5" as a rationale for having far fewer values, (i.e. the only values from 30 to 60 would be 30, 35, 40, 45, 50, 55, 60 on the basis that that's as close as you can intentionally differentiate your strikes anyway), crescendos would be far less smooth, i.e. if every subtle increase between 30 and 60 were rounded to the nearest 5. In reality, 33 and 37 should not sound the same, though both would round to 35. You might not be able to reliably produce a 33 or a 37, but the crescendo does not demand that. What the crescendo wants is for the second note to be louder than the first.
It's really not about your ability to repeat or generate something on demand. It's about the ability for the keyboard to respond to exactly what you play, regardless of any conscious intent. And uncontrolled small variations is part of what makes something sound human.
|
|
|
|
Joined: Dec 2016
Posts: 326
Full Member
|
OP
Full Member
Joined: Dec 2016
Posts: 326 |
For one thing, let's say you can only reliably repeat a note within 10 digits out of the standard MIDI 127 (i.e. plus or minus 5, which would still be pretty good). Repeating numbers 50, 51, 52, 53, 54, etc. would all be different, because 50 would repeat as something between 45 and 55, 51 would be something between 46 and 56, and so on.
And if you were to use the "+/- 5" as a rationale for having far fewer values, (i.e. the only values from 30 to 60 would be 30, 35, 40, 45, 50, 55, 60 on the basis that that's as close as you can intentionally differentiate your strikes anyway), crescendos would be far less smooth, i.e. if every subtle increase between 30 and 60 were rounded to the nearest 5. In reality, 33 and 37 should not sound the same, though both would round to 35. You might not be able to reliably produce a 33 or a 37, but the crescendo does not demand that. What the crescendo wants is for the second note to be louder than the first.
It's really not about your ability to repeat or generate something on demand. It's about the ability for the keyboard to respond to exactly what you play, regardless of any conscious intent. And uncontrolled small variations is part of what makes something sound human. Subdivision should be so fine that it's negligible when compared to what pianist can consistently repeat. If he can repeat within +/- 5 units, than it's up to people to decide whether 1 is negligible against 5. At least this principle works in physics. Want to model waves on a finite grid? Choose the grid spacing so that it's much smaller than the wavelength. Factor of 10, or 20 is often enough. Using the same spacing without leaving no margin is not reasonable.
|
|
|
|
Joined: Feb 2010
Posts: 6,730
6000 Post Club Member
|
6000 Post Club Member
Joined: Feb 2010
Posts: 6,730 |
I don't fully understand your post.
One more point though... "+/- 5" from your target also doesn't mean that each number in that range has an equal chance of being sounded. Probability is that the numbers closer to your target should be hit more often than the ones farther from your target.
Yet another variable is the skills of different players. One player may be able to stay closer to the target than another.
But if you don't think that having more values than you can consciously repeat is of value, this would be an interesting experiment: With software (like Pianoteq + a MIDI monitor), see what encompasses a repeatable value for you; then use software to round all input to a small enough number of output values such that you can reliably trigger each of the available values (by coming close enough that it would be rounded to that figure). You might end up with as few as about a dozen values that, with rounding, are entirely repeatable for you. So then use MIDI filtering/mapping so that Pianoteq (or whatever) only sees those dozen values regardless of what you play (by always picking the one you're closest to). Even though it will be at the limit of your repeatability, I think you will find it to play terribly.
|
|
|
|
Joined: Sep 2009
Posts: 14,439
Yikes! 10000 Post Club Member
|
Yikes! 10000 Post Club Member
Joined: Sep 2009
Posts: 14,439 |
It doesn't matter. My car can go at infinitely many speeds, but the speedometer only shows integers. I can't possibly know that I'm traveling 47.3 MPH when it reads 47 MPH. But it doesn't matter. The environment may present us with an infinite variety ... but the human senses are limited. For practical purpose, if you can't sense it, it doesn't exist. Again ... don't get lost in the numbers. Focus on the senses, the music, the art. Focus on the results. The numbers don't matter. It matters even less when you realize that the piano keyboard can produce only 127 possible velocities for a note. I don't know where you came up with that 11176 number. I think you've become lost in the numbers. Pianoteq gives you the ability to strike the piano with one finger in at any moment in time 11176 different ways when spread across 88 keys whereas with the use of sampled sounds ... you only a palette of around 880 possible ways to strike the piano with one finger. Woe is me! I have only 880 ways to play a note. I'm missing out on the other 10295 ways. I'm doomed! You realize when you play an acoustic that those levels raise to infinity. How do shape a phrase and maintain a beautiful line if you are only playing with a fraction of the palette sounds available? It's like when my little brother couldn't understand why my lego homes always looked better than his. I was playing with regular bricks and I always gave him the Duplos. He never figured it out. LOL
|
|
|
|
Joined: Sep 2009
Posts: 14,439
Yikes! 10000 Post Club Member
|
Yikes! 10000 Post Club Member
Joined: Sep 2009
Posts: 14,439 |
This is an interesting proposal for an experiment. I think you need to add one more thing: a manner in which to judge the results. If you don't think that having more values than you can consciously repeat is of value, this would be an interesting experiment: With software (like Pianoteq + a MIDI monitor), see what encompasses a repeatable value for you. Then use software to round all input to a small enough number of output values such that you can reliably trigger each of the available values (by coming close enough that it would be rounded to that figure). You might end up with as few as about a dozen values that, with rounding, are entirely repeatable for you. So then use MIDI filtering/mapping so that Pianoteq (or whatever) only sees those dozen values regardless of what you play (by always picking the one you're closest to). Even though it will be at the limit of your repeatability, I think you will find it to play terribly. You need to measure the sonic result, right? I don't care about variation in MIDI values ... except to the extent that it produces an audible variation.
|
|
|
|
Joined: Feb 2010
Posts: 6,730
6000 Post Club Member
|
6000 Post Club Member
Joined: Feb 2010
Posts: 6,730 |
My car can go at infinitely many speeds, but the speedometer only shows integers. I can't possibly know that I'm traveling 47.3 MPH when it reads 47 MPH.
Moreover, you CAN travel at 47.3. Your ability to intentionally drive at exactly 47.3 on demand may not exist, but imagine if some science fiction car could only drive at whole integers, and when gradually accelerating (i.e. as in a crescendo), it would instantaneously jump from each integer value to the next, what a rocky trip that would be!
|
|
|
|
Joined: Sep 2009
Posts: 14,439
Yikes! 10000 Post Club Member
|
Yikes! 10000 Post Club Member
Joined: Sep 2009
Posts: 14,439 |
If you sense it, it matters. If you cannot, it doesn't.
|
|
|
|
Joined: Apr 2007
Posts: 7,268
7000 Post Club Member
|
7000 Post Club Member
Joined: Apr 2007
Posts: 7,268 |
My car can go at infinitely many speeds, but the speedometer only shows integers. I can't possibly know that I'm traveling 47.3 MPH when it reads 47 MPH.
Moreover, you CAN travel at 47.3. Your ability to intentionally drive at exactly 47.3 on demand may not exist, but imagine if some science fiction car could only drive at whole integers, and when gradually accelerating (i.e. as in a crescendo), it would instantaneously jump from each integer value to the next, what a rocky trip that would be! Yes, but speed is a continuous function, it's a derivative of movement over time. However in pianos each strike has a single velocity, and then there's another strike, but the two are not connected. They are separate because you have to restrike. You can't make the piano produce a continuous change of velocity because strikes are separate events and not a function, hence it's difficult (or even impossible) to perceive steps. If a pianist can prove that presented with a piano that produces regular MIDI values he can consistently hit a particular value from 1-127 interval, then we can eventually move on to measuring his abilities in hitting exact HD-value. But other than that having HD-MIDI values is absolutely meaningless. And then there's another angle. The fact that a keyboard can produce HD-values doesn't mean its precision is the same. MIDI-controllers usually scan the triple (or double) sensors of the entire keyboard in succession. They are not event-driven, i.e. the controller logic won't get triggered the moment a particular sensor is active. Instead they will read it (it's a simple contact switch) at some point and will at that point in time conclude it's activated. However due to the scanning sequence it may turn out that the sensor was activated 87 key-scans earlier. Although the granularity of the MIDI values might be higher, the precision is lower.
Last edited by CyberGene; 06/15/19 07:17 PM.
|
|
|
|
|
|
Piano
by Gino2 - 04/17/24 02:34 PM
|
Piano
by Gino2 - 04/17/24 02:23 PM
|
|
Forums43
Topics223,408
Posts3,349,457
Members111,637
|
Most Online15,252 Mar 21st, 2010
|
|
|
|
|
|