sampled versus modeled piano app played on a semi-weighted keyboard

[LeesKeys]

I usually use a Studio Logic 73 with Pianoteq IOS. I recently purchased a Nektar GXP88 (semi-weighted key bed). I've tweaked the velocity curve using the GXP options as well as the curve withing Pianoteq. The result is not bad, certainly not like a fully weighted keyboard, but it's OK. One thing that continues to happen is that even though I've tweaked the curve for a faster keyboard, I still get occasional velocity spikes on individual notes, even though I THINK I'm playing all notes at the same velocity (of course I'm not). This type of thing has happened on other brands of semi-weighted keyboard I've previously owned.

 

Yesterday I downloaded Genuine Sounds sampled CFX piano. The quality of the sound is very good, but more to my point, when playing it on the GXP (and using the same velocity curve as with Pianoteq, I did not experience any random velocity spikes on individual notes. I wonder about this- is a sampled piano more "forgiving" than a physically modeled piano when played on a semi-weighted keyboard? Thoughts?

[Reezekeys]

Here's my theory and I'm sticking to it: a sampled piano has velocity layers where a given range of velocities can address the same layer. Your keyboard may indeed be "spiking" values but as long as the higher-value velocity is still within the same layer, the difference in sound may not be as pronounced as a modeled piano. Of course I have no idea how many layers the CFX piano uses or whether they simply switch between them or use crossfading, etc. If you put a midi monitor in your chain and make a note of the velocity when you hear a spike, that may help diagnose your issue. One of the settings on my Roland A800 Pro keyboard lets me assign any fixed velocity value to the keys - so I can play a note and slowly dial up the velocity value to listen for when the sound changes. If the Nektar does this, that might be something to do to help you get some answers as well.

 

 

[hrestov]

I don’t know. Also for me, with casio cts 500, that have a lightweight action, I prefer the Cfx of just piano Than Pianoteq. 

[ElmerJFudd]

Record your performance into a midi track in your DAW.  You be able to see the velocity variations and spiking from your controller clearly.  Use the same performance to send to one or the other plug-in.  Then you’ll be certain the same velocity data is delivering different sounds from these instruments.  It’s all in how the instrument was programmed to respond and how you have the velocity curves setup in their software.   Theoretically modeling software should have more gradual transitions in timbre than velocity layers in sampling.  With sampled instrument the solution is to add more velocity layers and carefully program their transition points and relative volume.   

[Jinkings]

Lowering Pianoteq’s dynamic fader might help some