Expression versus channel volume.
Sorry for the newb question, but I am, in fact, a newb when it comes to midi strings. I've been experimenting lately with the Expression control lane in Pro Tools (with NI Session Strings) and, as far as I can tell, the Expression feature is just another volume control. I'm curious to know if you midi veterans out there bother with the Expression controls at all or if you just add swells and crescendos to your synth strings by automating the Volume lane.
(Somebody please let me know if the PT forum would be a better location for this question.)
Hi,
The simple answer is 'it depends upon the sample player'.
Back in the older days of hardware sound modules (and sound cards) it was pretty well universal (not totally) that 'note on velocity' controlled not only the loudness of a note but also the timbre. In this case channel volume (cc#7) was used to set the relative levels between the instruments in a mix. That is, get
the balance correct, much as one would do with a hardware mixer and the main channel faders for each audio source. Expression would then be used to adjust the level, applying crescendo/diminuendo during the playing of a note.
This procedure is still primarily true for all sound fonts that I've used to date.
However, in virtual instruments, typically VSTi, this may be true or, in many I've come across, the overall level is still set with channel volume, but note on velocity is just used to define how strongly the note is 'attacked'. In many virtual instrument libraries expression (cc#11) is used to determine note volume for non percussive sounds, i.e. strings (except pizzicato), brass, winds, etc. Note velocity is still used for e.g. piano, harp, timpani, etc. With some libraries breath control and modulation can also be used to control loudness.
When expression is used in these libraries it is typical that they not only control loudness but also the timbre of the note will change subtly as the control is raised.
So my suggestion would be to use expression for just that purpose, to produce the crescendo and diminuendo in note volume and set channel volume (starting with the default of 100) to balance the strings with the other instruments in the mix. The key thing is that the maximum combined amplitude should remain below 0dB FSD. I usually aim for -3dB, which gives a little headroom should conversion to mp3 be needed (sometimes the conversion process can push the maximum amplitude up into clipping if the level is set too high).
And how does one know? The only recourse I've found is to refer to the manual or ask a user of the library you're using.
I have no experience with NI strings but I suspect it uses expression in the way I've described.
It's generally considered a better procedure anyway to use expression for just that purpose.
My own use is primarily of the Garritan libraries, but also of several of the East West products too.
With Garritan the main expression control is cc#1 or 2 or 11, they all work the same way.
Just some jottings off the top of the head.
Regards,
John.
