Re: Playback Statistics component: v3.x foo_playcount
Reply #383 – 2015-05-03 05:05:29
There's a way around this, which most people wouldn't bother doing, but I use the method below for monitoring the "real-world performance" of the DAR Rating algorithm, which assumes a sane play count system. So, for anyone interested: I have a portable installation which uses foo_preview and only plays the first 65 secs of each track. In my main foobar2000 I use foo_history and that only adds tracks (to a playlist called "history") after they've been played for 60% or more. Every day or so, I drag the tracks from the history playlist to my portable installation (and clear the history playlist) and then they get played (the portable installation uses "null" for sound driver - thus plays in silence) and the portable foobar2000 closes automatically once all are played, and runs a batch file which closes my main foobar2000 (if it's open) and copies across the portable version of its \foobar2000\index-data\ to the main installation, thus overwriting the playcount data. This way any files which were counted as played after just 60 secs but were less than 60% completed do not see any change in their stats, as they were not played by the portable installation, since they never made it into the history playlist. As I said, I do this because I need to make sure DAR is giving sane results for a sane playcount system, since DAR can be tweaked to account for the weakness in foo_playcount, but IMO it shouldn't inherit the weakness of a component to in some ways cover up that very weakness. So if it's extremely important for you to have sane playcount data, start a new thread and I'll happily flesh how to do what I've noted above. But, I doubt most people are that bothered to go to such lengths. C.