Jump to content

IMDB matching wierdness


Go to solution Solved by pwhodges,

Recommended Posts

Posted

Can anyone explain to me what is going on here please?

In my Movies library folder, I have all the files named with their IMDB IDs. So, for example, the file for the movie "Inception" is named tt1375666.mp4
99% of the time this works perfectly and everything gets automatically matched correctly with zero human intervention required.

But on the very odd occasion there will be a title where it just randomly mismatches.

Two examples from yesterday...

"Challenger Disaster: Lost Tapes" named tt5372500.mp4 got incorrectly partially mismatched to "Romantic Story" (tt0372500)
"Titanic: The Final Secret" named tt2978760.mp4 got incorrectly mismatched to "Finding Dory" (tt2277860)

See screenshots of how they appear in the Web UI...

screenshot2.png

screenshot1.png

Posted

Hello davepusey,

** This is an auto reply **

Please wait for someone from staff support or our members to reply to you.

It's recommended to provide more info, as it explain in this thread:


Thank you.

Emby Team

  • Solution
pwhodges
Posted

Emby does not use IMDB, 'cos $$$.

When an IMDB id is provided, it will be used as one of the keys in the lookup on TMDB or TVDB - these providers usually have the IMDB id in their database, but just sometimes it can be either missing or wrong.  Providing the id of the database which you have configured as the first lookup (TMDB by default, for films) will be appropriate when the IMDB id doesn't work (or indeed at any time).

Paul

Posted

So by default it's not loading the metadata directly from IMDB then? I thought that was the entire point of using the tt number.

Is there any way to make it do this?

pwhodges
Posted

No - it's too expensive.

Paul
 

Posted

Movie Naming  I use this.

Folder 10 Cloverfield Lane (2016) {TmdbId-333371}

Movie : 10 Cloverfield Lane (2016) {TmdbId-333371}.mkv

Posted

So lets look at your naming of your files 😉 

Posted

I'm trying to avoid putting anything in the filename besides the ID number.

This is how it looks on my test instance...

image.png.830f46bcacd1346ec480233de91bbbd4.png

The bottom two are how I normally do it, and those work perfectly... (1st and 4th items below)

image.png.6f91762741b34a1f63bbcbac8e9fa374.png

Posted

Not sure if ever will have success with that.

Read Movie Naming and understand why.

Posted

I don't see why it needs the name and year as well. The ID should be all the information it needs to find it.

Posted

Success?!

Don't even need my custom image anymore (only made one because IMDB had no image)

image.png.59810b9ab67d58c4e9f63913396e10a0.png

image.png.ef1a7d255b7db4834a271c2438388b88.png

Posted

Jep there are workarounds, but that might break down the line, as Emby evolves.

Posted

Now to find out if this will work with TV Series too.

What I have been doing is naming the first folder with the IMDB tt number. So, for example "Stargate SG-1" was named tt0118480 and then had the seasons and episodes below that with numbering that matched the "official" order on TVDB.

One of the reasons I chose to use IMDB IDs originally was so that I didn't have to get identification data from multiple sources for my own custom database that tracks where all my media came from (locations, boxsets, disks, ripping and encoding profiles, etc.)

Posted
1 hour ago, davepusey said:

I'm trying to avoid putting anything in the filename besides the ID number

Hi.  Why?

Doesn't that make it pretty hard to manage when just looking at the file system?

  • Agree 1
Posted

I'm trying to avoid having extra infomation in the name that will potentially confuse the matching process. This is something I've encountered before and going ID-only was the solution that I found worked best, with the occasion oddities like the ones above. I chose to use IMDB IDs because I thought that was the defacto standard that every system would support.

Posted

But it seems I'm going to have to migrate my naming system over to TMDB and TVDB IDs.

At least with the TV Series I can include the unique "slug" too for a more human readable reference...

image.png.aa4286fd155d6afca764aef67afe689b.png

Just need to see if TMDB have the same concept in their database too for the Movies.

pwhodges
Posted
3 hours ago, davepusey said:

I don't see why it needs the name and year as well. The ID should be all the information it needs to find it.

But it needs to know what the number is!  Some films have names which are just numbers (2001, anyone?)  I use [tmdbid=1234567] but various other ways work.

Paul

Posted
4 hours ago, davepusey said:

I don't see why it needs the name and year as well.

Because if you put the tvdb id in the filename, but Tmdb is your highest priority fetcher, then you are relying on Tmdb having that tvdb id in their database. If they don't, then it will end up searching with whatever is left to work with in the filename.

Posted (edited)

I checked. I have TMDB for Movies library type, and TVDB for TV Series library type.

I'm going to re-number everything based on this new information and see how that goes.

Edited by davepusey
Posted

The other reason would be that if you ever change your metadata fetcher priority.

pwhodges
Posted

Not to mention you have no idea what's on your disk without looking up the "name" of every file in turn!

Paul

Posted

I already have a custom database that has ID to Name mappings, as part of my record keeping.

Posted

You can obviously do whatever is best for you but I guess we are struggling with why you are making it so hard on yourself?  Using plain names (and IDs embedded if you wish) would work much better and you wouldn't have to maintain a database to know what you are looking at on your disc.

22 hours ago, davepusey said:

I'm trying to avoid having extra infomation in the name that will potentially confuse the matching process

The extra information is actually what would help the matching process.  Mis-matches (when using normal names) are the exception, not the rule.  At least 90% of items should match fine.  Then you can address the minority of ones that don't.

Posted
Quote

The extra information is actually what would help the matching process.

But the point is my workflow has already identified the correct movie/series to match it with, and I've given a unique indentifier for it. If Emby is unable to match with just that then either I've given an incorrect ID or something is wrong. I do not want Emby trying to figure out the movie/series matching by itself, I want it to match explicitly on the identifier I am providing ONLY.

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 account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...