Right now, when using the modulation envelopes, there is no visual aid with respect to the sample it affects. It would thus be nice, if each sample’s preview would also show an overlay of the envelope (which one should be toggable).
That sounds like a nice edition.
And which envelope should be shown if volume, pitch and resonance/cutoff are all used?
Volume by default, right click on sample preview allows to chose others.
Or simply allow a legenda pallet to select what to show together, where every envelope has its own color.
Sure
The highest zoom level in the envelope is 3ms = 132 samples ( distance between 2 points ) …
The highest zoom level for an audiofile is 1 sample …= 0.002 ms
Which means , zoom level for the envelopes have to be increased .
Looks like 1msecs to me…
I’m not sure what else 1msecs modulations are good for besides fast smooth and clickless attacks… let alone modulation of times-ranges that are even less than that.
maybe you shouldn’t rely so much on visuals.;
There is always a transition time between the points …pulling out 1 point will not give you 1 ms …;.render it an you’ll notice …
No i don’t. It is perhaps noticable if you do your best but the majority of people that will be listening to your stuff are not even aware they can bother themselves with this level of detail. 
The time it takes to play a sample really depend on the pitch it is being played at, while the modulation is synced to wallclock/beats.
So the implementation couldn’t possibly be as elegant as you might imagine - either it would consist of multiple overlaid graphics, one for each sample that was playing or alternatively, rapidly switching between different overlays.
I don’t understand. The modulations are defined with respect to the root note, but they will scale when playing a different note, right?
No, they don’t. If an envelope has a length of 4 beats (beat-timing) or 4 seconds (wallclock-timing), then this is the duration of that envelope.
Playing at a different note will not change the time that the envelope takes. Wallclock timing is constant, no matter what, while beat-timing will change as you modify the tempo of your songs.
But if we had such a third timing option (note-timing ?), then I completely agree that it would make sense to have an overlaid graphic either in the waveform editor, the modulation page, or both.
That sounds strange. This would mean that using AHDSR envelopes would be useless for any pitched instrument…How do other samplers handle this?
Why should the timings of the envelope be changed, when you pitch the sample?
If you have a looping sample (pad sound or something) with a specific attack and release time, it would be strange, when you are playing high notes and the attack and release time will be getting shorter.
That depends on what kind of samples that you use, to be honest, if i would use an AHDSR on pitch or volume, i really want to have their pace steady rather than being matched to the sample frequency. This is frankly what would make envelopes and A(H)DSR routines useless because this would more or less simulate the behavior how a sample would sound up or downpitched if they were recorded with that very effect and that is actually what one doesn’t want to have:that it sounds like a cheap sample with included modulated effects (a violin player in real life does not tremble their vibrato faster or slower if they play a note higher or lower).
Well, for instance, you use an AHDSR volume envelope to shape how a sample sounds. That is, how fast the attack is, etc. If that changes entirely if you play any note that is not C-4, then what is the point?
If you use multiple layers and a different sample for each note that has the same dynamic properties except for the pitch rate, then this should not be a problem. If you use one sample where any form of modulation effect is noticeable, yes this is a problem if you pitch up or down. But i have personally not seen any big sample library that relied on just one sample.
I could imagine some samplers perhaps use some pitch stretching or crunching wizardry so that certain dynamic aspects may sound like the sample itself is naturally recorded that way where in fact they are not. And if this is pure on starting areas for attck and ending for decay or quick fade outs, it perhaps might not even be very noticeable.
The trick might be more prominent if you use larger samples that have a clear fingerprint (which should definitely change even if time-stretching or crunching is used)