This custom search works much better than the built in one and allows searching older posts.
|
|
69902 Members
40 Forums
143537 Topics
2076760 Posts
Max Online: 15252 @ 03/21/10 11:39 PM
|
|
|
#1868262 - 03/25/12 02:06 PM
How many velocity levels for a good DP?
|
Full Member
Registered: 12/18/09
Posts: 29
|
How many dynamic levels and sample levels do you get out of your digital piano when stood alone, and are you satisfied with it? I ask because I tried a keyboard that played only three levels through its speakers, even though it sent 0-127 on MIDI. It had no adjustments on it, and to play "p" consistently was impossible on its cheap sprung keys. I noticed a cheap semi-weighted CDP-100 also claimed only three velocity levels... can that really be satisfactory? I mean for solo material.
I have used samplers that claim something like seven-nine velocity layers for each note though I tested them and, by sound, thought they could be pared down to five at a pinch. I wondered if for a budget digital piano, a pp-p-mf/mp-f-ff five layer dynamic with only three sample layers and not even interpolated would work, along with adjustable response. I'm only thinking aloud, so don't get over-exercised about it- any opinions on how few layers can be got away with are welcome.
|
|
Top
|
|
|
|
#1868276 - 03/25/12 02:25 PM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
4000 Post Club Member
Registered: 07/07/04
Posts: 4999
Loc: Vught, The Netherlands
|
I don't count the layers. I just play and if it responds as I expect it to, then it's doing its job.
I learned a long time ago not to base decisions on a spec sheet. You'll find your ears will carry you pretty far. (My old GranTouch only had 32 note polyphony. I never noticed any notes dropping out.)
|
|
Top
|
|
|
|
#1868278 - 03/25/12 02:30 PM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
3000 Post Club Member
Registered: 09/24/09
Posts: 3120
Loc: North Carolina
|
Don't confuse velocity values with sampling layers.
In principle, the keyboard can measure 127 levels of velocity. And the piano will respond with 127 different loudness levels.
The piano might have only 3 sampling layers, one for soft, one for medium, and one for hard. Based on the key velocity, the piano will choose one of the 3 sampling layers ... and then adjust the loudness to suit the velocity value.
We might imagine that velocities 1 through 42 will play using the first sample layer, velocities 43 through 84 with the second layer, and velocities 85 through 127 with the third layer.
So you'd get up to 127 loudness level, with three different layer timbres.
Maybe some of this is moot on cheapo keyboards? Those might not be capable of producing 127 loudness levels? I cannot say.
|
|
Top
|
|
|
|
#1868281 - 03/25/12 02:42 PM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
2000 Post Club Member
Registered: 02/20/10
Posts: 2456
|
Ears are much more important the spec sheets. Personally, I find that the single-layer Yamaha P-95 plays more musically than the 4-layer Casio PX-130. Some people feel differently. The GEM RP-X is a nice sounding module that uses a hybrid sampling + modeling approach, it has only one sampled layer, and I think most people agree that it sounds better than either of the other two I mentioned. The point is, it's all in the implementation, not the specs.
|
|
Top
|
|
|
|
#1868303 - 03/25/12 03:23 PM
Re: How many velocity levels for a good DP?
[Re: anotherscott]
|
Full Member
Registered: 12/18/09
Posts: 29
|
Personally, I find that the single-layer Yamaha P-95 plays more musically than the 4-layer Casio PX-130. Some people feel differently. The GEM RP-X is a nice sounding module that uses a hybrid sampling + modeling approach, it has only one sampled layer, and I think most people agree that it sounds better than either of the other two I mentioned. Thankyou for sharing your specific experience. I realise that on a stand-alone the samples will be limited, and for example I thought a separate p-pp, a mp-f and a ff tone layer, making 3 layers, might possibly be sufficient. The mid-tone in particular could be used on several volume levels. What I was after was how many velocity levels and sample levels players could distinguish in the digital pianos they have played, and how many they thought in practice were necessary. Assertions to trust my ear about keyboards I have never heard are not very useful in directing me towards a model...!
|
|
Top
|
|
|
|
#1868306 - 03/25/12 03:30 PM
Re: How many velocity levels for a good DP?
[Re: anotherscott]
|
Full Member
Registered: 02/27/10
Posts: 94
Loc: Spain. Cadiz.
|
I find that the single-layer Yamaha P-95 plays more musically than the 4-layer Casio PX-130 If I'm not wrong, the P85 is single layered, while the P95 has four layers.
_________________________
Started learning piano: 01 March 2010 - Ex: Yamaha P-85, Kawai ES-4 - Current: Kawai CA-63 - Videos
|
|
Top
|
|
|
|
#1868311 - 03/25/12 03:44 PM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
500 Post Club Member
Registered: 09/30/11
Posts: 597
Loc: Portugal
|
Theoretically it may be the case that all shades of sound necessary are contained in the loudest fff sample, and that all gradations below that, down to ppp can be obtained by subtle filtering - it is a matter of fashioning lower energy levels from a high start point. (This is not a statement of fact by any means - I said 'may be'!)
As far as 128 levels of volume/ potential timbres are concerned, it seems to me there is no reason what ever why that should be an industry standard. This was, after all, a protocol set up for communicating instruments in the very early 1980's. At that time, computers generally came with about 64k bytes of RAM or less. Now a cheap laptop will have at least 4GB of RAM - over fifty thousand times bigger...just for onboard RAM! So, apart from communication via MIDI, pianos and other keyboards could comfortably be dealing with many thousands of layers of volume/tonal character. Should they so wish.
_________________________
My piano is Roland HP 302 Other keyboards: Fender Rhodes Suitcase 88 piano (c. 1970), Yamaha SY85 (synth c.1991) Previously: Korg Polysix, Roland Juno 60, Ensoniq ESQ1 Favourite piano: Blüthner (6'early 20th century)
|
|
Top
|
|
|
|
#1868312 - 03/25/12 03:44 PM
Re: How many velocity levels for a good DP?
[Re: supertorpe]
|
Full Member
Registered: 03/09/10
Posts: 451
|
Many of the DP's seem to have 3-4 sample layers, while top virtual pianos like Ivory II and the Vintage D have 16-18 sample layers. Ivory II is interesting because you can vary the number of sample levels from 4 (I believe that's the correct number without looking) to 18 on its Steinway. That can give you some idea about how sample layers can affect the sound quality and playability. Of course there are many other variables involved, so it's just an example. But it's an interesting example.
_________________________
Macy
CVP-409GP, Vintage D, Ivory II GP's & American Concert D, Garritan Authorized Steinway, Alicia's Keys, EWQL Pianos, MainStage, iPad/forScore/PageFlip Cicada, Custom Mac MIDI/Audio Software Design, Macs Everywhere
|
|
Top
|
|
|
|
#1868313 - 03/25/12 03:46 PM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
1000 Post Club Member
Registered: 02/29/12
Posts: 1668
Loc: Ireland (ex England)
|
...how many velocity levels and sample levels players could distinguish in the digital pianos they have played, and how many they thought in practice were necessary... The point the replies were trying to make is that the answers to your questions are subjective. You can only trust your ear when making judgements about a DP, not specs.
_________________________
Richard
|
|
Top
|
|
|
|
#1868315 - 03/25/12 03:49 PM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
Full Member
Registered: 12/18/09
Posts: 29
|
re: Yamaha P-95 & Casio PX-130, the youtube page http://www.youtube.com/watch?v=UrM7TzRVWhI was useful and appeared to demonstrate 3 really discernable sample layers in the PX-130 with a change in volume too. Perhaps there was another one hiding in there but the three I recognised I might find enough. Now I need to find someone playing a real piece...
|
|
Top
|
|
|
|
#1868320 - 03/25/12 03:58 PM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
2000 Post Club Member
Registered: 02/20/10
Posts: 2456
|
Assertions to trust my ear about keyboards I have never heard are not very useful in directing me towards a model...! That's exactly the point. No answer here is going to be useful in directing you to a model. You have to play them. There is no answer to your question of how many velocity levels is good. There is no number that is big enough to assure that you will like the result, nor so small as to guarantee that you won't.
|
|
Top
|
|
|
|
#1868322 - 03/25/12 04:09 PM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
Full Member
Registered: 12/18/09
Posts: 29
|
I would have thought that since I was asking about dynamics and sample layers, it was obvious I was used to hearing them. I would think for instance that for someone used to hearing solo classical repertoire, it would be impossible for them to be satisfied by the Yamaha P-95 which quite clearly only has one sample layer. But we shall see.
|
|
Top
|
|
|
|
#1868365 - 03/25/12 06:08 PM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
500 Post Club Member
Registered: 09/30/11
Posts: 597
Loc: Portugal
|
Clavier_watcher, the youtube clip you linked to above states that the Yamaha P95 changes in volume only. If that is the case, then, in that respect, that is remiss of Yamaha, it seems to me. I tried a similar experiment on another known 1 layer Yamaha piano: CLP220 and it did indeed seem to vary only in volume. Surprising.
However, in answer to your question: how many sample layers do you need for a high quality simulation of a piano? My answer is 1, because volume, as has been pointed out above, is but one of the variables that are altered at each of the (theoretically virtually unlimited) steps that can be recognised by a DP, even if it only has one sample layer per note.
Masking, filtering, eq, different interactive algorithms etc are just some of the parameters that can be varied along with the volume.
Different sample layers may ultimately be seen as way stations from silence to fortissimo. But a program that is cleverly enough designed may be able to get from 0 - 60 with just the one fortissimo sample,I would have thought.
Of course I could be talking absolute bollocks here, but if I am, I imagine someone will tell me.
_________________________
My piano is Roland HP 302 Other keyboards: Fender Rhodes Suitcase 88 piano (c. 1970), Yamaha SY85 (synth c.1991) Previously: Korg Polysix, Roland Juno 60, Ensoniq ESQ1 Favourite piano: Blüthner (6'early 20th century)
|
|
Top
|
|
|
|
#1868369 - 03/25/12 06:22 PM
Re: How many velocity levels for a good DP?
[Re: toddy]
|
Full Member
Registered: 12/18/09
Posts: 29
|
However, in answer to your question: how many sample layers do you need for a high quality simulation of a piano? My answer is 1, because volume, as has been pointed out above, is but one of the variables that are altered at each of the (theoretically virtually unlimited) steps that can be recognised by a DP, even if it only has one sample layer per note. I was really thinking how limited the capacities of a digital piano could be whilst still getting away with it, hence I conceived of it simply in terms of volume and sample layers. I am aware of synthesizers/modellers and this is what Truepianos seems to do quite agreeably as you could check for yourself on another thread recently. What I would miss with more limited samples would be the sympathetic resonance.
|
|
Top
|
|
|
|
#1868376 - 03/25/12 06:39 PM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
500 Post Club Member
Registered: 09/30/11
Posts: 597
Loc: Portugal
|
..... What I would miss with more limited samples would be the sympathetic resonance. Surely this could be better achieved (and probably already is achieved) through computational means rather than multi-sampling?
_________________________
My piano is Roland HP 302 Other keyboards: Fender Rhodes Suitcase 88 piano (c. 1970), Yamaha SY85 (synth c.1991) Previously: Korg Polysix, Roland Juno 60, Ensoniq ESQ1 Favourite piano: Blüthner (6'early 20th century)
|
|
Top
|
|
|
|
#1868421 - 03/25/12 08:27 PM
Re: How many velocity levels for a good DP?
[Re: supertorpe]
|
2000 Post Club Member
Registered: 04/16/07
Posts: 2667
Loc: Pennsylvania
|
I find that the single-layer Yamaha P-95 plays more musically than the 4-layer Casio PX-130 If I'm not wrong, the P85 is single layered, while the P95 has four layers. It turns out that you are wrong, and the P95 also has only one layer. In fact, dewster has pretty convincing evidence that it's the exact same layer that is found in the P85. I've heard people say it has multiple layers in several places now. Not sure where that got started, but it's not correct. In response to the original post, I think people have given you a good idea of how many layers is typical. Personally I have a hard time with single layer pianos, but I don't find the lack of layers in, for example, a 3-layer piano to be particularly offensive. To me it's the actual timbre of the piano that matters, not how much or how smoothly it changes with velocity. I can't tolerate the timbre of my MP8 and it's not great in most other onboard sounds either. That's why I use software pianos. If Vintage D had only 4 layers, I would still like it (as long as no one told me it only had four layers, but then I'd know it's not the latest and greatest and I would be sad).
Edited by gvfarns (03/25/12 08:28 PM)
|
|
Top
|
|
|
|
#1868579 - 03/26/12 05:58 AM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
4000 Post Club Member
Registered: 07/07/04
Posts: 4999
Loc: Vught, The Netherlands
|
It's not cute, it's excellent advice.
Your research stage should be the stage of the various piano stores where you can play the keyboards in question, period.
A piano has to be played to be judged.
You can take the opinion of professional players and you will find that they always don't agree with each other. The only opinion that counts is yours ... and your opinion comes after you've actually played the pianos.
Sorry, that's not the answer you want, but that's the best advice you're going to get.
|
|
Top
|
|
|
|
#1868586 - 03/26/12 06:51 AM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
2000 Post Club Member
Registered: 02/20/10
Posts: 2456
|
I understand wanting some guidance about what pianos to try to find and play, before locating a store. But the question "how many layers should I look for" is almost like asking "how many knobs and buttons should it have." There's no right answer, and by itself, the parameter doesn't mean a lot.
A more useful query, IMO, would be something like "My needs are xxxxx, and I'm looking for something in the price range of up to about $yyy. What are some models I should look like?" Needs would be things like how portable it should be, the importance of any non-piano sounds or functions, whether it should have speakers, etc. Ideally you could also mention some models you've played in the past for reference, i.e. things you have liked the sound or feel of, or ones that you have not liked the sound or feel of.
If you're curious about technical specs, i.e. how many velocity layers a piano has, how well blended they are, how many notes are sampled vs. stretched, how long the decays are, how noticeable the loops are, whether there are various kinds of string resonances, and so forth, you can read dewster's informative evaluations in the DPBSD thread. But there is no single element among those that will define a superior piano; in number of velocity layers, you've just chosen one of many possible attributes to consider. Moreover, while it is of academic interest and may guide someone to understanding why they like certain models over others and what other ones may offer similar attributes, the fact is that even those tests will not tell you which of those pianos you will like best, and you can easily end up preferring one that "specs out" worse than some other in some notable way.
|
|
Top
|
|
|
|
#1868589 - 03/26/12 07:01 AM
Re: How many velocity levels for a good DP?
[Re: gvfarns]
|
2000 Post Club Member
Registered: 02/20/10
Posts: 2456
|
the P95 also has only one layer. In fact, dewster has pretty convincing evidence that it's the exact same layer that is found in the P85. I've heard people say it has multiple layers in several places now. Not sure where that got started It might have been me, as a while back I did post that I thought I heard at least two layers in it. But in going over some sound files with dewster, he persuaded me that it's just some clever use of filtering and such, and have since referred to it as single layer, Sorry if my earlier posts led anyone astray! (Though at least I never said I was certain it had more, though it is what I though to be the case.)
|
|
Top
|
|
|
|
#1868591 - 03/26/12 07:06 AM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
500 Post Club Member
Registered: 09/30/11
Posts: 597
Loc: Portugal
|
AnotherScot: the question "how many layers should I look for" is almost like asking "how many knobs and buttons should it have." There's no right answer, and by itself, the parameter doesn't mean a lot.
Yeah - it's a bit like trying to judge a film by doing comparative ratios of red/ green/ blue pixels. That analysis is going to yield all sorts of information about the dynamics and shape of the film, perhaps. But it would be very difficult indeed, from such a procedure, to find out if the film is any good.
Edited by toddy (03/26/12 07:09 AM)
_________________________
My piano is Roland HP 302 Other keyboards: Fender Rhodes Suitcase 88 piano (c. 1970), Yamaha SY85 (synth c.1991) Previously: Korg Polysix, Roland Juno 60, Ensoniq ESQ1 Favourite piano: Blüthner (6'early 20th century)
|
|
Top
|
|
|
|
#1868606 - 03/26/12 07:51 AM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
Junior Member
Registered: 03/16/12
Posts: 14
Loc: UK
|
At the OP, According to the website, the CDP100 has 3 types of sensitivity levels. I haven't read the manual and could be mistaken, but I'm thinking this is not the same as the 0-127 level that is generated for midi, rather, it is for how hard to press to deliver a 127. E.g. with the "Light" level, you need minimum force to get the piano to sound at max volume, whereas at the "Heavy" level, you'd need to apply a significant amount of force to get a loud value. Bottom line: You may be confusing velocity mapping to actual velocities being generated. The CP33 I'm looking at also has 3 velocity levels (and off, like the CDP100), and it definitely has a very wide range of response depending on how hard I press the keys.
|
|
Top
|
|
|
|
#1868613 - 03/26/12 07:59 AM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
2000 Post Club Member
Registered: 02/20/10
Posts: 2456
|
Sensitivity levels refers to changing how soft or hard you have to hit a key to achieve a certain MIDI velocity level. No matter what, on standard MIDI equipment, there are a total of 127 possible velocity levels. They can theoretically be mapped to anywhere between 1 and 127 velocity layer samples. A single sample can be triggered over multiple velocities, but that doesn't mean it will sound the same at all velocities... it can still be adjusted in volume and with filters. , Most DPs have between 1 and 4 velocity layers, and that's mostly whet we've been talking about here. The Kronos has 8. Software pianos may have more.
|
|
Top
|
|
|
|
#1868682 - 03/26/12 11:16 AM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
Junior Member
Registered: 02/04/12
Posts: 18
Loc: Russia
|
Clavier_watcher, As for dynamic levels, I think, the minimum amount of them is what the music requires (I mean, all what is written in the music: p, f, pp, ff,... ). And how many dynamic levels are therw in AP?  I think, countless amount. That is the maximum  And, as for the sample layers, it seems to me, sometimes modelled layers sound better than sampled ones - it depends on what you like.
|
|
Top
|
|
|
|
#1868775 - 03/26/12 01:40 PM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
Full Member
Registered: 12/18/09
Posts: 29
|
Perhaps I could go some way to averting certain dear friends from wasting their time. I did invite experiences of digital pianos along with their specifications, whilst wondering aloud how limited a digital piano could be whilst still being satisfying on its own. It was a pity this thread has attracted mainly those who can say nothing about these pianos but instead try to enforce dogmas about 'buying with ears', on pianos I never heard of. So let me take some time out to soothe these people with "You're oh so right", and then hopefully now they can leave it and let the rest of us chew the cud about digital pianos we have heard...!
|
|
Top
|
|
|
|
#1868783 - 03/26/12 01:54 PM
Re: How many velocity levels for a good DP?
[Re: Clavier_watcher]
|
4000 Post Club Member
Registered: 07/07/04
Posts: 4999
Loc: Vught, The Netherlands
|
Perhaps I could go some way to averting certain dear friends from wasting their time. I did invite experiences of digital pianos along with their specifications, whilst wondering aloud how limited a digital piano could be whilst still being satisfying on its own. It was a pity this thread has attracted mainly those who can say nothing about these pianos but instead try to enforce dogmas about 'buying with ears', on pianos I never heard of. So let me take some time out to soothe these people with "You're oh so right", and then hopefully now they can leave it and let the rest of us chew the cud about digital pianos we have heard...! Martin, your customer.
|
|
Top
|
|
|
|
#1868922 - 03/26/12 07:04 PM
Re: How many velocity levels for a good DP?
[Re: anotherscott]
|
Full Member
Registered: 06/21/11
Posts: 226
Loc: Sydney, AU
|
I understand wanting some guidance about what pianos to try to find and play, before locating a store. But the question "how many layers should I look for" is almost like asking "how many knobs and buttons should it have." There's no right answer, and by itself, the parameter doesn't mean a lot.
A more useful query, IMO, would be something like "My needs are xxxxx, and I'm looking for something in the price range of up to about $yyy. What are some models I should look like?" Needs would be things like how portable it should be, the importance of any non-piano sounds or functions, whether it should have speakers, etc. Ideally you could also mention some models you've played in the past for reference, i.e. things you have liked the sound or feel of, or ones that you have not liked the sound or feel of.
If you're curious about technical specs, i.e. how many velocity layers a piano has, how well blended they are, how many notes are sampled vs. stretched, how long the decays are, how noticeable the loops are, whether there are various kinds of string resonances, and so forth, you can read dewster's informative evaluations in the DPBSD thread. But there is no single element among those that will define a superior piano; in number of velocity layers, you've just chosen one of many possible attributes to consider. Moreover, while it is of academic interest and may guide someone to understanding why they like certain models over others and what other ones may offer similar attributes, the fact is that even those tests will not tell you which of those pianos you will like best, and you can easily end up preferring one that "specs out" worse than some other in some notable way. You cant get a more straightforward answer than that. This should be in a sticky thread along with an explanation of " my digital is broken - the top notes sustain too long and dont cut out like the rest of the notes".
|
|
Top
|
|
|
|
|
|