cpd73 wrote:
> That config item is set to false by default for non-Linux, if it is not
> set in the config file. However, later you should have seen other lines
> 'Tracks have not been analysed with Essentia, attribute filtering
> disabled' or 'Will use Essentia attributes to filter tracks (highlevel)'
> -- do you see those? The code checks if Essentia is enabled then queries
> the DB to see if tracks have highlevel attributes. They would be in the
> section you have removed from your example output.
>
> Just noticed this check might not have been 100% so have made some
> slight changes.
>
There was nothing I snipped from the log that related to highlevel
attributes. I pulled your latest changes (9e0fda) and didn't see a
change in the startup logging. (Doesn't mean there wasn't any.)
cpd73 wrote:
> Your example 'FILTERED' lines are all Essentia low-level - BPM, Key,
> loundness.
You're correct. After I added an "essentia" stanza to my config.json
file to force high-level recognition (it wasn't enough to do so from the
Music Similarity plugin settings' page), then I did see a difference in
both the startup and in the attributes returned with each song.
cpd73 wrote:
> No. And, I'm not sure the current filtering of these makes that much
> sense.
I'm not sure exactly what you mean by 'current filtering of these,' but
I have some ideas. Some of us on these forums fondly remember Michael
Herger's deprecated SmartMix plugin. It was a fantastic tool that
allowed us to create mixes based on a variety of attributes just like
the ones in the music-similarity database. It had slider bars that
allowed us to set a range for various attributes and then build a
playlist from the tracks that met that criteria.
Imagine I want to query the music-similarity database for all songs with
these values and produce a playlistthat was SmartMix and that's what I
think the purpose of those values would or could be.
- happy: > 0.9
- aggressive: 0.4 - 0.6
- danceable: >0.75
- electronic: >0.8
- dark: >0.5
That criteria in my music collection finds 124 songs. I'd like to put
-that- on play.
With the voice attribute we have a way to create playlists based on
(perceived) male or female voices, toosomething SmartMix couldn't do
(that I recall).
I actually see these low-level and high-level attributes being the
greatest power of what you've done with the Similarity project. You've
opened up nearly 20 different ways to slice and access our music
collections beyond DTSM. And if we could hook into ratings from
TrackStat or RatingsLight... (well, a boy can dream)
cpd73 wrote:
> The similarity score is how similar a track is to the seed - a low value
> => high similarity. So, it would make no sense to have a similarity
> score to itself - it will always be 0.0, as its the same track.
>
> There is only one score.
Thank you for the clarification. Things are starting to make sense for
me.
------------------------------------------------------------------------
prabbit's Profile: http://forums.slimdevices.com/member.php?userid=11142
View this thread: http://forums.slimdevices.com/showthread.php?t=115609
_______________________________________________
plugins mailing list
plugins@lists.slimdevices.com
http://lists.slimdevices.com/mailman/listinfo/plugins
