Help for Tune Smithy
From Tune Smithy
Pitch artefacts at the start of the note =
This happens with some soft synths. If you apply an "instant" pitch bend immediately before a note, the note doesn't respond to it instantly, instead it responds with a glide to the desired pitch from the usual equal tempered pitch. I can see no reason programmatically why a note shouldn't be able to start exactly at the pitch set by the most recent pitch bend, so perhaps this is intentional - maybe as a way to respond more smoothly to the pitch bend wheel? Anyone reading this know more about why it happens?
The solution that works best is to apply pitch bends to all the channels well in advance as soon as the scale changes. This completely eliminates the problem if you have enough channels available to send notes for each pitch bend on a different midi out channel.
The problem will still occur for large scales and non octave scales, or if several instruments are played simultaneously with varied pan position (left / right) or other effects. For instance if there are more than 15 pitch bends required for a scale, then as the music goes on, from time to time one of the existing channels will need to be retuned.
When this happens, the only solution is to add delays before the new pitch bends from time to time. There's an opton to add these delays - see Out | Options | Out (menu) | Midi Out / Save Timing (Ctrl + 58) in FTS. If you need this just for Quicktime there is an opton to just switch on the option for QT in FTS.
With this option in place, then you will hear occasional pauses as you play in the more demanding tunings - also when you play some of the more demanding fractal tunes. But that may be preferrable to the pitch glides, and in many non demanding pieces they never happen.
Pitch bend artefacts at the end of a note =
Gliding resonances after the end of a note played on harp or other instruments that continue to resonate after the note is switched off
This happens if FTS needs to use the same channel for a new pitch bend immediately after the note is switched off. Although the note is switched off, instruments like harp may continue to resonate for a while after the note stops. These resonance type effects will get pitch bent by the next pitch bend in that channel.
You may also notice that long high pitched notes tend to go slightly flat they fade away at the very end of the note, and low pitched ones tend to go slightly sharp
This is something you can notice in twelve equal as much as for any other tuning - it is nothing to do with the pitch bend retuning.
The reason for this is that we tend to hear high notes of constant frequency as sharper as they get louder, and flatter as they get quieter. Low notes work the other way. In real life, players compensate for it automatically by playnig notes a little sharp as they fade out, and low notes a little flat as they fade out.
I did an FFT analysis of the samples for the 'cello arco / vibrato in the Dan Dean's solo strings and found that the player compensates by raising the c6 by about 27 cents as it fades almost to silence, and by lowering the c2 by about 20 cents. That is for the fade out right at the end when the note is nearly silent.
Giga and other samplers have to fade out the notes you play when they are switched off, maybe before the end of the recorded sample, as an abrupt cut off would sound strange. But it isn't able to raise the pitch to compensate in exactly the same way a player of the instrument would do.
Though 20 cents seems like quite a large pitch change of nearly a fifth of a semitone - actually this is quite a subtle effect, because it only happens when the note is very very quiet. The note becomes almost inaudible before you need a pitch compensation of as much as 20 cents. It is most noticeable if the instrument fades out slowly, e.g. for plucked sounds.