Closed fietew closed 5 years ago
bug fix seems ok to me, just corrected the labeling of axis in the test script.
The test runs fine o my machine.
Is this the desired output:
Fig1
Fig 2
Fig 3
Is this the desired output:
It seems, there is a problem regarding the Octave output. The greens lines in Fig1 and Fig3 (upper left) should be horizontal. I will have a look.
The behaviour is, again, related to a bug in zp2sos in Octave (see https://savannah.gnu.org/bugs/?51936), where single or non-existing zeros are handled incorrect. This was already mentioned in #169 . I used a similar workaround for this.
Looks good now, here are the new figures I'm getting by running the test:
Fig 1
Fig 2
Fig 3
Is it ok to squash the commits during merge?
Alternatively I would propose to condense them to two, one for the fix and one for adding testing functions.
For some filter orders (even, but not divisible by four), there was a wrong sign for LR Allpass Filters. This PR fix this and adds a testing function.