While I’m on the subject of soundfont editing, I discovered something interesting last night while working on my current track. But first, I need to give a quick follow-up to my last post.
That is, while I successfully edited soundfonts to respond to breathe control (CC#2), I have been unable to find any VSTi synths that will PLAY BACK CC#2! I tried every soundfont player in the KVR Audio list, and everyone one of them ignored CC#2. Well, VST Synthfont didn’t ignore it – it played it back as CC#1 data, producing very bizzare, warbly noises.
The only way to get CC#2 playback was to use the soundcard, which does not allow off-line rendering. That opened up another whole can of worms that resulted in me going back to my old DAW to render the tracks. Nuff’ said, but if you have or know of a soundfont playback VST that will respond to all CC’s programmed into the soundfont itself, please let me know.
So, enough about the past, now to move forward. First of all, what the heck is a release sample?
Well, when a typical musical sample (single note type) is played back with a keyboard, the volume of the sound is often shaped by an ADSR envelope. This means that when the key is released, all that happens is that the volume is brought to 0 at whatever rate the ADSR envelope specifies. Careful listening to real instruments will reveal that there is often a very unique sound accompanying the end of the note – the scrape of the bow, the drop of the damper, that sort of thing.
Until last night, I didn’t thing you could do that with soundfonts, but have learned how by looking at a soundfont that does just that. In this case, it is a harpsichord soundfont that has the dropping of the plectrum at the release of the key. Here’s how it is done:
Each sampled note has an accompanying release sample assigned to the same note range. However, the release sample has a short period of silence at the beginning (a few milliseconds). The attack and decay portion of the ADSR envelope is set to 0, and this silence is then looped at the sustain portion of the sound. This way, the actual sample doesn’t play until the release portion of the envelope. Brilliant, I think.
However, there are some drawbacks that almost eliminate it’s usability. The primary one concerns sustaining instruments and legato passages, which is where release samples are most useful. The problem is that the release sample will play for every note in a legato passage, when reallyonly the last note should have one. There may be a way to control this, but I haven’t found it yet.
The other problem with this soundfont specifically, is that the release samples of all those plectrums dropping was adding an unacceptable amount of noise to the track. So, after learning how to duplicate the process, I promptly went in and deleted all the release samples.
Oh well, live and learn.
And somebody send me a vst that will play back CC#2 for soundfonts!