Is the existing codeclib mdct being used now (which is already heavily
optimised)? Or is optimisation lower down the priority list and/or mdct not
the bottleneck?

Reply via email to