Skip to main content
Topic: foo_beatit (Read 3161 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

foo_beatit

Implement's fraganator's BPM algorithm, which is multithreaded to run on multiple files at once, depending on amount of CPU cores.
Also implements two others, which are also multithreaded.

http://mudlord.info/crap/audio/foo_beatit.dll

foo_beatit

Reply #1
fcuk yeah !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!    Thank you !
However I found a problem - when I choose some track on list of calculated BPMs and I choose "double" or "halve bpm" window closes and analysis results are lost. Additionally resizing of window with analysis results causes that buttons "double bpm" and "halve bpm" doesn't move with window edge - they float over track list and simply don't work...

foo_beatit

Reply #2
Very fast
But haven't got good results with default settings. With FFT size 1024 results are much better.

Btw. i copied the settings to foo_bpm and it was nearly as fast as foo_beatit on my system...
But results are very different (almost wrong with foo_bpm then).
Although fraganator's algorithm was chosen in foo_beatit.
Shouldn't both don't give the same results then?

foo_beatit

Reply #3
fcuk yeah !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!    Thank you !
However I found a problem - when I choose some track on list of calculated BPMs and I choose "double" or "halve bpm" window closes and analysis results are lost. Additionally resizing of window with analysis results causes that buttons "double bpm" and "halve bpm" doesn't move with window edge - they float over track list and simply don't work...


Fixed.

foo_beatit

Reply #4
Very fast
But haven't got good results with default settings. With FFT size 1024 results are much better.

Btw. i copied the settings to foo_bpm and it was nearly as fast as foo_beatit on my system...
But results are very different (almost wrong with foo_bpm then).
Although fraganator's algorithm was chosen in foo_beatit.
Shouldn't both don't give the same results then?


Not sure, could be due to the different FFT library.

foo_beatit

Reply #5
As far as wrong results belong to the problem of doubled or halved bmp there was once made (if i remember correct) in the foo_bpm thread: defining for a certain genre a certain range in preferences. That means, if you have a hip-hop track with 190 BPM you can be sure that it is 95 BPM instead, if you have a trance track of 73 BPM it should be 146 then.

foo_beatit

Reply #6
Well, reported bug was solved only partially. Buttons don't float anymore and window isn't closed suddenly but halving and doubling BPM does not work anyway. When I press "Double BPM" or "Halve BPM", then calculated value is copied from "BPM" column to "Status" column and tags are stored immediately - for all analyzed files, but unmodified - they are never doubled or halved, they are written just like they were calculated.
But still speed is briliant

foo_beatit

Reply #7
Sounds like this needs more alpha testing, sorry for rlsing it so early.

edit: there, known bugs fixed. Not sure about the FFT issue though.

foo_beatit

Reply #8
How is this different or better than foo_bpm?

foo_beatit

Reply #9
Multithreaded scanning, thats it.

foo_beatit

Reply #10
Hello!  Sorry for bothering, but the link in the starting post is dead.  Is it possible to re-upload it?


Re: foo_beatit

Reply #12
Thanks, Bro!  Five years have passed... :)

 

 
SimplePortal 1.0.0 RC1 © 2008-2020