Server Version#: 1.19.3.2740
Player Version#:3.0.2
I understand that Show Lyrics needs to be clicked for every track to discourage unattended wasteful use and so limit licensing fees from LyricFind but PLEASE make them persist between tracks for offline lyrics stored in .lrc/.txt files, without a need to navigate and re-click every track. Surely a work around would be to auto-show offline lyrics unless n/a in which case just show “Artist - Title” at top with a LyricFind logo button just below, that could be clicked to get LyricFind Lyrics or say none available if none are.
Surely this anomaly, ongoing for years, needs to be addressed properly with a different licensing arrangement or LyricFind just dropped completely rather than crippling the lyrics feature. Due to this and the Plex Alexa skill rarely working properly, Plex has been abandoned by my family in favour of Amazon Prime music, due to ease of use, which is a shame after all the effort put in, creating a perfect music library, including writing Plex Plug-Ins. I’m hoping Plex Amp will lure them back in.
It would also be nice if the visualisation could be displayed behind the lyrics with a fade setting in options 0-100%, obviously the font would need outlining to not disappear. Also tilting the phone into landscape (or clicking a fullscreen icon) could make Lyrics/Visualisation go full screen. Oh and a timer option (end of track or after n minutes)
I decided about 8+ years ago that the only way to have sane metadata was to go down the route of having all metadata in .nfo/.jpg/tags/.lrc/.txt. That way any errors or custom metadata can be easily corrected/added once then works forever. So I wrote an app that bulk gets the metadata via API’s from themoviedb, thetvdb, theaudiodb, musicbrainz etc and saves to .nfo/.jpg/.lrc files in a kodi compatible format. I used the existing plex .nfo reader plug-ins and wrote 2 more for music & musicvideo.
The result is perfect data that can be read and re-read by Plex, Kodi, Emby etc without ever needing to go online again so is a lot more efficient, especially if say database needs rebuilding. I originally tried embedded metadata (still used for music) BUT it’s not as easy to edit/fix as a text based xml .nfo file where one .nfo format works for ALL different media that all use different ways of tagging stuff.
So I wonder if you have ever considered separating out the Plex scanners from the importers in this manner. If the importer only ever read in/monitored .nfo files then keyed off name for art/media it could be optimised to be very fast, as many assumptions about the data being correct could be made (as you’ve produced it in first place). I do still embed tags in music/audio files the data could just as easily come from the album.nfo files. It would simplify the code and reduce problems a lot IMHO by a separation of concerns.
I even considered writing my own importer to just directly inject nfo into the Plex DB just to see how fast it would be. I’ve done this already in past for RomsCollectionBrowser (went from 6 months to 20 minutes to import) + NextPVR in Kodi. Reverse engineering Plex DB didn’t look like too much fun though. Maybe one day!
