Jump to content

Plugin - Chapter Editor (ChapterApi)


TeamB

Recommended Posts

rbjtech
On 07/10/2022 at 23:51, TeamB said:

I added avg into duration per season to the summary page in the plugin. 1.3.0.6

My apologies @TeamB - I've only just seen this.   Thanks for the updates - Much Appreciated!

Link to comment
Share on other sites

  • 2 weeks later...
Khalimerot

Hi, 

I have discover this plugin this morning.

I have made some new entries in the database...sadly i'm more a doer than a reader, so i have a few bad entries in it, the bad one are hidden and can be removed, sorry for that.

I have questions 

I have a few shows with multiple language, and these show have different intro sound because someone speak in it (babylon 5 for example).
Does the ffmpeg command do a chromaprint for all tracks or only one? 
Can i select the track i one to chromaprint? 
If i have to do a chromaprint in french, must i tell somewhere in the database de print come from a french audio source?

Thank you!

Link to comment
Share on other sites

rbjtech
5 hours ago, Khalimerot said:

Hi, 

I have discover this plugin this morning.

I have made some new entries in the database...sadly i'm more a doer than a reader, so i have a few bad entries in it, the bad one are hidden and can be removed, sorry for that.

I have questions 

I have a few shows with multiple language, and these show have different intro sound because someone speak in it (babylon 5 for example).
Does the ffmpeg command do a chromaprint for all tracks or only one? 
Can i select the track i one to chromaprint? 
If i have to do a chromaprint in french, must i tell somewhere in the database de print come from a french audio source?

Thank you!

Is the 'Intro' portion of the Intro in French ?  ie contains spoken French language ?

If not (and it's the same as any other language because it's 'music') then there is no need to create a separate fingerprint as it will make no difference and will work regardless of 'language'.

Link to comment
Share on other sites

Khalimerot

Yes some show have a spoken part in french, usually a little speach presenting the show, not the resume.

It happens often in show from the 90th than recent one, but it happens.

There is also a lot of anime with different intro music in french.

Link to comment
Share on other sites

rbjtech
31 minutes ago, Khalimerot said:

Yes some show have a spoken part in french, usually a little speach presenting the show, not the resume.

It happens often in show from the 90th than recent one, but it happens.

There is also a lot of anime with different intro music in french.

ok cool - in that case yes, you need a different fingerprint - but you can't specify the language as it's not needed.

The plugin scans/uses all available fingerprints for that episode/season/show - and then tries to match them all to get a hit - and then uses the one with the best confidence I believe.

ie if the show had an English fingerprint and a French fingerprint - then both would be used, but it wouldn't only 'hit' on the French version.

 

Link to comment
Share on other sites

TeamB
2 hours ago, rbjtech said:

The plugin scans/uses all available fingerprints for that episode/season/show - and then tries to match them all to get a hit - and then uses the one with the best confidence I believe

yes it works like that, but only at a show/series level, it will pick the best match for each episode from all the chromaprints available for the show/series.

however and i have not tested this, i dont know which audio stream will be used to match against if there are multiple, probably the default but i am not sure.

i wonder how the core handles this.

Link to comment
Share on other sites

Cheesegeezer
10 minutes ago, TeamB said:

yes it works like that, but only at a show/series level, it will pick the best match for each episode from all the chromaprints available for the show/series.

however and i have not tested this, i dont know which audio stream will be used to match against if there are multiple, probably the default but i am not sure.

i wonder how the core handles this.

The core will match one fingerprint against the other, same as the plugin did. if that doesn’t match it will find one that does, once its found a match or 2 matches it should then test all episodes against the master in that season to establish if the episode matches the master intro. The detection is then set and results are based on this.


 

 

Link to comment
Share on other sites

TeamB
7 minutes ago, Cheesegeezer said:

The core will match one fingerprint against the other, same as the plugin did. if that doesn’t match it will find one that does, once its found a match or 2 matches it should then test all episodes against the master in that season to establish if the episode matches the master intro. The detection is then set and results are based on this.


 

 

yep, i understand the detection logic of the approach used in the plugin/core. my question was on what audio track is used if there are multiple.

Link to comment
Share on other sites

Cheesegeezer
15 minutes ago, TeamB said:

yep, i understand the detection logic of the approach used in the plugin/core. my question was on what audio track is used if there are multiple.

There shouldn’t be multiple during detection, if there is… it’s probably down to different encoded audio streams or differences in the streams with non matches… so it would select the first and throw out non matches as no intro detected.

Link to comment
Share on other sites

TeamB
3 minutes ago, Cheesegeezer said:

There shouldn’t be multiple during detection, if there is… it’s probably down to different encoded audio streams or differences in the streams with non matches… so it would select the first and throw out non matches as no intro detected.

so it does try to extract all the audio streams of an episode and do matching against them all?

Example, an episode has 3 audio streams (3 different languages), the plugin/core extracts them all and tries to find a match?

Link to comment
Share on other sites

rbjtech

My understanding is it uses the 'first' available Audio track. 

Language Tracks do not get used because 99.99% of the time, no languages are actual spoken in Intro's - it's music only.

In all the time we have been involved with 'Intros', this is the first time I have heard that there is a different 'Intro' based on a language.

Remuxing the correct 1st Audio track (not necessarily the 'Default' - they are not the same thing) would also solve this.

 

Link to comment
Share on other sites

Khalimerot

I must admit having differents intro based on language, because the intro have spoken elements, or because the intro music is completely different in some countries (Prison Break for example) is an edge case.

It just happens that i have a lot of these cases, as i have a lot of show from the 90th and early 2000.

For myself, it's not a problem, the original language is always the default one, but i was hoping to make intro print for people having files with only the french tracks, wich is an other edge case on the top of the previous :D

 

 

  • Thanks 1
Link to comment
Share on other sites

rbjtech
23 minutes ago, Khalimerot said:

I must admit having differents intro based on language, because the intro have spoken elements, or because the intro music is completely different in some countries (Prison Break for example) is an edge case.

It just happens that i have a lot of these cases, as i have a lot of show from the 90th and early 2000.

For myself, it's not a problem, the original language is always the default one, but i was hoping to make intro print for people having files with only the french tracks, wich is an other edge case on the top of the previous :D

 

 

Really insightful and useful info - thanks.

An enhancement to the choice of 'Intro' track used could be made by using the 'default' track - rather than the first Audio track.

But I think you are correct that generally, a localised version of a source file, will have it's local language as the first and likely the 'default' track anyway. 

Link to comment
Share on other sites

Khalimerot

I don't know if there is a better way, but right now i'm generating prints with ffmpeg is a command prompt, i'll do a batch soon to make it more easier for me.

I don't know what's the plan for the plugin in the future but, would it be possible to automate the fingerprint generation, as ffmpeg is include in emby ?

For example, i would put the intro start and intro end myself in a few episodes of a show in the plugin with manual edition, and when emby is free of work, or when in a set schedule, it would generate fingerprint and wav check files base on my manual chapters.

But perhaps there is already a better way wich i don't know?

 

 

  • Like 1
Link to comment
Share on other sites

rbjtech
1 minute ago, Khalimerot said:

I don't know if there is a better way, but right now i'm generating prints with ffmpeg is a command prompt, i'll do a batch soon to make it more easier for me.

I don't know what's the plan for the plugin in the future but, would it be possible to automate the fingerprint generation, as ffmpeg is include in emby ?

For example, i would put the intro start and intro end myself in a few episodes of a show in the plugin with manual edition, and when emby is free of work, or when in a set schedule, it would generate fingerprint and wav check files base on my manual chapters.

But perhaps there is already a better way wich i don't know?

 

 

I've also thought about automating it (easy enough to do with a script as you say) but the key thing I have found is the accuracy of the submission to the Intro dB in the cloud needs to be good and correctly tagged and therefore imo needs to be done manually.

What I've been doing is as I get a new show - I verify the Intro and then add the Intro to the dB for future (new) episodes.

For existing shows that already have their Intro detected via other means, then there is a level of confidence/verification that can be used - Intro duration per episode (they 'should' all be the same) - but it needs a bit of thought to get the logic correct - the last thing we want to do is upload inaccurate data leading to missed detections for other users.

Link to comment
Share on other sites

rbjtech
9 minutes ago, Khalimerot said:

I don't know if there is a better way, but right now i'm generating prints with ffmpeg is a command prompt, i'll do a batch soon to make it more easier for me.

Just use the Plugin GUI to do it for you - it will extract the WAV file for you, you then verify it - once ok, then just create the JSON directly.   Then import that into the cloud dB.   No need to do it manually .. :)

  • Like 1
Link to comment
Share on other sites

Khalimerot

It's better after having read the docs :D

Stupid me didn't notice the download json / wav icon in the episode list, cause i was jumping from the chapter edition to ffmpeg :D

Link to comment
Share on other sites

Is there a way to insert the same chapter data into multiple episodes at once?

I wanted to add the intro chapters to all the episodes of Bleach (366 episodes), since pretty much all of them (up till now) is at the beginning of the video (1m30s).

Link to comment
Share on other sites

rbjtech
42 minutes ago, testbug said:

Is there a way to insert the same chapter data into multiple episodes at once?

I wanted to add the intro chapters to all the episodes of Bleach (366 episodes), since pretty much all of them (up till now) is at the beginning of the video (1m30s).

Just edit the details page and remove (delete) the Episode value (to cover all episodes in that season) or remove the Season value as well for all the Episodes in all Seasons.

You'll note in the list, many items only have a single entry - but cover all episodes - this is how that is achieved.

 

Edited by rbjtech
Link to comment
Share on other sites

1 hour ago, rbjtech said:

Just edit the details page and remove (delete) the Episode value (to cover all episodes in that season) or remove the Season value as well for all the Episodes in all Seasons.

You'll note in the list, many items only have a single entry - but cover all episodes - this is how that is achieved.

 

Ah, yep, I managed to figure it out. Thanks!

  • Like 1
Link to comment
Share on other sites

  • 2 weeks later...
crusher11

I've had to add a second NAS, due to running out of storage space. Now I have a show with some episodes on one NAS, and other episodes on the other NAS. The ChapterAPI UI just lists the show name twice, instead of merging them together, which means I have to guess which one I want each time.

Link to comment
Share on other sites

sydlexius
51 minutes ago, crusher11 said:

I've had to add a second NAS, due to running out of storage space. Now I have a show with some episodes on one NAS, and other episodes on the other NAS. The ChapterAPI UI just lists the show name twice, instead of merging them together, which means I have to guess which one I want each time.

Forgive me if you've already validated this, but did you configure this option in your TV library?

image.png.28ec1072d4fe837773482b33ea4f3e38.png

Link to comment
Share on other sites

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...