yaksplat 66 Posted January 13 Posted January 13 I don't want my music mixed with videos. Is there a preferred folder structure for Music videos? Or a preferred naming convention? <Artist> - <title>.<ext> I tested with: Faith No More - Epic (Official Music Video).mp4 I left it in the root of the MusicVideo directory which is used in a library that is specified as Music Videos. The video was imported and called Epic. No Artist. So i identified it with the IMVDB number. Still no artist. Then I hit the IMVDB API Get http://imvdb.com/api/v1/search/videos?q=faith no more epic with this response { "total_results": 93641, "current_page": 1, "per_page": 25, "total_pages": 3746, "results": [ { "id": 131667161071, "production_status": "r", "song_title": "Epic", "song_slug": "epic", "url": "https://imvdb.com/video/faith-no-more/epic", "multiple_versions": false, "version_name": null, "version_number": 1, "is_imvdb_pick": false, "aspect_ratio": null, "year": 1990, "verified_credits": false, "artists": [ { "name": "Faith No More", "slug": "faith-no-more", "url": "https://imvdb.com/n/faith-no-more" } ], "image": { "o": "https://s3.amazonaws.com/images.imvdb.com/video/131667161071-faith-no-more-epic_music_video_ov.jpg?v=2", "l": "https://s3.amazonaws.com/images.imvdb.com/video/131667161071-faith-no-more-epic_music_video_lv.jpg?v=2", "b": "https://s3.amazonaws.com/images.imvdb.com/video/131667161071-faith-no-more-epic_music_video_bv.jpg?v=2", "t": "https://s3.amazonaws.com/images.imvdb.com/video/131667161071-faith-no-more-epic_music_video_tv.jpg?v=2", "s": "https://s3.amazonaws.com/images.imvdb.com/video/131667161071-faith-no-more-epic_music_video_sv.jpg?v=2" } }, The artist is available in the first result but emby didn't connect the dots. What do i need to do to make this work? I've tried a folder structures of: <artist>/<music video name>/<music video file> <music video file name>/<music video file> with permutations of filenames using just the <music video name>, or <artist> - <music video name>, but nothing has resulted in correctly identified videos.
arrbee99 1815 Posted January 13 Posted January 13 Personally I just have Music Videos\Artist\Video name. Can't say it works very well so there's a lot of manual stuff. I've only got about 100 favourite (individual not album) videos though. Maybe you could +1 this, should you so desire...
Lazarus_Long 80 Posted January 13 Posted January 13 For music videos, I used to previously use Media Companion to generate NFO files (and they provided a lot of nicer information, including summaries from Wikipedia). Then Emby would pull the NFO data in and it looked nice. Unfortunately Media Companion's scraping for music videos has been borked for a few months now, waiting on the dev to get time to fix it...
yaksplat 66 Posted January 14 Author Posted January 14 I have thousands of music videos to bring in and I'd rather do it right the first time than find out i have to remove and readd them. I have them all in a folder and I'm waiting to write a script to sort them in the right manner.
yaksplat 66 Posted January 19 Author Posted January 19 Here the result of 311, All Mixed up. It found info, populated it right, but then decided that it was a movie, despite being in a Music Video Library It found the correct IMVDB Id.
yaksplat 66 Posted January 19 Author Posted January 19 Bad religion Stranger than fiction. IMVDB id found. Nothing populated
Luke 42077 Posted January 20 Posted January 20 The artist will need to be filled in. That either comes from embedded metadata in the same fashion as music, or by using the metadata editor first.
yaksplat 66 Posted January 20 Author Posted January 20 But why? That information comes back on the IMVDB api call. Shouldn't everything available be mapped? Otherwise, what good is even having the IMVDB id?
Luke 42077 Posted January 20 Posted January 20 Because the rate of false matches is too high when searching by file name alone without knowing the artist in advance. Additionally it’s very common for the spelling of artists on imvdb to be different from music databases such as musicbrainz and this can lead to a duplicate artist being added to your server database.
yaksplat 66 Posted January 20 Author Posted January 20 That's why I was asking about the convention of naming. If there is a method, whether it's naming a video by artist - song or files named by song in a folder of artist, there's a method to grab the correct information. As it is, right now, the results are a mess. As far as multiple artists in the db, that's a case of crap in, crap out. The software can't be responsible for incorrectly named files. If there's no result from IVMDB, then it should be up to the user to populate the data. But it sounds like it'll be up to me to write a script for generating nfo files.
Solution yaksplat 66 Posted January 20 Author Solution Posted January 20 (edited) This script will look at the IMVDB id in the nfo files and confirm that it matches the file title. if it does not match the file name, the correct id is added. This will also update artist, title, sorttitle and year. if the item is not in '<artist> - <song title>' format, it will be written to an error file for manual intervention. To run: python imvdb_fill_nfo_sharable.py "path to music video folder" it works quite well imvdb_fill_nfo_sharable.py Edited January 20 by yaksplat 1
brothom 177 Posted January 21 Posted January 21 (edited) Hooking into this functionality, I gave IMVDB a try as well. It seems that whenever an IMVDB identification is found, the artist becomes hidden/disappears from the overview: The artist field is also empty, but this item seems to contain that info on the IMVDB side: https://imvdb.com/video/o-town/all-or-nothing. Scrolling through this list is now a pain because there's no way to check which artist the video is associated to unless the image contains a band member (I'm not memorizing every band's members lol) or by viewing the original filename which is in this case "O-Town - All or Nothing.mp4" Any idea what might be going on here? Edited January 21 by brothom 1
yaksplat 66 Posted January 21 Author Posted January 21 5 hours ago, brothom said: Hooking into this functionality, I gave IMVDB a try as well. It seems that whenever an IMVDB identification is found, the artist becomes hidden/disappears from the overview: The artist field is also empty, but this item seems to contain that info on the IMVDB side: https://imvdb.com/video/o-town/all-or-nothing. Scrolling through this list is now a pain because there's no way to check which artist the video is associated to unless the image contains a band member (I'm not memorizing every band's members lol) or by viewing the original filename which is in this case "O-Town - All or Nothing.mp4" Any idea what might be going on here? yeah, this was my exact concern. If you have nfo files enabled, my script will help that. I have an updated version that allows manual intervention for all of the files with issues. I think I'll create a new thread for it and drop a link here.
yaksplat 66 Posted January 21 Author Posted January 21 My library looks like this now, with years, artists and titles. 1
Luke 42077 Posted January 21 Posted January 21 5 hours ago, brothom said: Hooking into this functionality, I gave IMVDB a try as well. It seems that whenever an IMVDB identification is found, the artist becomes hidden/disappears from the overview: The artist field is also empty, but this item seems to contain that info on the IMVDB side: https://imvdb.com/video/o-town/all-or-nothing. Scrolling through this list is now a pain because there's no way to check which artist the video is associated to unless the image contains a band member (I'm not memorizing every band's members lol) or by viewing the original filename which is in this case "O-Town - All or Nothing.mp4" Any idea what might be going on here? Hi, The artist will need to be filled in. That either comes from embedded metadata in the same fashion as music, or by using the metadata editor first.
Lazarus_Long 80 Posted January 21 Posted January 21 (edited) 14 minutes ago, Luke said: Hi, The artist will need to be filled in. That either comes from embedded metadata in the same fashion as music, or by using the metadata editor first. Why can't it pull it from IMVDB? The data is already there. That's sort of the whole point...otherwise there's not much point to even pulling from IMVDb Edited January 21 by Lazarus_Long
brothom 177 Posted January 21 Posted January 21 (edited) 1 hour ago, Luke said: Hi, The artist will need to be filled in. That either comes from embedded metadata in the same fashion as music, or by using the metadata editor first. When the item is NOT identified, Emby uses the correct artist. When it IS IDENTIFIED, it doesn't show or apply the artist tag at all anymore. My formatting is "artist - title.ext" Everything that has been identified now needs to be edited manually? Do .mp4/m4a files even háve an "artist" meta tag? Edited January 21 by brothom
yaksplat 66 Posted January 21 Author Posted January 21 1 minute ago, brothom said: When the item is NOT identified, Emby uses the correct artist. When it IS IDENTIFIED, it doesn't show or apply the artist tag at all anymore. My formatting is "artist - title.ext" Everything that has been identified now needs to be edited manually? Do .mp4/m4a files even háve an "artist" meta tag? just use nfo files
brothom 177 Posted January 21 Posted January 21 1 minute ago, yaksplat said: just use nfo files I have millions of files, I don't want a million extra nfo files polluting my drives. I'm already considering myself insane for formatting each and every entry as described and even separating subs. It's also beside the point: the data is available in IMVDB, in the exact item Emby has associated to each entry even. 1
yaksplat 66 Posted January 21 Author Posted January 21 I know what you mean. With the jpg and nfo files, I suddenly have an extra 120k files in my music video directory. I don't see any interest from the emby staff to change how this works, despite it being blatantly wrong as it is. With nfo files, i can easily manipulate them with a script to do anything that i want. They're easy to find and easy to delete. Emby sees the change and handles it appropriately. I'd rather manipulate the database directly, but that doesn't seem to be as open.
Lazarus_Long 80 Posted January 21 Posted January 21 (edited) I'm retesting using Media Companion to create NFO files...it only runs under Windows and until yesterday there was a bug in it, but I pulled a beta/test version the dev released for me and it seems to be working properly again. It can pull from Wikipedia (what I tend to use) and also IMVDB. This is an older video, not sure how well it works with newer content: Edited January 21 by Lazarus_Long
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now