[PD] hip~ nyquist gain bug test patch [Was: arraysize]

[Claude Heiland-Allen]

On 03/10/12 15:22, Claude Heiland-Allen wrote:
> On 02/10/12 16:15, Miller Puckette wrote:
>> 1. there's a bug in hip~ - its DC gain is slightly (and possibly
>> considerably)
>> greater than 1.
>
> Did you mean nyquist-frequency rather than DC here?

See attached test patch that waits 10 seconds for settling to steady 
state (extreme overkill no doubt), at 48k sample rate the nyquist gain 
for hip~ goes up dramatically as you increase the cutoff...

> IIRC DC gain for hip~ should be 0 or very close to it

verified.

> and nyquist-frequency gain 1.

here is the bug, these get much louder than the -1 1 -1 1 input

hip10:
-1.0007 1.0007  -1.0007 1.0007  -1.0007 1.0007  -1.0007 1.0007
hip100:
-1.0066 1.0066  -1.0066 1.0066  -1.0066 1.0066  -1.0066 1.0066
hip1000:
-1.07   1.07    -1.07   1.07    -1.07   1.07    -1.07   1.07
hip10000:
-2      2       -2      2       -2      2       -2      2


Claude
-- 
http://mathr.co.uk
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: pd-hip-nyquist-offset-test.pd
URL: <http://lists.puredata.info/pipermail/pd-list/attachments/20121005/726e261c/attachment.asc>