Hi, this is the question to the developers. Can you be so kind to confirm if Pianoteq actually use MIDI controller message #88 to generate Hi-Res velocity input data? Since it is not so obvious to detect it by hearing the official info will be very useful. Thank you.
It does.
Ok, I did some synthetic tests in a DAW trying to catch the influence of MIDI controller #88. I was playing with some extreme Volume - Dynamics - Velocity curve settings plus actually tweaking the #88 MIDI controller. My tests show that due to the level of randomness and/or complexity that is in the nature of Pianoteq algorithms #88 controller have absolutely zero practical influence on the generated sound.
At the extreme experimental "microscope" settings I was able to notice:
a) The notes of the exact same velocity are uneven in volume
b) +1 / -1 changes of standard velocity is clearly audible though extremely inconsistent.
c) Controller #88 presence and changes was absolutely unnoticeable.
The tests was done inside Ableton Live 9 with the VST instance of the Pianoteq. Effects - Off. There are possible some factors that I could not take into account though I hope to publish some video report for this forum users consideration in the near future.
===================
Added: the test is need to be redone since I did not know that there is a setting in the MIDI preferences of Pianoteq that I did not enable (Dialect / Hi-Res #88).
AKM: I'm not at all surprised - that's what I've always suspected.
AKM: I'm not at all surprised - that's what I've always suspected.
Just wanted to note that by no means I was trying to state that Pianoteq is a kind of "bad" software, exactly opposite actually But, according to my test, it is kind of incorrect to say that Pianoteq is a Hi-Res capable software. Again, it need to be double checked, there may be some influences that I did not considered.
While I don't want to argue, I really don't think a Casio can transmit the nuance that some claim.
And me think that a human can't perceive differences between a normal res and high res midi recording. 128 steps is enough already.
please note that cc88 is only taken into account as hi-res midi velocity if you select "hi-res cc88" in the "dialect" menu of the "options / midi" dialog, otherwise it is simply discarded
please note that cc88 is only taken into account as hi-res midi velocity if you select "hi-res cc88" in the "dialect" menu of the "options / midi" dialog, otherwise it is simply discarded
Thank you so very much, sure I did not know that.
While I don't want to argue, I really don't think a Casio can transmit the nuance that some claim.
You actually don't need to think, you can just SEE it in the MIDI monitor in the Pianoteq preferences MIDI page that Casio do send CC#88 Hi-Res messages. The only theoretical thing it can be is that they actually fake it and it is just sending out some random mess, is that what you saying?
Again, thank you so very much, Julien. I will re-check it all again later and post the results.
Ok, new tests are ready with Hi-Res switch On.
In less than a half cases I was able to hear the expected crescendo from, say, velocity 11 - 11.5 - 12. In more than half cases the 11.5 was as loud as 12 or 11 was as loud as 11.5 or any other combination, though some level of #88 influence was obvious. So my conclusion is quite same as after the first incorrect test: since Pianoteq have some level of natural randomness in notes volume Hi-Res implementation is practically quite useless, though it is great to have it anyway just in case.
Next, just some notes. When playing from Casio PX-150 keyboard I can notice that in about 1 of 25 cases Pianoteq discard the #88 Hi-Res messages showing just an integer velocity number. While doing the synthetic tests in a DAW I notice that the timing for the #88 changes occur is very important. I could get 100% recognition of #88 when controller changes happen at exact same time as the note plays. If you move controller change message slightly ahead or after(!) the note the Hi-Res information is added in much less percent of events. The behaviour I'd expect is that since you sent the #88 message at any given time once it will be kept unchanged until the new one arrived - but it does not work this way. I guess it is because of the internal specific of MIDI protocol realization, maybe some data flow in serial or some kind of parallel way, or some mono-poly behaviour, idk. Though it is too specific topic.
Added: not sure who read this all, seems it is all for my own enjoyment. Was able to test the incoming messages from Casio using some NI Reaktor debug tools in parallel with Pianoteq. There is some unpredicted behaviour happens in some cases from seems the Casio PX-150 side, maybe because of 3-d sensor sometimes, maybe anything else but sometimes it looks like it just does not send #88 data at all (maybe it is meant to be a zero value from it) or send it kind of not exactly in time (about 1 millisecond before the note data, so Pianoteq truncate it) ... or whatever... I'm going to stop here and just play some music.
Eigenharps are capable of sending very high data output and so i always have Hi-Res selected, it does feel more "in control".
Eigenharps are capable of sending very high data output and so i always have Hi-Res selected, it does feel more "in control".
Not knowing much about these, it seems their keys might be more pressure sensitive than velocity sensitive.
It's both.