I think there’s a bit of an issue with the philosophy behind this.
Whilst a section of quavers sounds great at say 80%, this does not hold true at all for semibreves, and they drop out way too early.
I get that there should be some gap between normal notes, but it needs to be more an absolute gap rather than a percentage gap. I was wondering why my tied semibreves (2 bars) were cutting out half way through the second bar.
So I turned it up to 100%, which works fine for horn pads, but no good then for faster bowed passages in the violins.
It’s almost like this could be set based on a technique text, e.g. marcato vs staccatissimo etc etc. in addition to the dots (staccati) and lines (tenuti)
What we have done in other situations is to enforce a maximum amount of reduction that one of these options can allow, so that we don’t allow a gap of more than a certain amount of real time based on the percentage.
Of course, the idea is that this is adjustable on a per-playing technique basis, via the mechanism of VST Expression Maps. This may not work correctly in all cases at the moment, but you will be able to specify changes in dynamic and note length etc. for individual playing techniques in the expression maps.
Yes, I was very pleasantly surprised to see you’d set up a framework for rule-based control like that.
Don’t get me wrong, it’s very cool. Like you say there needs to be a little more finesse in some parts (e.g. cap the amount of absolute time that is trimmed from notes).
I suspect that what you are hearing could be due to the slow attack of the string patches rather than the note playback duration. We expect that this is something that will improve in future versions when we provide automation (so that we can automatically reduce the attack for faster passages).
Also the method we use isn’t purely a percentage - there is a maximum reduction so that you don’t end up with very long notes being shortened to much. This rule did change slightly in 1.1 so it’s possible that it introduced a bug. It certainly shouldn’t reduce more than a couple of beats.
Yeah like I said it trimmed half a bar of the tail end of my tied semibreves.
I’d suggest max trim off the end should never be more than a semiquaver for a normal note. What you’re emulating for strings is a bow change. For winds a breath.
When playing the music you don’t think in terms of percentages but absolute length of a note based on markings and period style. The percentage is an engineering approach but with the rule framework you could easily change the way note length is defined. You’ve already done most of the work by recognising all these things in the first place.