Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Concerns about buying the MP11SE due to possible MIDI issues

Do you have anything special in the setup?

What firmware do you have?

@dmd: I am happy for you :-) But please do not dismiss the reports of the other users who cannot consistently replicate your findings and get note-on values below 30. This is PianoTeq with a CA98 on Light+ velocity curve playing ppp/pp.


But as many have already said, I do not find this a major limitation of this action and you can always set a proper curve with external software. However, the OP asked about the MIDI note-on values one can get. And those raw values in my case are limited to the range [30-40, 100-110].

It may just be a natural variation in manufacturing where some units wound up more sensitive than others. Alternatively, it may have to do with wear and age - either positively or negatively. For example, I know that the Roland actions which I have had over the years have tended to be a bit stiff out of the box but to loosen up with use.

@arc7urus -

Your exhibit above looks a lot like what I was getting with an MP11SE using MIDI Monitor on a Mac. I didn't think to do a screen grab, but that's pretty much what I was seeing.

Also like you, this doesn't bother me in the least. When I'm using the Kawai as a MIDI controller for various external instruments, I feel perfectly "in control" of the relevant dynamic range.

Suggestion: Perhaps rather than our continuing to focus on our MIDI data values, we might offer the OP impressions of our experience playing acoustic piano VSTs through the Kawai keybed used as a controller... which I'm guessing is actually what, as a classical musician, he was primarily interested in anyhow.

Toward that end, I'll repeat what I've offered above: The MP11SE has proven itself to be a highly expressive keyboard both for its internal sounds and external control over grand piano VSTs (in addition to other styles, I play Chopin, Mozart and other "expressive" classical piano compositions). For any such purpose, I would highly recommend it.

- OneWatt

That's exactly right for me as well - an apt focus and description.

(I play mostly classical)

re: firmware ... I have whatever is current.

Cheers,

Jack

I agree




This is not the case, since I tested with the highest C, which I never play (It still has the grand feel sticker as new)

Very odd. I don't understand how some people seem to have no trouble whatsoever hitting any low velocity they want while others can't go below 30 (and this threshold of ~30 seems to be consistent)... I guess I'll have to make the two hour drive with a laptop and do some testing myself. Thanks for all the help

What is odd are the claims that is "easy" to get low note-on velocities. Have a look at other forums (e.g. pianoteq, audiobus, synthzone, DAW and VST forums). You will notice that minimum note-on velocities in the range 20-40 have been consistently reported whenever digital pianos are used as controllers.These reports have been going on for years. There are also limitations to maximum velocity, but the upper boundary is not as consistent. Having access to the full velocity range is usually not an issue with dedicated MIDI controllers.

My Clav cannot produce velocity 127. The highest I've seen was 108.
I could get a sledge hammer and beat really hard on the keys, trying to achieve 127 ... but you'll forgive me for not doing so.

At the opposite extreme ... it is indeed hard to get really low velocity values. But with care I can get numbers below 10. I've not seen the 20-40 minimum.
However, it is difficult to get those low numbers. I cannot get the low numbers while playing.
I can get them only when doing careful testing ... which is artistically pointless.
Is that the Clav's fault? Or is it just my heavy-handed tendencies? I cannot say.

I see. So then, how about something like the VPC1? It's supposed to be a MIDI controller, but it has an action that's very similar to the MP11.

Not sure about the VPC1. I suspect it will behave as the MP11. There are several VPC users around here who can report.

I guess the circuit maps (near) zero time to maximum velocity. Which is easy to archive with an unweighted action, but heavy weighted hammers don't move in almost zero time. So you bottom out at about 100-115.

This be tested by opening the piano, removing a key and its hammer, then pushing down all two/three sensors at the same time. The piano should report velocity 127 as a result.


In digital pianos there is a silent note cutoff (when the hammer failed to hit the string). This seems happens to be at some arbitrary point way before velocity zero and then the piano suddenly reports velocity 1.

Maybe this mapping was calibrated to get a somewhat realistic (acoustic piano like) response from the internal tone generator. Those MIDI values are used internally too and get sent out unchanged.

I'm not sure this is always the case. . .

The values coming out from the sensors are a kind of high-resolution timestamp. These values are then used as the input to a function that determines the attack velocity, which is then used by the sound engine to generate the corresponding sound. These internal values are not MIDI messages due to the precision required. The MIDI note-on velocity events are calculated by mapping (i.e. quantizing) the internal high-res values to a limited range according to a given curve. So, the process is the opposite you are describing since the MIDI values are not used internally... they are just a conversion of the internal high-res values.

I understand that the position and design of the sensors may make difficult to detect velocities at the upper and lower boundaries of the range. However, this will only have impact on the range of the internal values as it is always possible to map them to whatever MIDI range. It is also fine if a DP uses a calibrated curve that has an artificially limited MIDI velocity range if that somehow helps to better simulate an acoustic. However, the user should be able to select curves that cover the whole range. Most DPs have other instruments than acoustic pianos and many are used as VST controllers. So, this restriction makes no sense to me.

I hope nobody minds me resurfacing such an old thread. But since I came across it after some initial struggles with creating a velocity curve for the MP11SE for VSTs and this thread pops up on Google when you research the topic, I thought I should provide an update.

On my MP11SE, purchased new last week, I am, in fact, able to regularly hit velocities below 30.

The notable thing for me is that it's much easier to do this if I stop pressing the key right at the escapement notch. If I go all the way through the escapement, it is very hard (albeit not impossible) to do anything less than 20-25 or so. Coming from a Casio PX-560 without escapement, this took some getting used to.

I'm not sure how realistically that translates to an acoustic piano. It reminds me of the notion of "playing above the escapement" but that's not a technique I've ever tried to do on purpose until now, so I'm not sure how accurate the behavior is. My understanding is that on an acoustic, playing above the escapement helps play fast note sequences softly, but unless I'm mistaken, I don't think it's required to play single notes or chords as softly as possible. But it's something I've noticed consistently as I've been messing with the velocity curve in VSL to best control ppp-pp velocities.

I'm curious to see if anyone else with an MP11SE finds the above to be the case.

It is also extremely difficult -- but again, not impossible -- to hit 120+ velocities with the 'normal' touch curve. Reaching 127 with the normal curve requires some ridiculous slamming of the keyboard, to the point I worry about damage. So I prefer to adjust the curve to make those top velocities a little more reasonable.

So I'd say my MP11SE, at least, exhibits the full range of velocities. They just require more work than I'm used to. To a somewhat impractical degree, I'd say -- but then again, an acoustic has no upper limit to its velocities until something breaks... Perhaps it is more realistic for someone trying to play over an orchestra!

It sounds like a user calibration feature is needed, right? As in ...... for each of the 88 keys ... a machine is needed to strike at a particular velocity (that should really align with your own max velocity that you're capable of) .... and knowing when to stop, without damaging the key etc. And a software setting would allow you to set the upper limit value of 124 or so. Also noting that graded hammer weight etc means different force to generate the same velocity. Actually ...... we probably expect acceleration to be involved as well. But if velocity is what they estimate, then we just go with it. Sounds like calibration could involve a fair bit of work and time and effort ... for each key. And each person has different performance characteristics ..... so this could get fiddly.