snip
To me the most "obvious" improvement would be to drop the toggle for time
stretch (despite RAM cost, this is the sort of feature that should be
always on
I disagree - with the current algoithm the RAM cost is significant, and
the impact timestretch has on sound quality (e.g. mono output and plenty
of artifacts) mean that in many cases (i.e. non-speech) it's better to use
the old pitch-only option.
Isn't there an improved algorithm in a task on the tracker? I believe the
last comment I saw for it was that someone couldn't get it to work. If this
is still the case, has anyone tried contacting the original author asking
about how it should be used or if an even more up to date version exists?
