Optimal VPC1 Curve using Midi Shape Shifter

why did you use the shape shifter instead of the VPC1 curve editor ?

I looked up the Wiki documentation, which strikes me as needing to be read by the type of geek I'm not, just to winkle out a bit of the what-else-to-this there is - so there's plenty but I'm not the type to handle it, found THAT out.

However while I was at it, saw that the PreSonus AudioBox 44VSL I bought, to interpose (since this box is a cheaper device than the KB so burnout via USB 2/3 misplay won't matter so much) between the VPC1 and the laptop USB it ultimately will connect to, IS ALSO probably a perfect solution to hosting MIDI Shape Shifter. Since not only did PreSonus make AudioBoxes low-latency, they followed through with low-latency in the DAW which comes with it. Ta da : the host!

What's more, it looks as though this method will sidestep any need for the VPC1 Editor and its waittimes at start. (Mac users notice this replacement-role also.)

Plus, during past discussions I held that curves ending on MIDI 127, after running lower than linear, get improper above ff, bad more bluntly. What do you know, MSS's inventor, in his Youtube demo section describing the Power Curve you use, shows that rolling off the curve would be dead easy, in the steeply-rising region 127 would otherwise be approached.

All up, had to vote Yes.

Thank you Custral, I'm pleased you found this helpful,

Kindest Regards,

Chris

After studying the curves produced in Midi Shape Shifter, I went back to the VPC EDITOR still dissatisfied. I have gone back to working on a tripoint curve i.e. velocity=[0, X, 127; 0, Y, 127].

The best I have managed so far is around the 90, 60 area - specifically 94, 62 i.e.,

velocity = [0, 94, 127; 0, 62, 127]

This can be used in the PianoTeq velocity pane, but I would recommend rather to use it in the VPC Editor.

I'd also be very grateful if anyone else wishes to suggest a curve for the VPC1 here in this thread. Mabry? Custral? Feline? et al

Kindest Regards,

Chris

Well I've got a worry. You can interpolate your MIDI output sure, but not the MIDI input, that's just integers 0..127. Result that seems to follow is that with some 94 input steps up to your 'mid' point (~3/4 of the input range), and 64 output steps to your 'mid' point (~1/2 the output range ....your attempt to interpolate any output can't work), what that curve produces is 1) inputs aliased such as needing 4 input steps to climb 3 output steps for the first range (zero to 'mid' point, where you'll have to hit harder than normal to make any difference); then 2) output steps aliased such as to climb four steps for every 3 input steps for the last range ('mid' point to endpoint....and you will struggle to keep output even, it'll want to lurch).

I'm tired so my numbers mightn't be utterly right, but the principle seems to be.

Thank you James, I did wonder that myself many times!!! But something convinces me that a tripoint curve is a simple answer that is actually possible and, as you may have guessed, I have modified it since. In fact, I've made use of my VRM Box (Focusrite) utilizing the virtual ADAMS reference monitors (which sound supurb by the way) to test out this one as it sounded beautiful through my monitors ( nEAR EXPERIENCE 05's) so I wanted to check through headphones. I have recorded a demo which I shall post soon.

Here are the coordinates:

0,0; 83,52; 127,127

Also, playing the above 'curve' having previously pushed the let-off arms out of play, I am able to get great control of pianissimos. And I have great control of fortes also. I shall demonstrate this in my mp3 as alluded to.

One last thing.

The key touch feel of the RM3 GRAND II is much improved since bypass of the let-off. I do miss let-off per se, but not the spongy-ness that came with it! If I had a gripe, which I don't, I would sincerely ask KAWAI to re-look at their let-off simulation and see if they could make it less so (spongy).

P.s. James, (and interested VPC1 owners) please try out the above curve and let me know your honest opinion(s). Obviously the presence of let-off simulation for you and others will mean you won't feel what I feel in the sense of the playing experience, but do give it a go anyway if you would as I would be interested to know if 1. It performs well, and 2. It performs well with let-off simulation?

Thank you,

Kindest Regards,

chris

Hi,

Thanks a lot for the Midi Shape shifter. It works great for velocity curves !

About the VPC1 let off simulation, I may add that, as a piano student, I find it much useful when practising, because the real let off on grand pianos is really not my cup of tea and I had many trouble to deal with it when practising only on my upright piano at home.
As a recent owner of a Vpc1 (wich, by the way is a real great keyboard !), with its heavy touch (much heavier than my upright) and the let off simulation, i am now able to work a lot better.

Having said that, I will moderate my opinion about the let off. I think that it works not so bad for slow pianissimos (in fact, it's almost as annoying as in a real piano, haha, but that's what i'm looking for...).

On the other hand, one would notice that on a real piano action, the let off is not effective when you play lightly and quickly without pressing the key all the way (the hammer is then throwed by its inertia and the let off is not engaged). That's obviously not the case with the VPC1 let off, wich is effective all the time and then slow the hammer course.
This means that quick and light pianissimos (as in Mozart or Chopin's works) are very difficult to realize with this keyboard... So I will sure disengage the VPC1 let off as soon as my warenty will be off (in 3 years !) but for the design of the future vpc2 (?), maybe Kawai engineers could have a look at this...

But I still think the the Vpc1 is a bloody good keyboard :-)

95, 57 is my latest and best offering to date

Kindest Regards,

Chris

Seems to be more presence body and  substance to chord grouping in your samples (as well as independent hands, but it's the chords that seem a newer thing). Is it so, or my imagining?

Could take eons, like natural selection. Lotta perms possible, but enough eons to weed out to the fittest candidate, and it'll win. Quicker to use an intelligent design approach, only are the same prefs everyone's?