>> At 06:04 PM 1/18/00 , Nicolas Leveille. wrote:
>> yep and their characteristics are not so attractive -from my point of
>> view ;) - I think the slope is 6db.
> the standard says nothing about the slope, it just says that the
>parameter you pass is the cutoff point at -6 dB. It is the same for all
>the others, or so I understand.
I think the issues here are:
-- For many musical applications, especially ones where the filter
parameters are moving at k-rate, the actual slope of the filter is
key to the sonic effect. When used in this way, the 6dB/octave slopes
that the lopass() of both saolc and sfront have will just sound
wimpy, while a nice 24dB/octave filter will sound strong.
-- Since the standard (and by your comment, I assume the conformance
tests as well) leaves the slope to the implementor, a sound designer
using filters in a way where the filter slope is sonically import
needs to either craft their own filter from scratch (either using
delay() opcodes or the biquad() or the fir/iir/firt/iirt filters),
or use instrs or opcodes written by others that are normative with respect
to cutoff slope.
I think this isn't really at odds with Giorgio's earlier post about
the conformance tests nailing down some aspects of lopass(), there
are also many applications where the slope of the filter isn't going
to make a noticable difference in the sound, and the convenience
and/or efficiency tradeoff of using the stock lopass() might make
it preferable to using a custom-designed filter,
This archive was generated by hypermail 2b29 : Mon Jan 28 2002 - 11:46:37 EST