Plugin: EmbyCredits, detect end credits and add auto skip.

[Babatom 34]

I installed it via the plugin store, but I'm still on v2.65.0; there's no way to trigger it manually, is there?

[yocker 1821]

1 hour ago, Babatom said:

I installed it via the plugin store, but I'm still on v2.65.0; there's no way to trigger it manually, is there?

Choose the latest version in the drop down and then press install.

[Babatom 34]

Oh man, thanks a lot

 

[DarWun 369]

@yockerI have a (possibly stupid) question about your tesseract container (yock1/embycreditocr). Most of the time Tesseract's status in Task Manager on my Synology is "Sleeping". But on occasion, it will go active and start utilizing the CPU even though the EmbyCredits plugin is not running. CPU usage is roughly 26% when it is active at those times. Is that expected behavior? If it is, what is Tesseract actually doing? I'm assuming it is doing maintenance (i.e. clean-up of remnant files). But I thought I would check just in case there is something else going on.

Edited July 2 by DarWun

[yocker 1821]

9 minutes ago, DarWun said:

@yockerI have a (possibly stupid) question about your tesseract container (yock1/embycreditocr). Most of the time Tesseract's status in Task Manager on my Synology is "Sleeping". But on occasion, it will go active and start utilizing the CPU even though the EmbyCredits plugin is not running. CPU usage is roughly 26% when it is active at those times. Is that expected behavior? If it is, what is Tesseract actually doing? I'm assuming it is doing maintenance (i.e. clean-up of remnant files). But I thought I would check just in case there is something else going on.

Sadly don't know, sorry.

Tesseract and PaddleOCR are docker containers i have modified slightly to clean up the images after them self when done with them, in fact you could use the originals with the plugin if you wanted.
You can find Tesseract at: tesseract-ocr/tesseract: Tesseract Open Source OCR Engine (main repository)
They are open source, highly known and used so i'm 99.9% sure it's nothing nefarious or people would have caught it by now.

[yocker 1821]

New beta version (v2.6.9.4) in the catalog.

Added TheIntroDB lookups.
This, when enabled, looks for credit markers on TheIntroDB.org before running any detections. If found it uses those and skips the detection run on that/those files.
Please note, the amount of credit markers on TheIntroDB is rather limited atm. but should get better as time goes on.

I'm considering adding an upload function but not sure detections are precise enough for the site, so any uploads should at least be manually inspected before any uploads happen.. Don't know if it's worth adding.

Edited July 16 by yocker

[ebr 16648]

17 hours ago, yocker said:

I'm considering adding an upload function but not sure detections are precise enough for the site, so any uploads should at least be manually inspected before any uploads happen.. Don't know if it's worth adding.

Just my 2c - you gotta be really careful with this because that is basically going to be viewed as sending personal information about the media off of the device to ....wherever.   Which is just a really sticky thing these days.

[yocker 1821]

12 minutes ago, ebr said:

Just my 2c - you gotta be really careful with this because that is basically going to be viewed as sending personal information about the media off of the device to ....wherever.   Which is just a really sticky thing these days.

Agree, it's going to be a manually thing if anything plus I'm going to ask the people behind theintrodb before even considering it.

[yocker 1821]

Small update (v2.9.6.7)

Added ability to use locally installed Tesseract instead of being Docker only.

Tesseract can be found at: https://github.com/tesseract-ocr/tesseract

[GMEMedia 2]

@yockerI'm currently using EmbyCredits and https://github.com/TheIntroDB/emby-plugin. I see your latest changes include TheIntroDB lookups.

Does this mean I no longer need the other plugin?

[yocker 1821]

57 minutes ago, GMEMedia said:

@yockerI'm currently using EmbyCredits and https://github.com/TheIntroDB/emby-plugin. I see your latest changes include TheIntroDB lookups.

Does this mean I no longer need the other plugin?

When enabled in EmbyCredits it will only look for end credits for any file that it runs detection on.
If found it will use the time stamps from TheIntroDB and skip detection.

For anything more you will need the other plugin.

[yocker 1821]

New Beta version in the catalog (v2.7.0.0) for those who dare.

Added:

1) Support for locally installed Tesseract OCR, a locally installed Tesseract does not require docker.
2) Ability to run multiple PaddleOCR sessions at once, this way you can run detections on multiple files at once with PaddleOCR. Please note: This requires a powerful and/or dedicated computer to not create bottlenecks.
3) Improved chromaprint detection slightly.

Bug fixes:

1) Episode comparison when running chromaprint detection would compare episodes against them self.
2) Clean up would in rare cases fail to clean up causing a small memory leak.
3) Other fixes and optimizations not noticeable to users.
 

[yocker 1821]

New version (v2.7.0.5) available in the catalog.

Added:

1) Support for locally installed Tesseract OCR, a locally installed Tesseract does not require docker.
2) Ability to run multiple PaddleOCR sessions at once, this way you can run detections on multiple files at once with PaddleOCR. Please note: This requires a powerful and/or dedicated computer to not create bottlenecks.
3) Improved chromaprint detection slightly.
4) Added TesseractOCR/PaddleOCR readiness indicator.

Bug fixes:

1) Episode comparison when running chromaprint detection would compare episodes against them self.
2) Clean up would in rare cases fail to clean up causing a small memory leak.
3) Every 200th frame would be discarded and not be run through detection.
4) Other fixes and optimizations not noticeable to users.
5) Strm files are now automatically skipped (Thank you Serph91P on Github).

Edited Sunday at 07:30 PM by yocker

[frankythedog 6]

Hello,

The last version working for me is the 2.4.7.0

 

Each episode is scanned in second, and give error.

Like if smb path were not recognized anymore.

I checked on a library without smb:// but with real path, and it works.

 

[2026-08-20 22:36:47.368] [WARN] [OCR Detection] FFmpeg coarse scan exited with code 1 - hardware acceleration (vaapi) may have failed
[2026-08-20 22:36:47.368] [DEBUG] [OCR Detection] FFmpeg pre-input args: -hwaccel vaapi -vaapi_device /dev/dri/renderD128 -hwaccel_output_format vaapi
[2026-08-20 22:36:47.401] [DEBUG] [OCR Detection] FFmpeg stderr: ffmpeg version 5.1-emby_2023_06_25_p6 Copyright (c) 2000-2022 the FFmpeg developers and softworkz for Emby LLC
  built with gcc 13.4.0 (crosstool-NG 1.28.0)
Execution Date: 2026-08-20 22:36:47
smb://192.168.1.11/NAS001_6/Séries/Loki/Loki - 01x03 - FHD.mkv: Protocol not found
Did you mean file:smb://192.168.1.11/NAS001_6/Séries/Loki/Loki - 01x03 - FHD.mkv?
EXIT

I attached a Dry Run with Debug log

 

Do you have  an idea ?

 

Thanks in advance.

credits-detection-debug-2026-08-20T20-36-50.log

[yocker 1821]

37 minutes ago, frankythedog said:

Hello,

The last version working for me is the 2.4.7.0

 

Each episode is scanned in second, and give error.

Like if smb path were not recognized anymore.

I checked on a library without smb:// but with real path, and it works.

 

[2026-08-20 22:36:47.368] [WARN] [OCR Detection] FFmpeg coarse scan exited with code 1 - hardware acceleration (vaapi) may have failed
[2026-08-20 22:36:47.368] [DEBUG] [OCR Detection] FFmpeg pre-input args: -hwaccel vaapi -vaapi_device /dev/dri/renderD128 -hwaccel_output_format vaapi
[2026-08-20 22:36:47.401] [DEBUG] [OCR Detection] FFmpeg stderr: ffmpeg version 5.1-emby_2023_06_25_p6 Copyright (c) 2000-2022 the FFmpeg developers and softworkz for Emby LLC
  built with gcc 13.4.0 (crosstool-NG 1.28.0)
Execution Date: 2026-08-20 22:36:47
smb://192.168.1.11/NAS001_6/Séries/Loki/Loki - 01x03 - FHD.mkv: Protocol not found
Did you mean file:smb://192.168.1.11/NAS001_6/Séries/Loki/Loki - 01x03 - FHD.mkv?
EXIT

I attached a Dry Run with Debug log

 

Do you have  an idea ?

 

Thanks in advance.

credits-detection-debug-2026-08-20T20-36-50.log 40.25 kB · 0 downloads

Try turning off hardware acceleration in the settings.

[frankythedog 6]

Same result with Hardware detection disabled.

 

2026-08-20 23:41:16.204] [DEBUG] [DetectionCoordinator] Running OCR Detection...
[2026-08-20 23:41:16.236] [DEBUG] [OCR Detection] FFmpeg disk pipeline stderr: ffmpeg version 5.1-emby_2023_06_25_p6 Copyright (c) 2000-2022 the FFmpeg developers and softworkz for Emby LLC
  built with gcc 13.4.0 (crosstool-NG 1.28.0)
Execution Date: 2026-08-20 23:41:16
smb://192.168.1.11/NAS001_6/Séries/Loki/Loki - 01x01 - FHD.mkv: Protocol not found
Did you mean file:smb://192.168.1.11/NAS001_6/Séries/Loki/Loki - 01x01 - FHD.mkv?
EXIT

[2026-08-20 23:41:16.252] [ERROR] [OCR Detection] FFmpeg frame extraction failed with exit code 1
[2026-08-20 23:41:16.252] [ERROR] [OCR Detection] FFmpeg error output: ffmpeg version 5.1-emby_2023_06_25_p6 Copyright (c) 2000-2022 the FFmpeg developers and softworkz for Emby LLC
  built with gcc 13.4.0 (crosstool-NG 1.28.0)
Execution Date: 2026-08-20 23:41:16
smb://192.168.1.11/NAS001_6/Séries/Loki/Loki - 01x01 - FHD.mkv: Protocol not found
Did you mean file:smb://192.168.1.11/NAS001_6/Séries/Loki/Loki - 01x01 - FHD.mkv?
EXIT

[2026-08-20 23:41:16.252] [DEBUG] [DetectionCoordinator] OCR Detection failed: FFmpeg frame 

credits-detection-debug-2026-08-20T21-41-20.log

[yocker 1821]

18 minutes ago, frankythedog said:

Same result with Hardware detection disabled.

 

2026-08-20 23:41:16.204] [DEBUG] [DetectionCoordinator] Running OCR Detection...
[2026-08-20 23:41:16.236] [DEBUG] [OCR Detection] FFmpeg disk pipeline stderr: ffmpeg version 5.1-emby_2023_06_25_p6 Copyright (c) 2000-2022 the FFmpeg developers and softworkz for Emby LLC
  built with gcc 13.4.0 (crosstool-NG 1.28.0)
Execution Date: 2026-08-20 23:41:16
smb://192.168.1.11/NAS001_6/Séries/Loki/Loki - 01x01 - FHD.mkv: Protocol not found
Did you mean file:smb://192.168.1.11/NAS001_6/Séries/Loki/Loki - 01x01 - FHD.mkv?
EXIT

[2026-08-20 23:41:16.252] [ERROR] [OCR Detection] FFmpeg frame extraction failed with exit code 1
[2026-08-20 23:41:16.252] [ERROR] [OCR Detection] FFmpeg error output: ffmpeg version 5.1-emby_2023_06_25_p6 Copyright (c) 2000-2022 the FFmpeg developers and softworkz for Emby LLC
  built with gcc 13.4.0 (crosstool-NG 1.28.0)
Execution Date: 2026-08-20 23:41:16
smb://192.168.1.11/NAS001_6/Séries/Loki/Loki - 01x01 - FHD.mkv: Protocol not found
Did you mean file:smb://192.168.1.11/NAS001_6/Séries/Loki/Loki - 01x01 - FHD.mkv?
EXIT

[2026-08-20 23:41:16.252] [DEBUG] [DetectionCoordinator] OCR Detection failed: FFmpeg frame 

credits-detection-debug-2026-08-20T21-41-20.log 34.94 kB · 0 downloads

Looking in the logs, it might be because you are using SMB shares for your videos.

Will look into it right away.

[frankythedog 6]

SMB problems occured in the first version then has been fixed, and is no longer working after 2.4.7.

So, one days it worked

 

 

Thanks for your investigation.

[yocker 1821]

1 hour ago, frankythedog said:

SMB problems occured in the first version then has been fixed, and is no longer working after 2.4.7.

So, one days it worked

 

 

Thanks for your investigation.

As more features have been added i have had to change the arguments sent to FFmpeg and the fix could get lost with that.
It should be fixed now and works in my testing as it was just reverting/copying from the old version.

It will be added in the catalog as beta version 2.7.0.8 in a minute.
Please report back if it fixed the problem.

[frankythedog 6]

10 hours ago, yocker said:

As more features have been added i have had to change the arguments sent to FFmpeg and the fix could get lost with that.
It should be fixed now and works in my testing as it was just reverting/copying from the old version.

It will be added in the catalog as beta version 2.7.0.8 in a minute.
Please report back if it fixed the problem.

I understand.

You have done and continue to do a fantastic job.

But I feel silly because I can't find any beta versions...

[yocker 1821]

12 minutes ago, frankythedog said:

I understand.

You have done and continue to do a fantastic job.

But I feel silly because I can't find any beta versions...

It should be in the catalog inside emby if you choose beta.

I'm home in an hour, then I can upload it here for you if you want

[frankythedog 6]

1 hour ago, yocker said:

It should be in the catalog inside emby if you choose beta.

I'm home in an hour, then I can upload it here for you if you want

I finally found it in the catalog (I was looking into Github...)

 

It works well  now.

Thanks again for your responsiveness and good work.

 

[yocker 1821]

2 hours ago, frankythedog said:

I finally found it in the catalog (I was looking into Github...)

 

It works well  now.

Thanks again for your responsiveness and good work.

 

Nice to hear.  

Btw sorry, never got around to uploading here either, been a busy day.
Anyway.. Thank you as well, i'm glad you like the plugin!  

[Baenwort 122]

So now that this has support for locally installed Tesseract OCR (FreeBSD and Docker don't get along) I'm looking to see if I missed tuning guidance? I've got this on a 32 core AMD 5950x and I'm guessing I want to tweak the settings to use number of images * number of threads =32?

I'm looking at about around 200+ hours spent on my TV collection and I'd like to reduce it. The server this is running on is mostly used for broadcast archiving so there isn't a large demand on the CPU but I'd like to get this done faster.

[yocker 1821]

27 minutes ago, Baenwort said:

So now that this has support for locally installed Tesseract OCR (FreeBSD and Docker don't get along) I'm looking to see if I missed tuning guidance? I've got this on a 32 core AMD 5950x and I'm guessing I want to tweak the settings to use number of images * number of threads =32?

I'm looking at about around 200+ hours spent on my TV collection and I'd like to reduce it. The server this is running on is mostly used for broadcast archiving so there isn't a large demand on the CPU but I'd like to get this done faster.

Sadly noting i can do about improving Tesseract performance as it's not my program. Making it run multiple instances at the same time (like with PaddleOCR) is not possible either.
How ever, in the performance section in the settings there are a number of settings you can fine tune for better performance.

Paralle OCR Processing.
Direct Memory Pipeling.
Make sure CPU limit is at 100
And a couple more.

In the end it is limited by the speed of the Tesseract though as extracting frames from a video is faster than the tesseract can "read".
It will take some time with that amount of episodes, sorry.  

Edit: Forgot to write that i suggest you update to the latest EmbyCredit (v2.9.0.7) in the catalog as it has a fix for direct memory pipeline and local Tesseract server.
It should save you for a good amount of disk writes.

Edited 19 minutes ago by yocker