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

[yocker 1957]

@fbourgAre the videos french? I don't mean just the audio but french speaking with french end credits and so on?

I noticed this in your settings file:
<OcrDetectionKeywords>associate producer,based on,cast,casting,cinematography,co-producer,composer,costume design,created by,credits,developed by,directed by,director of photography,editing,editor,end credits,ende,executive producer,fim,fin,fine,guest starring,music by,produced by,producer,production company,production design,screenplay,series producer,sound,special thanks,starring,story by,the end,visual effects,written by</OcrDetectionKeywords>

OCR needs keywords to look for and if it doesn't find them it will fail.
So you might need to add keywords like for example "histoire" and "acteurs" for it to look for.
The more you add the more precise it will be.

The Chroma/hash detection seems to work as it should.
If audio is too different between episodes it will fail because it can't find a common sound to compare with.

[yocker 1957]

@fbourgPlease quickly edit your message with the file as it's not allowed here. I should have clarified that it should have been in PM.. Sorry about that.

[fbourg 5]

2 minutes ago, yocker said:

@fbourgAre the videos french? I don't mean just the audio but french speaking with french end credits and so on?

I noticed this in your settings file:
<OcrDetectionKeywords>associate producer,based on,cast,casting,cinematography,co-producer,composer,costume design,created by,credits,developed by,directed by,director of photography,editing,editor,end credits,ende,executive producer,fim,fin,fine,guest starring,music by,produced by,producer,production company,production design,screenplay,series producer,sound,special thanks,starring,story by,the end,visual effects,written by</OcrDetectionKeywords>

OCR needs keywords to look for and if it doesn't find them it will fail.
So you might need to add keywords like for example "histoire" and "acteurs" for it to look for.
The more you add the more precise it will be.

The Chroma/hash detection seems to work as it should.
If audio is too different between episodes it will fail because it can't find a common sound to compare with.

Should I also set it to eng+fra in the OCR settings?

[yocker 1957]

Just now, fbourg said:

Should I also set it to eng+fra in the OCR settings?

That settings is only for if extra training files have been given to the OCR engines.
It doesn't really matter for what is being done here as long as you add the right keywords.

[yocker 1957]

EmbyCredits 2.9.0.0 has been released in the catalog.

Changed:
1) Much better and precise chroma/hash detection and precision!
It is now where i could call it almost on par with OCR detection and much faster.

2) When it encounters a series that uses different music between episodes it will instead fallback to using blackframe detection.
This fallback has also had it's precision improved greatly.

3) OCR detection has had the way it communicates with the OCR engines changed slightly.
It no longer just keeps sending frames for the engine to run detections on but patiently waits for it to be done with each frame. This makes sure that no frame is ever missed for detection.
This also improves precision.
Note: This can make OCR feel slightly slower on rare occasions but only because no frames were lost, talking about 100 frames per multible 1000's on a normal machine.

4) Changed a few things about the chroma/hash and blackframe detection that should improve performance slightly.

Fixed:

1) Removed the plugins own episode comparison from chroma/hash detection as it messed with the detections.

2) Fix a bug that could in very rare occasions cause the settings page to stall and make it look like the plugin had crashed.

3) Fixed blackframe detection fallback not engaging when it should.

4) Other fixes and optimizations.

Special thanks to @Neminemfor help with testing.

[fbourg 5]

thank you

[yocker 1957]

@RedNo7 I almost forgot to mention.
I suggest you try the chroma/hash detection now instead of OCR, it's much more CPU friendly and with the improved precision it's almost as good.

[RedNo7 35]

Thanks, I’ll do that!

[RedNo7 35]

Is this the setting?

[yocker 1957]

3 minutes ago, RedNo7 said:

Is this the setting?

Enable Hash Detection, the one you have chosen there uses OCR as fallback in case Chroma/Hash fails.

[RedNo7 35]

thanks!

edit: wow it's bombing along!

Edited 14 hours ago by RedNo7

[fbourg 5]

The plugin stopped again overnight. Here are the files. Version 2.9

 

embyserver-63925725666.txt embyserver.txt

[fbourg 5]

its new version 2.9.0.1 ?

[yocker 1957]

@fbourgDoes the plugin act like it was done or has it stalled mid detections?

Looks like a lot of episodes was added to the queue that already had time stamps or failed and the plugin skipped them.
If you look purely on the ETA then the time left can be very deceptive and can fluctuate A LOT.

You add items to queue.
It start detections on episodes and the ETA goes up to lets say 2 hours.
Then 80% of the files suddenly doesn't have to have detections run anyway and the ETA can seem to be wrong because most of the work suddenly disappeared. 

That said, looking in the logs i might have found a problem.
I will have a version ready for you to try soon.

Edited 1 hour ago by yocker

[fbourg 5]

Hi @yocker,

To answer your question: it definitely stalled mid-detection rather than finishing early or skipping files.

Out of roughly 62,000 episodes in my library, only around 3,000 have been processed so far. The remaining ~59,000 are untouched and do not have any credit markers or previous timestamps.

Looking closely at the log, the batch processing kicked off for Les Griffin Season 5 at 00:15:20.395, and then the plugin stopped logging completely while server-side TMDB queries were running. The scheduled task just hung right there.

I'm ready to test your upcoming build as soon as it's available. Thanks for looking into it!

[yocker 1957]

19 minutes ago, fbourg said:

Hi @yocker,

To answer your question: it definitely stalled mid-detection rather than finishing early or skipping files.

Out of roughly 62,000 episodes in my library, only around 3,000 have been processed so far. The remaining ~59,000 are untouched and do not have any credit markers or previous timestamps.

Looking closely at the log, the batch processing kicked off for Les Griffin Season 5 at 00:15:20.395, and then the plugin stopped logging completely while server-side TMDB queries were running. The scheduled task just hung right there.

I'm ready to test your upcoming build as soon as it's available. Thanks for looking into it!

Doesn't look to be stalled, there's just a lot of noise in the log from another plugin (TMDB api lookups).
FFmpeg did cause an error at one point, i'm working on adding a guard againts that but other than that i don't really see a problem in the log.

When that version is done can i get you to make the plugin save individual logs?


That will remove all the noise.

[fbourg 5]

3 minutes ago, yocker said:

Doesn't look to be stalled, there's just a lot of noise in the log from another plugin (TMDB api lookups).
FFmpeg did cause an error at one point, i'm working on adding a guard againts that but other than that i don't really see a problem in the log.

When that version is done can i get you to make the plugin save individual logs?


That will remove all the noise.

Already done! "Save Debug Logs to File" is enabled and ready to go. Just let me know whenever the new build is up, and I'll run it right away to grab the isolated logs!

[yocker 1957]

8 minutes ago, fbourg said:

Already done! "Save Debug Logs to File" is enabled and ready to go. Just let me know whenever the new build is up, and I'll run it right away to grab the isolated logs!

An hour or so and it should be ready if everything goes to plan.