(2024-02-01, 20:37)black_eagle Wrote: I use a 3rd party media manager. Usually MediaElch
That muddles the waters a bit, as MediaElch will output everything as "music videos". So there will be a 3 minutes clip and a 2h concert right next to each other. As you say I also much prefer to scan my concerts as movies and keep them on their own section, not mixed with MVs; the users of this forum often seems to want something else.
I'm gonna go on a verbal detour cause I love MV's and have 5 min to kill
. As an MV collector (and performance videos and many other things) I'm kind of hardcore about their metadata. In my opinion the best source for info for (true) music videos is Wikipedia. Like, on the right page one gets (example Dua Lipa - Houdini)
Quote:Music video
The music video for "Houdini" was directed by Manu Cossu and released alongside the song. It is set in a dance studio, where a redheaded Lipa performs a choreography with a group of redheaded male dancers. According to Vogue, the video appears to be a reference to the video for "Hung Up" (2005) by Madonna.
That tells us details about the actual item being scanned - the MV, which is what I want, and not about the song itself or other related items.
The problem is obvious - there is no scrapper for Wikipedia, and from my experiments that wiki Music Video section is all over the place (it can exist or not, it can be on various places on the page, it can exist multiple times if the song was covered by multiple artists and there are multiple music videos, etc). I'm no coder but it doesn't seem easy to just grab Beautiful Soup and some python code and boom! have a scrapper for the necessary part,
without fail.
There have been efforts (Media Companion) but from my trials it misses the target pretty often (grabbing the release date of the audio single, and that's not a given it's the exact same one for the MV, the producer of the song as the director of the MV - and that's wrong, the music label of the artist as the producing studio, etc). It's not easy.
And so, for when I'm making my own nfos for MVs, I use an over-engineered Autohotkey script (when one doesn't know how to code ends up writing complex scaffolding in Autohotkey) that semi-automatically gets that description from wikipedia. It gets to the page and literally pauses and waits for me to select what I want off the page and copy it to clipboard, then cleans it up, paste it in the right place in Notepad++ that has the NFO file open etc. It's a lot of insanity, and not 100% bulletproof. Maybe I'll make a post about it if I ever decide my clumsy efforts are worth sharing.
This would be much easier of course, if everybody would agree on one place to contribute to (preferably one with an API), with all the MV details - theaudiodb.com or imvdb.com (there used to be a scraper for this, right?) or whatever. But I don't see a lot of traction on that front. Maybe I'll learn Python in the next 5 years and do it myself. Maybe the MV collectors will unite. One can hope
.