know to avoid delay
timings between 50-120 mS on vocals, for example, as this is
approximately the length of a short syllable or at least roughly
the length of time it takes to recognise a segment of vocal sound
as linguistically meaningful
Although this can be an effect in its own right - last week we were
trying out LADSPA Tape Delay, trying to get the sound of a Watkins
Copicat (explanatory picture follows for the post-analogue
generation):
http://www.chrisguitars.com/wem-copy.jpg
Using a short delay and about 2 taps, we managed to get the sound we
were after. Maybe we need a user-contributed bank of presets for all
the LADSPA plugins - it's much easier to get started from a good
basic effect.
Cheers
Daniel