Plugin: TheIntroDB – Emby Plugin

[sh0rty 749]

I am not the creator, but I recently found this Plugin now available for Emby (for people that don't want to spend time on extracting Intro, Credits etc. on their own systems). 
The DB relies on community feedback. For sure it's not as perfect like e.g. @yocker's tesseract extraction since there are lots of different versions and releases of movies/TV shows, but dev is very responsive and the Emby plugin simply works since last version. Perhaps interesting for a few users here like me.

As far as I can see, the plugin sets up an own db in emby data folder and injects identified markers as chapters into Emby.

https://theintrodb.org/

https://github.com/TheIntroDB/emby-plugin

Edited May 27 by sh0rty

[Pasithea 17]

Hey, I'm Pas, the creator of TheIntroDB! I'm glad to see the community already adopting the plugins and sharing them before me lol.

Currently, TheIntroDB has 84,789 public segments across 2,061 TV shows and 282 movies! And because the times are submitted and verified by users, the segments are much more reliable automated identification. Plus, it not only supports intros and credits, but also recaps and previews!

I might not be very active on the forum here, but I'm responsive on Discord: https://discord.gg/ZbkSHAwWPg
You can find more info and help contribute here! https://theintrodb.org 

[seanbuff 1378]

Interesting project @Pasithea

Are the submissions based purely on timestamps or chromaprints for more accuracy?

If only the former, how are markers handled across different sources of the same episode?

[Pasithea 17]

3 hours ago, seanbuff said:

Interesting project @Pasithea

Are the submissions based purely on timestamps or chromaprints for more accuracy?

If only the former, how are markers handled across different sources of the same episode?

All segments are submitted by users and contain only the timings. Different media releases (e.g. theatrical, extended cut, etc.) are identified based on the total duration of the media.

https://api.theintrodb.org/v3/media?tmdb_id=1084242&duration_ms=7200000 

Edited May 28 by Pasithea

[yocker 1684]

14 hours ago, Pasithea said:

Hey, I'm Pas, the creator of TheIntroDB! I'm glad to see the community already adopting the plugins and sharing them before me lol.

Currently, TheIntroDB has 84,789 public segments across 2,061 TV shows and 282 movies! And because the times are submitted and verified by users, the segments are much more reliable automated identification. Plus, it not only supports intros and credits, but also recaps and previews!

I might not be very active on the forum here, but I'm responsive on Discord: https://discord.gg/ZbkSHAwWPg
You can find more info and help contribute here! https://theintrodb.org 

You trying to make my plugin obsolete?  
Please do! Anything that makes time marks easier and better is absolutely awesome!
Good work!

Edited May 28 by yocker

[Amanade3 61]

for @Pasithea cool plugin but problem is in recap section if the recap starting time is at the begining it just ignores to insert the recap chapter and only the recap end chapter is injected would have been nice if it contains Recap (TheIntroDB) at 00:00:00 . 

[Pasithea 17]

11 minutes ago, Amanade3 said:

for @Pasithea cool plugin but problem is in recap section if the recap starting time is at the begining it just ignores to insert the recap chapter and only the recap end chapter is injected would have been nice if it contains Recap (TheIntroDB) at 00:00:00 . 

I can add that in the next update

[Babatom 16]

@Pasithea

Thanks for the plugin. Unless I've missed something, does it only scan at a specific time? It would be great to have an option that scans new media files straight away when you add them.

[Pasithea 17]

4 hours ago, Babatom said:

@Pasithea

Thanks for the plugin. Unless I've missed something, does it only scan at a specific time? It would be great to have an option that scans new media files straight away when you add them.

I can look into this

[Pasithea 17]

On 5/28/2026 at 12:23 PM, Amanade3 said:

for @Pasithea cool plugin but problem is in recap section if the recap starting time is at the begining it just ignores to insert the recap chapter and only the recap end chapter is injected would have been nice if it contains Recap (TheIntroDB) at 00:00:00 . 

I've added this in v1.0.2.3.
https://github.com/TheIntroDB/emby-plugin/releases/tag/v1.0.2.3

Hopefully this is the last "testing" release before I can submit to the official Emby plugin catalog. If people could test and give me more feedback before I do, please do!

[soderlund 167]

Great plugin! Keep it up!

I have been testing the plugin and have a few things that i think would be a good improvement. I just at this moment finished scanning my library of 15000+ movies and 6000+ tv-shows (took about 20 hours). 

• Add ability to only scan selected shows. For example, in my big library, it takes a long time to scan it all, but if I had the ability to only do a scan for Game of thrones and a few others that I'm currently watching it would be nice. So i don't have to wait all that time for THE ONE I really want. 
• Ability to select what library to run. Same reason as above. Maybe i only want to run this on some of my libraries
• Ability to test API key 
• Can we limit the log output (without loosing information)? I got like 1,2 millions rows of log from the plugin.. 
• Look into rate limiting, see below

Since the log is pretty big, I asked Claude to summarize and here is the result:

Library	168,648 items
Speed	~3 items/second
Movie %	0%
TV-Show %	9,4% (13,429 of 143,571)
Errors	1,200+ of API rate-limit

 

Is it possible to add some settings the user can change, like "extend the scan to include results with similar timings" (and ofc it could lead to false positives) 

Don't get me wrong, this plugin is freaking awesome, but I did expect some better results. I guess this is the limitations with user reported databases. 

When I do get results, they seem to be spot on! 

 

Edited June 1 by soderlund

[Pasithea 17]

1 hour ago, soderlund said:

• Add ability to only scan selected shows. For example, in my big library, it takes a long time to scan it all, but if I had the ability to only do a scan for Game of thrones and a few others that I'm currently watching it would be nice. So i don't have to wait all that time for THE ONE I really want. 
• Ability to select what library to run. Same reason as above. Maybe i only want to run this on some of my libraries
• Ability to test API key 
• Can we limit the log output (without loosing information)? I got like 1,2 millions rows of log from the plugin.. 

These all are great ideas!. I've created issues on GitHub for all of them: https://github.com/TheIntroDB/emby-plugin/issues
As for the rate limiting, I can't exactly remove it, however your idea for library or show specific filtering would help. 

I'll start work on adding these features, but it might take me some time. Currently I'm the only active developer for 6 different TIDB plugins (If anyone has dev experience, PRs welcome!) and the project is only a couple months old, but growing fast!

[soderlund 167]

6 minutes ago, Pasithea said:

As for the rate limiting, I can't exactly remove it, however your idea for library or show specific filtering would help. 

Yeah I understand that, but maybe look at the timings between calls if that could have an impact? I don't mind it takes a bit longer to scan if it means that it doesn't skip any episodes. 

I would love to help, but i need to put my limited time on my own plugin and the family... But i would like to help you test out new releases or any alpha/beta testing! 

[Pasithea 17]

@soderlundI've added some of the features you requested in the latest update!

• Pre-filter submissions before media segment scan
• Add per file/folder selection for scan
• Add API key test

https://github.com/TheIntroDB/emby-plugin/releases/tag/v1.0.3.0

[sydlexius 322]

@PasitheaPerhaps support for GraphQL to allow for bulk mutations and queries might lighten the load?

[Pasithea 17]

I'll consider it. It's not a priority right now.

[Pasithea 17]

TheIntroDB plugin is now in the official Emby plugin catalog!
You can find it under the general section

[AngelSing 73]

Hi,

I would like to report an issue that I was able to reproduce with the TheIntroDB plugin on Emby Server 4.10.0.16 beta running on Synology DSM 7.2.

At first, I thought the issue was related to Emby's Restart option or to the Synology package itself, because Emby remained stopped and did not start back up automatically. However, after more testing, I found the actual way to reproduce the problem.

The issue happens when I add new content to a TV library, for example a new season or new episodes, and then Emby starts scanning/processing that content. At that point, TheIntroDB is automatically triggered for the new items through item_added / item_updated events. After that, Emby stops completely. It does not restart; the package remains stopped in Synology DSM and I have to start it manually from Package Center.

In the log embyserver-63918344396.txt, the issue can be seen while new episodes from the series From are being scanned. For example:

TheIntroDB: On-demand segment fetch completed (item_added) for El Corazón es un Cazador Solitario: 2 segments, markers applied
TheIntroDB: On-demand segment fetch triggered (item_updated) for Los Mejores Planes (Episode)
TheIntroDB: GetMediaSegmentsAsync called for ItemId=706fcf54-7b5f-b1e0-145d-8831a245ebe8
TheIntroDB: Episode: Name=Los Mejores Planes, Series=From, S4E7
TheIntroDB: Fetching from TheIntroDB API
TheIntroDB: On-demand segment fetch triggered (item_added) for Los Mejores Planes (Episode)

After that activity, Emby stopped. The next log, embyserver (1).txt, is already the next Emby startup at 15:39:56, after I manually started the package again.

To confirm the cause, I uninstalled only TheIntroDB and repeated the same test: I added new content to the TV library and ran the scan. This time Emby did not shut down, and the content was added correctly. So I can confirm that, at least in my environment, the issue was related to TheIntroDB.

There is also an earlier log where this error appears during Emby shutdown:

System.Threading.SynchronizationLockException: The lock is being disposed while still being used
at System.Threading.ReaderWriterLockSlim.Dispose(Boolean disposing)
at TheIntroDB.Data.TheIntroDbSegmentRepository.Dispose()
at TheIntroDB.EntryPoints.TheIntroDbChapterMarkerPersistenceEntryPoint.Dispose()
at Emby.Server.Implementations.ApplicationHost.Dispose(Boolean dispose)

Looking at the plugin code, this seems to match the behavior observed. TheIntroDbChapterMarkerPersistenceEntryPoint subscribes to ItemAdded and ItemUpdated, starts OnDemandFetchAsync, and during Dispose() it calls _segmentRepository.Dispose(). In turn, TheIntroDbSegmentRepository.Dispose() calls _lock.Dispose().

My impression is that when new content is added and TheIntroDB starts processing several episodes through on-demand operations, something may remain active using the lock or waiting for it. Then, when Emby tries to shut down or the process fails, the lock is disposed while it is still in use.

Attached logs:

• embyserver-63918344396.txt: shows TheIntroDB processing the newly added episodes shortly before Emby stops.

• embyserver (1).txt: shows the later Emby startup after I manually started the package again.

• I can also attach the earlier log where the TheIntroDB SynchronizationLockException appears during shutdown.

After uninstalling TheIntroDB, the issue disappeared in my tests.

embyserver-63918344396.txt embyserver (1).txt embyserver-63918336666.txt embyserver.txt

Edited 1 hour ago by AngelSing

[Pasithea 17]

Thanks @AngelSingfor reporting this! Can you try v1.0.7.1 and see if it’s been fixed?

[yocker 1684]

Just a heads up.
To help the developer in the best way possible it's usually a good thing to shut down Emby, start it back up (to reset logs and start a new one), reproduce the problem and then send the log right away.
This lessens the amount of irrelevant stuff in the logs that the dev has to look through.

10MB of logs is A LOT to go through!  

[AngelSing 73]

9 minutes ago, Pasithea said:

Thanks @AngelSingfor reporting this! Can you try v1.0.7.1 and see if it’s been fixed?

Thanks. I’ll test it as soon as it becomes available. I just checked the plugin from Emby, but right now the latest version available to install is still 1.0.7.0, and 1.0.7.1 is not showing up yet.

[AngelSing 73]

7 minutes ago, yocker said:

Just a heads up.
To help the developer in the best way possible it's usually a good thing to shut down Emby, start it back up (to reset logs and start a new one), reproduce the problem and then send the log right away.
This lessens the amount of irrelevant stuff in the logs that the dev has to look through.

10MB of logs is A LOT to go through!  

Thanks for the advice, you are right.

In this case, at first I did not know what was causing the issue and I also could not reproduce it. The failure happened unexpectedly, so the only log I had available was that large one. That is why I attached it, because if I restarted or cleared the logs, I could have lost the only evidence where the cause of the problem might appear.

Later, almost by coincidence, I found a way to reproduce it: it happened when adding new content to a TV library while TheIntroDB was installed.

For future testing, now that I know how to reproduce it, I will do exactly that: restart Emby, generate a clean log, reproduce the problem, and send the log right away.

[Pasithea 17]

@AngelSingoop my fault, forgot to push it from GitHub to the Emby library. V1.0.7.1 should be available now. Lmk if it works now!

[AngelSing 73]

12 minutes ago, Pasithea said:

@AngelSingoop my fault, forgot to push it from GitHub to the Emby library. V1.0.7.1 should be available now. Lmk if it works now!

Thanks. I was able to install version 1.0.7.1 and test it. It is working correctly now, and Emby no longer shuts down when adding new content.

Thank you very much for the quick fix.

[Pasithea 17]

Awesome!