TMDB as Metadata Source

I am using Trakt list import, and since Trakt is using TMDB for metadata bad things happen.
The show will still be imported if found on TVDB, but then it makes Trakt link that is actually Trakt search of TVDBID on Trakt and if there is no TVDBID you get dead Trakt link for the show you imported from Trakt (lol).
The only solution so far is to edit TMDB database to have TVDB ID in its data.
So i pull list from Trakt, Sonar finds it on TVDB, then links it to Trakt that is dependent on TMDB database having TVDBID.

I have 186 shows in all in Sonarr and Trakt and TMDB has never missed one while TVDB has missed at least 15%.

Bigger the library, bigger the problem.
:smile:
Capture

And that means what? Nothing?

Going to have to disagree with this. TMDB is wrong quite often as well, very wrong. In fact, TVDB and TMDB are very much in sync with each other. The TVDB doesn’t align with IMDB, nor does TMDB. The only one that does is OMDB. This option would be GREATLY preferred over TMDB or TVDB. Or perhaps even IMDB directly.

Want a prime example? Highlander: The Series. OMDB and IMDB both have the correct, proper viewing order. It also matches the order on the DVD’s. Airing order is irrelevant with older series as they were frequently aired out of order. That’s not a problem anymore these days, but both TMDB and TVDB have blatantly refused to correct this inaccuracy (among many others) because all they care about is airing order.

The ORIGINAL air order is what matters and that IS what both TVDB and TMDB use. DVD order is irrelevant. As for IMDB most of us know that’s another joke like TVDB is.

The new episodes of Doctor Who will appear as a separate series on the BBC website, that has been confirmed multiple times and as has been mentioned, the recent specials have already appeared as part of that new series. TVDB used to consider the broadcaster website as the ultimate authority when it came to series/episode details, I have no idea if that has changed since they closed their user forums but if it is still the case they ought to have changed the Doctor Who details by now. (In actuality, the ultimate authority always turned out to be “Whatever we (the TVDB) feel like going with,” so if they haven’t changed it by now I don’t hold out much hope for them bringing it in line with the BBC website.)

1 Like

There is nothing wrong with my previous post and you damn well know it.

For people reading this topic and looking for an answer to the question asked, here you go:

I made an account just to beat a dead horse, fight me.

I have two Sonarr instances running on my server. One for TV shows, and one for anime.

I could not be any happier with the TV instance. Truly, Sonarr, alongside the other Arrs, is a work of art, and you are doing God’s work.

HOWEVER,

I will not commission the anime instance, because TVDB completely screws up anime unapologetically. TVDB is not for anime, period. Full stop, hands down. I will not let Sonarr touch my collection until I can add AniDB as the scraper and disable TVDB. TVDB is hot garbage when presented with anime, especially really old anime. Remember, this is nothing against the Arrs team, just a choice to lock something behind a menu that seems like it was always meant to be user changeable.

If I have to click a hundred “Don’t do this, you have been warned” buttons and check boxes, so be it…but I beg you with tears in my eyes…please, let us choose what scraper/metadata aggregator we want to use. Until then, I have to do my anime by hand and no donations will be given.

But… but it’s not just “locked behind a menu” or something like “just change the link from where the info comes”. You make it sound like the option is there, it just isn’t. Full stop. That’s not how coding works, that’s not how any of this works.

The devs have chosen to not write all the lines of code that would give you that option. They have not chosen to withhold the (non-existing!) option from you…

2 Likes

I just want to know how to switch to TMDB, my data is wrong by at least 15 points

Not just animations, at least 15 pieces of data are wrong

You can’t. Asked and answered a couple replies up.

Please add TMDB + AniDB as options, TVDB is subpar to my current usage of Sonarr and the current setup of TVDB is a headache try and find a workaround.

Adding onto this because I absolutely need new metadata sources to either be baked in or add plugin support for it. I don’t care how they do it as long as it’s done. Anime with TVDB is a nightmare and I just straight up cannot use sonarr to watch Justice League Unlimited at all. Complete mess

as much as we would all appreciate having the option, the amount of code changes and new code needed to handle multiple sources would be massive. sonarr fixes and updates would effectively stagnate while that happened, or if you still want those then you have to accept an even longer window.

even if they managed to get that sorted out and give us multiple metadata source options the next, and bigger, part of the issue is the amount of data a sonarr user would need to pull from that source to update our local data, and the sheer number of users there are that would be doing it.

any metadata source that becomes available is going to get hammered by us - they are not going to like it - they are going to throttle or block us, or perhaps they will make the api a paid offering so they can buy more bandwidth to handle the extra traffic.

theres currently a skyhook server acting as a sort of cache proxy that makes the api calls to tvdb for the data we want and caches the results (to return for future queries from other users) so we dont turn tvdb into a bigger pile of crap than it already is (or force them to make it a paid api).

so if you dont want the users killing the sources the only real option is that you need a skyhook variant for each metadata source that is made available so we dont kill the source.

it would also mean that only the sonarr devs could create new sources, they cant be plugin based that anyone can write.

forgot - the very first thing needed from a new potential metadata source is an api that works and has all the required information. if it doesnt it cant be a source. do not even think about suggesting scraping.

XEM also relies on TVDB, so there would be a loss of functionality.

This thread really should be locked, ‘same here’ comments do not help, people just need to read.

2 Likes

The problem is people will also open new threads, even when this one exists, I’m inclined to not close this and encourage people to mute it and avoid the replies.

TMDB has free APIs to retrieve multi-language title variants as we see in Radarr, offers significantly better solution for non-english contents (especially anime). Users can apply for API keys then enable TMDB metadata retrieving if we don’t want to flood their servers with a proxy. I’m just wondering if we can use this as a supplementary source for multilingual search querys rather than relying on Xem, which actually pulls manually curated scene mappings all at once.

For example I’m using a dedicated Sonarr service to search for Chinese-subed Anime content, currently most of the titles would only have one alias in Romanji, which already fails to match many releases with Hiragana or Chinese translation titles. There are even cases where the Romanji versions differ. For instance, today I encountered a title called “Shuumatsu Train Doko e Iku” in Xem, but most Chinese indexers refer to it as “Shumatsu Torein Dokoe Iku”. This should be fixed if the client is able to search with Chinese, Japanese or any other configured indexer-specific language preferences like Radarr do.

This might include a new preferred language setting for indexers (with frontend), a new SceneMappingProvider, and few logic changes in ReleaseSearchService. In fact, it was only this afternoon that I noticed the issue with Japanese search queries, then cloned and debuged the repo for the first time. I understand that the development of this probably requires much effort and and I have great respect for the devotionof the developers. I’m not asking anything to be implemented; just a suggestion from a personal user perspective of how TMDB APIs can help us.

1 Like