SySPro Note Vel-control of some long articulations is very badly broken

[Macker]

Some Long and Legato articulation types suffer from a horribly big discontinuity in velocity-controlled volume-mapping for note velocity values below and above 82.

For example, my tests show that in Legato Regular vibrato, velocity-controlled volume leaps up by 11 LU (EBU Momentary) when note velocity changes from 81 to 82; then for velocity values up to 127 it continues upwards through its velocity-controlled volume map, relative to that egregious leap in volume level. This huge and painfully obvious discontinuity makes a nonsense of both live playing and note-programming, and is very tedious to correct by subsequent editing.

Long Molto vibrato behaves as it should (though on purely subjective listening, I'd have preferred its upper vel-controlled volume range to be just a tad louder, such that the entire loudness contour is more smooth - but that's probably splitting hairs).

I really don't want SysPro to have serious faults, any more than VSL does. I'm sorry but the painful fact is, it has.

[Macker]

Test Setup and Evidence of SysPro Note Vel-control behaving badly.

 

Synchron Player (v1.1.1954) AU directly plugged into LPX (v10.4.8).

VSL Factory preset: SySPro 1st Violins-14 Pro XF; Mixer preset: Default: VXF=Off, 0; Dynamic Range = Max.

MIDI sequence of 127 half notes of 1.75 secs length, all G3; note velocities successively increased by 1, such that the whole sequence plays the full range of note velocities increasing from 1 to 127; played back at 60 BPM.

Volume measured by Melda MLoudnessAnalyzer (v14.08), captured as EBU Momentary loudness vs Time graph.

Results of 2 test cases uploaded here: Long Molto Vib Vel-Ctrl (OK); and Leg Reg Vel-Control (Fault). The pictures both show the loudness (vertical axis) of all 127 test notes played back over a 256 second period (horizontal axis).

A 3rd pic is also uploaded, showing the MIDI note test sequence in LPX.

[Macker]

In case some are wondering what a good quality Note Velocity-controlled volume contour should look like, a fine and pretty much flawless example can be seen in my attachment No 1 below. This exemplary case is Appassionata Violins in VI Pro, played by the same MIDI note test sequence I've used above. I've also attached a pic of the VI Pro to show the test settings.

However, the same test applied to the Synchronized version of App Strings produces a very different result, as seen in my attachment No 3 below. 

Moreover, I'm also seeing similarly broken Note Vel-Controlled volume contours in SyzDimStrings. I don't want to post those results because, believe it or not, I'm really not into publishing negativity.