Returning to the core issue of whether to update the standard to define a
consistent method of determining the length of delay lines for all core
opcodes when a non-integer length is requested...
The calculation of delay line lengths will always(?) occur at i-rate, ie.
not very often, so I'm not convinced that the performance hit will be so
great as to justify not using Floor().
The best of both worlds (efficiency and consistency) would be to update all
of the core opcodes to specify round-to-nearest to determine the length of
delay lines.
Best,
Ross.
This archive was generated by hypermail 2b29 : Mon Jan 28 2002 - 12:03:58 EST