On 220928-13:48+0700, Georgy Salnikov wrote:
On Tue, 27 Sep 2022, Miroslav Rovis wrote:
Here's how my tentative with the shaker2.xml and the video went: https://croatiafidelis.hr/foss/cap/cap-220925-cinelerra/pg-220926-motion.php The files necessary are there (as I got them in email that missed to be sent to Cin ML).
Why it did not produce the expected result? (I haven't examined the XML-after yet, but it's available on the page.)
Miroslav, you have tested everything right. Your result is the same as mine.
OK, the result is in reality not perfect. By finer tuning the plugin parameters and keyframes placement achieving better results is in principle possible. But this particular example was only a testcase for me - to verify that the algorithm works. Getting visually perfect results was not my goal. _______________________________________________________________________________
Georgy Salnikov NMR Group Novosibirsk Institute of Organic Chemistry Lavrentjeva, 9, 630090 Novosibirsk, Russia Phone +7-383-3307864 Email [email protected] _______________________________________________________________________________
Well, that clears some of my doubts regarding motion plugin. Thanks! And there is one good thing this demostrates, if your result and my result are identical: consistency of the program, It may be long to go to get reproducibility [*]. I hope Cinelerra will be reproducible one day. Regards, kind developer! ---- [*] such as e.g. Debian is striving for: https://wiki.debian.org/ReproducibleBuilds . I for myself can state that when I compiled Getmail, I did get the exact same binaries out as the then current getmail package in Debian, as I reported in https://github.com/getmail6/getmail6/issues/123 However, long to go for Debian to be really reproducible as well. -- Miroslav Rovis Zagreb, Croatia https://www.CroatiaFidelis.hr my PGP-key: https://www.croatiafidelis.hr/FCF13245ED247DCE443855B7EA9884884FBAF0AE.asc