Topic: What is all that velocity curves really about?

Seriously, though I technically understand how it works, why somebody wants a nonlinear behaviour? Some Jazz or Pop pianists who play only in mezzo forte - forte diapazone? For classical music I believe the dynamic range and linearity is very important. Ok, I can imagine that in theory velocity curve can be used to make it actually linear solving the issues with a controller keyboard that have some buggy nonlinear sensors behaviour or, what is more realistic and interesting, some mechanical nonlinearity. But I'm quite skeptical about that such data can be gathered by users, I'd rather suppose it can be researched precisely by keyboard manufacturer or a Pianoteq engineers for the specific keyboard model. Your opinions?

Re: What is all that velocity curves really about?

Jazz pianists only playing mezzoforte? Where did you get that?

Re: What is all that velocity curves really about?

Now let,s face it, Andree, nothing concerning the living beings acts totally linear; most of the time a log or e-function better describes things. like our hearing for instance. So it is quite understandable, that also our muscles are most of the time actuated in a non-linear way.

So, what are we talking about then?? It is one of the pretty things with virtual machines, that you can taylour things so that they work to your personal satisfaction.

Of course, when one has to switch between a certain acoustic piano and the virtual one, it is wise, to match the velocity settings more or less to get the same results.

I see, that there are people here on the forum, who like to share their settings. But, that is rather ridiculous in my opinion; one needs to find his own settings. Its like playing the clarinet or other reed instruments; the strength of a certain reed works well for someone, while the other cannot produce the same soundquality with that  reed. So we all work our reeds.

Are you actually playing a virtual piano, and what is your experience with it??

Greetings,

Geert

Re: What is all that velocity curves really about?

robsogge wrote:

Jazz pianists only playing mezzoforte? Where did you get that?

Mate, don't be so serious about that statement, please :)


@Mr. geert

Totally agree about the human nonlinearity. There may be an analogy with dB metering and adjustments for loudness which are not naturally linear but rather adopted for human perception. But I don't see how velocity curves helps for controlling the keyboard. When I tested it adding some nonlinearity for increasing (or decreasing) the highest velocity dynamics or lowest or middle I can't say it helps. Achieving something in one place sacrifices in the other. Linear is the best for me for now at least. Look I'm not trying to insult anybody, just sharing my experience and just want to hear yours. And, answering you question, yes, sure I used to spent some time with blacks and whites both acoustic and digital.

Last edited by AKM (11-11-2014 12:42)

Re: What is all that velocity curves really about?

AKM wrote:

Seriously, though I technically understand how it works, why somebody wants a nonlinear behaviour? Some Jazz or Pop pianists who play only in mezzo forte - forte diapazone? For classical music I believe the dynamic range and linearity is very important. Ok, I can imagine that in theory velocity curve can be used to make it actually linear solving the issues with a controller keyboard that have some buggy nonlinear sensors behaviour or, what is more realistic and interesting, some mechanical nonlinearity. But I'm quite skeptical about that such data can be gathered by users, I'd rather suppose it can be researched precisely by keyboard manufacturer or a Pianoteq engineers for the specific keyboard model. Your opinions?

I don't have the Pianoteq software, but based on my experience with my Casio, the voice I like best is "mellow," while it is good over all, it definitely lacks brightness in the upper register.  I believe velocity curves would be used to compensate for the output of a a given MIDI keyboard controller, which are obviously not created equal.  It's not a question of classical style or jazz style, rather getting a given MIDI equipment assemblage to be playable and pleasant sounding

By the way Chopin was clearly a "jazz" pianist.  That is to say he improvised beautifully, but only wrote down his compositions with great difficulty.  In essence he was an "ear" player.  This is well documented if you do the research.

As an aside, the hammers on an acoustic grand need to be "voiced".  It a process of adjusting the hardness of the hammers so that they have more of less same quality of sound as their neighbors.  It's been my experience that the treble hammers need to be harder, while the bass section would be softer in general as in my opinion you don't want a twangy bass.  Hammers need to be replaced, not just because the felt is worn out, but it can become too hard and brittle over time depending on the climate.

Of course velocity and hammer brightness are not the same thing.  Generally one does not adjust the "velocity" of an acoustic grand but on the other hand key dip, and the moment of jack release, and hammer drop along with the back check all play into it, not to mention the scaling of the strings themselves. Better instruments are better conceived from the very beginning, but also need to be properly regulated.

The idea of at least the Pianoteq Pro software is to give you the potential of minute adjustments to the sound of individual notes.  That would be a good thing, be you, Scarlatti, Chopin, Debussy, Hampton Hawes or Stockhausen.

You have to understand that on an electronic instrument, a key or note is just an "on / off" event. Velocity is in essence the loudness of a given on / off event.  However this is complicated by the fact that the timbre of an actual acoustic grand changes with increased "velocity."   Notes struck hard in the bass do become more twangy no matter the hammer hardness.  This is the challenge for the creator of an electronic piano voice, and also one of the reasons most electronic pianos sound shallow or thin, while being comparatively lifeless.

To conclude, it's less about musical style, and more about instrument setup, and how well an electronically created sound can deceive your ear.

Last edited by GRB (11-11-2014 13:00)
Pianoteq Pro 7.x - Kubuntu Linux 19.10 - Plasma Desktop - Hamburg Steinway

Re: What is all that velocity curves really about?

AKM wrote:
robsogge wrote:

Jazz pianists only playing mezzoforte? Where did you get that?

Mate, don't be so serious about that statement, please


.

sorry, I'm a bit touchy on the topic because jazz piano is my job... no hard feelings

Re: What is all that velocity curves really about?

robsogge wrote:
AKM wrote:
robsogge wrote:

Jazz pianists only playing mezzoforte? Where did you get that?

Mate, don't be so serious about that statement, please


.

sorry, I'm a bit touchy on the topic because jazz piano is my job... no hard feelings

I think "hip" players play with a lot  of "pop."  They know how to put it in the "pocket."  The better a player can play jazz, the better he or she will play alleged "classical".  Classical being mostly antique music which was written down due to the lack of any ability to record it.  The Miles Davis "Sketches of Spain" must be considered a "classic." at this point in time.

Last edited by GRB (11-11-2014 12:57)
Pianoteq Pro 7.x - Kubuntu Linux 19.10 - Plasma Desktop - Hamburg Steinway

Re: What is all that velocity curves really about?

AKM wrote:
robsogge wrote:

Jazz pianists only playing mezzoforte? Where did you get that?

Mate, don't be so serious about that statement, please


@Mr. geert

Totally agree about the human nonlinearity. There may be an analogy with dB metering and adjustments for loudness which are not naturally linear but rather adopted for human perception. But I don't see how velocity curves helps for controlling the keyboard. When I tested it adding some nonlinearity for increasing (or decreasing) the highest velocity dynamics or lowest or middle I can't say it helps. Achieving something in one place sacrifices in the other. Linear is the best for me for now at least. Look I'm not trying to insult anybody, just sharing my experience and just want to hear yours. And, answering you question, yes, sure I used to spent some time with blacks and whites both acoustic and digital.


Perhaps your set up is best with the velocity curve or non curve as it came out of the box.  That would be very understandable.  Just because a feature is there does not mean you have to use it.  However, I believe the velocity curve on Pianoteq would be used to subtly adjust the timbre of a given section, not necessarily the loudness.  In the end, you just have to adjust the sound and playability to what you like best or to what is most comfortable for you.  No one can do that better than yourself.

Last edited by GRB (11-11-2014 13:12)
Pianoteq Pro 7.x - Kubuntu Linux 19.10 - Plasma Desktop - Hamburg Steinway

Re: What is all that velocity curves really about?

GRB wrote:
robsogge wrote:
AKM wrote:

Mate, don't be so serious about that statement, please


.

sorry, I'm a bit touchy on the topic because jazz piano is my job... no hard feelings

I think "hip" players play with a lot  of "pop."  They know how to put it in the "pocket."  The better a player can play jazz, the better he or she will play alleged "classical".  Classical being mostly antique music which was written down due to the lack of any ability to record it.  The Miles Davis "Sketches of Spain" must be considered a "classic." at this point in time.

yeah brother

Re: What is all that velocity curves really about?

robsogge wrote:
GRB wrote:
robsogge wrote:

sorry, I'm a bit touchy on the topic because jazz piano is my job... no hard feelings

I think "hip" players play with a lot  of "pop."  They know how to put it in the "pocket."  The better a player can play jazz, the better he or she will play alleged "classical".  Classical being mostly antique music which was written down due to the lack of any ability to record it.  The Miles Davis "Sketches of Spain" must be considered a "classic." at this point in time.

yeah brother

Obviously you know the difference between being "hip" and "hep."  We all know Horace Silver was the "hard bop grand pop," but that's not puttin' Chopin down.

Last edited by GRB (11-11-2014 13:41)
Pianoteq Pro 7.x - Kubuntu Linux 19.10 - Plasma Desktop - Hamburg Steinway

Re: What is all that velocity curves really about?

I,m a fan of Bill Evans. But, just lioke Brubeck, they have such big hands........while mine are  way too small.

More on topic: for me the preset "Pianoteq"curve in my VPC-1 suits me wonderful. My guess is, that people at Kawai listened rather well to some Pianoteq virtual instruments.

Of course, the weighting of the keys also plays in this; if I remember well, our Andrej likes keys that do not have much inertia in the form of built-in weights, but it is my experience, that a proper weighted keybed helps me a lot in varying the touch-velocity.
So many people, so many meanings.....

Greetings,

Geert

Re: What is all that velocity curves really about?

geert wrote:

I,m a fan of Bill Evans. But, just lioke Brubeck, they have such big hands........while mine are  way too small.

...

Greetings,

Geert

I fell in love with Evans' music in my teens... I've always envied his hands too, but that's sort of a false problem imo. We small handers find our way to do things (I can play most tenths btw)

Re: What is all that velocity curves really about?

robsogge wrote:
geert wrote:

I,m a fan of Bill Evans. But, just lioke Brubeck, they have such big hands........while mine are  way too small.

...

Greetings,

Geert

I fell in love with Evans' music in my teens... I've always envied his hands too, but that's sort of a false problem imo. We small handers find our way to do things (I can play most tenths btw)

But not as easy as Art Tatum playing 10ths

https://www.youtube.com/watch?v=F737M573SYM

and:

https://www.youtube.com/watch?v=kACt0FM0Kf8

Ian

Last edited by Beemer (11-11-2014 15:17)

Re: What is all that velocity curves really about?

geert wrote:

if I remember well, our Andrej likes keys that do not have much inertia in the form of built-in weights, but it is my experience, that a proper weighted keybed helps me a lot in varying the touch-velocity.
So many people, so many meanings.....

Lol, it's not a matter of love (or like) tbh )) Just play Chopin #1 op. 10 etude and you'll get the idea of what I was talking about )))

Last edited by AKM (11-11-2014 15:19)

Re: What is all that velocity curves really about?

Beemer wrote:

But not as easy as Art Tatum playing 10ths


Ian

  he was incredible... not just because of the hands though

Re: What is all that velocity curves really about?

All my love goes to Erroll Garner, check him on videos, it is just something. You should really "see" it.

Re: What is all that velocity curves really about?

geert wrote:

More on topic: for me the preset "Pianoteq"curve in my VPC-1 suits me wonderful. My guess is, that people at Kawai listened rather well to some Pianoteq virtual instruments.

Ok, I beleive that I may be wrong, but for me that VPC1 factory curves look like a marketing trick, that they call "approved by". What I think is that they just randomly moved something slightly - it can't seriously spoil the whole thing - and thats it, now they can "sell" it to public as an "exclusive feature". What purpose are they solve? Do they mean Pianoteq is nonlinear, so they corrected it by their "approved" curves? Or their keyboard is not linear, but why different curves for different manufacturers then? I say again, there is a big chance I keep that I'm wrong and just don't understand or know something, but just shared my way of thinking. Reminds me "Super Mega Bass 3D" sticker on a boombox.

...what was the method for getting that VPC1 curves, can anyone guess?

Last edited by AKM (11-11-2014 19:35)

Re: What is all that velocity curves really about?

AKM wrote:
geert wrote:

More on topic: for me the preset "Pianoteq"curve in my VPC-1 suits me wonderful. My guess is, that people at Kawai listened rather well to some Pianoteq virtual instruments.

Ok, I beleive that I may be wrong, but for me that VPC1 factory curves look like a marketing trick, that they call "approved by". What I think is that they just randomly moved something slightly - it can't seriously spoil the whole thing - and thats it, now they can "sell" it to public as an "exclusive feature". What purpose are they solve? Do they mean Pianoteq is nonlinear, so they corrected it by their "approved" curves? Or their keyboard is not linear, but why different curves for different manufacturers then? I say again, there is a big chance I keep that I'm wrong and just don't understand or know something, but just shared my way of thinking. Reminds me "Super Mega Bass 3D" sticker on a boombox.

...what was the method for getting that VPC1 curves, can anyone guess?

From the Kawai website:   Preset Touch Curves  -  Normal, Ivory, Pianoteq, Galaxy, Alicia's Keys

Yeah, what's the difference between "normal" and "ivory"?  Why isn't there one called "golden" or "super funky"?  And they left out "Genuine Chopin".  Worst of all this particular controller doesn't have any knobs or sliders.  I would never consider it for my needs.  I'm a fan of Nord.

Last edited by GRB (11-11-2014 21:52)
Pianoteq Pro 7.x - Kubuntu Linux 19.10 - Plasma Desktop - Hamburg Steinway

Re: What is all that velocity curves really about?

GRB wrote:

Why isn't there one called "golden" or "super funky"?  And they left out "Genuine Chopin"

:DDDDDDDDDDDDDDDDDDD

Re: What is all that velocity curves really about?

AKM wrote:
GRB wrote:

Why isn't there one called "golden" or "super funky"?  And they left out "Genuine Chopin"

:DDDDDDDDDDDDDDDDDDD

The Casio PX-150 comes stock with the Mussorgsky velocity curve which many consider superior to the Rebikov curve.  You can literally paint portraits with the Mussorgsky, but if you are a  fan of pre-impressionism you choose the Rebikov.

Pianoteq Pro 7.x - Kubuntu Linux 19.10 - Plasma Desktop - Hamburg Steinway

Re: What is all that velocity curves really about?

Hi, new user here.

I think the important part is to know if your keyboard transmits all 127 velocity values, or has big jumps. From some other forums I have come to knwo that not all keyboards send the 127, so it's difficult to get full dinamic range with them. In part is important that software as pianoteq has a way of setting velocity curve response to overcome this problem.

Some specific midi controllers allow you to adjust the velocity curve, something not quite common on DP specially the entry level ones,so it is ncie to overcome this limitation in software pianos.

I think a database where people submitted the velocity values they can get out of their keyboard could be very usefull to make pruchasing decissions.

Re: What is all that velocity curves really about?

login wrote:

Hi, new user here.

I think the important part is to know if your keyboard transmits all 127 velocity values, or has big jumps. From some other forums I have come to knwo that not all keyboards send the 127, so it's difficult to get full dinamic range with them. In part is important that software as pianoteq has a way of setting velocity curve response to overcome this problem.

Some specific midi controllers allow you to adjust the velocity curve, something not quite common on DP specially the entry level ones,so it is ncie to overcome this limitation in software pianos.

I think a database where people submitted the velocity values they can get out of their keyboard could be very usefull to make pruchasing decissions.


I agree with you totally.  Another issue is exactly how is the velocity measured?  Is it the impact on the sensor or what?  Optically you could measure the speed of travel between two points.  I really have no idea how it is done electronically.

Pianoteq Pro 7.x - Kubuntu Linux 19.10 - Plasma Desktop - Hamburg Steinway

Re: What is all that velocity curves really about?

OK. First of all: I love the VPC-1 just because there are no sliders, knobs or other ugly and distracting things. Remember, it,s not an airoplane cockpit!

Of course, when you want a Midi master keyboard with all sorts of splits, CC sends, pitchbend, modulation, you use a different unit....
I have them all over here, and my experience is, that a comfortable piano-style touch is far more important for my playing pleasure then all those adjusters at hand.
Although a Korg Kronos X-88 could be satisfatory in both worlds??

Then Andrej: you often like to exaggerate a bit., in your evergoing suspicious mind towards manufacturers. You don,t need to buy the stuff....!
The velocity curve that i made myself, looks for 90% the same as the factory preset in the VPC. so it certainly is NOT a gadget only for selling this device.

Perhaps when you diminish critisizing somewhat, you get happier?? ( I hope).

The "velocity" is indeed measured as the time between  two contacts closing one at 10% keydown or so, and one at 90% keydown or so.
This is speed figure is directly coded into 1--127 in a lineair way.

Greetings.,

Geert

Re: What is all that velocity curves really about?

geert wrote:

Then Andrej: you often like to exaggerate a bit., in your evergoing suspicious mind towards manufacturers. You don,t need to buy the stuff....!
The velocity curve that i made myself, looks for 90% the same as the factory preset in the VPC. so it certainly is NOT a gadget only for selling this device.

Perhaps when you diminish critisizing somewhat, you get happier?? ( I hope).

Actually I'm not so much criticizing rather saying I'd like to understand. BTW, how you developed that yours velocity curve? Teach me please, no sarcasm. When I, say, follow the Pianoteq calibration instructions I get the different results all the time. The sum of them seems to me more or less as a linear graph.

Last edited by AKM (12-11-2014 02:29)

Re: What is all that velocity curves really about?

Ok, I thought out some calibration algorithm, that could be implemented into Pianoteq: the instruction ask you, in how many steps do you want to calibrate the graph and in how many passes. I'm thinking about choosing, say, 10 steps from ppp to FFF and another 10 back and, let me think, do it in about 300 passes, with no sound, no visual velocity feedback, just your muscle perception and some indicator of that events are received. With about 100 passes just for practice earlier. The intermediate position of all actual passes is the result. That approach seems to make some velocity curve I could believe being perceived linear by myself.

...even better, during the calibration the velocity indicator may actually show ideal velocity position for the certain step graphically, but not your actual velocity.

...thought more about it, I'd really be interested in trying it for scientific purposes, though I still don't clearly understand what's wrong with just linear.

...just thought more about it. First I thought that that advanced calibration procedure could actually play a sound from an ideal velocity setting for each step (not the real one) the same way I suggested for the visual feedback. But then I thought about that "ideal" volume linear curve - it could also be pre- calibrated by user with no actual numeric data feedback from the interface - to not let your brains fool you. Say, just by mouse blind dragging arrange 10 notes loudness from ppp to FFF linearly only by hearing. Provide some user adjustable dynamic range. Do it 100 times, get the intermediate result. That will build the pre- keyboard playing perceived volume linear curve. Or we can stay with Pianoteq scientifically modelled linear volume curve. And after that - go to calibration from the keyboard. And, opposite to what I suggested previously, in this case the actual sound feedback could be acceptable. And, if there is a will to go crazy about it, the intermediate calculations should be more or less advanced. I'd suggest to just truncate 10% of the values for a given step removing sequentially all the most extreme values, to get rid of occasionally made wrong entries, and only after that do the intermediate calculation of the remaining values.

...and, btw, maybe, if I'll not be so lazy as usually, I'll develop all that stuff by myself with a Reaktor software (it could be some other visual programming MIDI friendly software such as Max/MSP which I'm not so familiar). Just maybe.

...just to prove to myself that it was just a waste of time and stick to just linear.

Last edited by AKM (12-11-2014 13:45)