Here's the deal, using such objects only make sense if converting to and back, right? I mean, at least for FFT that's the way it is, and I can't think of another example.
Oh, and let it be clear that only the phase value is coming out wrong (negative when it should be positive). And that is by using [cartopol~], of course.
Anyway, for any kind of spectral processing and stuff, it doesn't really matter if the phase is positive or negative. In other words, it sounds the same wether the phase value is theoretically correct or not. But this is only true if the backwards conversion with [poltocar~] is equally "incorrect".
This is an example where two wrongs make it right. And they were both wrong, but working. On the other hand, [cartopol] and [poltocar] where always correct, by the way.
So the issue was just that if you wanted plot or print the phase from [cartopol~] you get a wrong value. Not that much of a big deal. But it would hurt to make the audio objects rights like the data versions.
Now get ready for the ultimate issue. I had complained about all this way way long ago. And when Pd Extended 0.43 came out, apparently, something was done about it, but only one of the objects where changed (I guess [cartopol~]).
What doesn it mean? As of then, NONE of the spectral processing do work at all! Cause when [poltocar~] converts it back. It's just wrong.
In other words, it was better before, when two wrongs made it right.
How many now working patches will break when the object behaviour changes? A workaround could be to create new objects, e.g. cartopol2~ and poltocart2~ with the correct behaviour.
So, it was an honest concern. But the bottom line of what I'm telling you is: Nothing will break if this is corrected, and, actually, my patches using cartopol and poltocar are all broken in Pd Extended 0.43
So the only reasonable thing to do right now is just fix it as it is supposed to be. What do you say?
Can you follow all that I'm saying? I hope I was clear. Let me know if you need more details.
Thanks
2014-12-13 11:33 GMT-02:00 Fred Jan Kraan fjkraan@xs4all.nl:
Hi Alexandre,
Thanks for the clarification, I changed the overview. But this is a good example for the type of problem that maybe shouldn't be fixed in the current objects. How many now working patches will break when the object behaviour changes? A workaround could be to create new objects, e.g. cartopol2~ and poltocart2~ with the correct behaviour. A correct working cartopol~ and poltocart~ would then have to wait for a new library.
Some community input is welcome on this.
The plan now is to further investigate the items on the list, create bug-report tickets at https://sourceforge.net/p/pure-data/bugs, try to fix the problem, test it and create and insert code-patches at https://sourceforge.net/p/pure-data/patches/ and then have them apply to the code. The mechanism is just for this and it is operational. The list on my site should be seen as just a preliminary overview of tasks.
Greetings,
Fred Jan
awesome, though I wouldn't point th issue with cartopol~/poltocar~ as "accuracy", it's just positive when it should be negative and vice versa, but I assume you know what's going on.
cheers
2014-12-12 13:03 GMT-02:00 Jonathan Wilkes via Pd-list <pd-list@lists.iem.at mailto:pd-list@lists.iem.at>:
Great work! I'll take a look at them for Pd-l2ork. -Jonathan On Friday, December 12, 2014 8:20 AM, Fred Jan Kraan <fjkraan@xs4all.nl <mailto:fjkraan@xs4all.nl>> wrote: Hi all, The set of updated help-patches for the cyclone library is nowcomplete.
I'll check them at least once more time. For now they are at: http://fjkraan.home.xs4all.nl/digaud/puredata/cyclone/help-patches Meanwhile there is also a list of issues to fix:http://fjkraan.home.xs4all.nl/digaud/puredata/cyclone/cycloneToDo.html
If more people want to help, I could put the list in the wiki if I manage to figure out how. Fred Jan _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list _______________________________________________ Pd-list@lists.iem.at <mailto:Pd-list@lists.iem.at> mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list