Info on subtitle fetching

[f3bruary]

I've been using filebot since a few weeks for subtitle fetching and I have some questions about the inner workings.

When I start the task using 'filebot -get-missing-subtitles --lang en /path/to/video.mkv'

It first starts a query using filehash. When it's done and hasn't found subtitles for certains files, it falls back on name checking. Is there any way to prevent this ? My SickBeard and CouchPotato automatically rename files so the name checking doesn't guarantee perfect subs like filehash based searching does.

I also would like to now what exactly the -non-strict parameter doesn. I see many people use it and I've read what filebot -help says about it, but it still isn't clear to me.

If someone could clarify this for me, I would appreciate it.

-f3b

[rednoah]

In 4.0 it does use name search to some degree even in strict mode, that's changed in 4.1 RC3 where no name search is performed in strict mode.

In strict mode it will now only use "hash" and "tag" look up, the latter being an exact match for the file name.

-non-strict just means it'll be less strict in various aspects of operations, e.g. what is considered a "good enough" subtitle match, or episode match

[f3bruary]

And if I don't want 'tag' search ? Cause this could fetch wrong subs too.

Also, I just downloaded 4.1 RC3 and when I started a query, it couldn't see the existing subtitles, it tried to fetch everything again.

[rednoah]

Then you're out of luck. Better grab subs before you rename the files. But as long a there's a hash match first this can't be an issue, so if you find a bad subtitles, make sure to upload a better one with filebot.

Also, I need logs if you want to report an issue.

[f3bruary]

filebot 4.1 RC3 does indeed ignore existing subtitles.

Example: Command used first: filebot -get-subtitles --lang tr --encoding utf8 /location/*
Second pass: filebot -get-missing-subtitles --lang tr --encoding utf8 /location/*

This is just an example. The actual command I just used is: filebot -get-missing-subtitles --lang tr --encoding utf8 --output srt /volume1/video/movies/A* --log-file /volume1/public/filebot.log --log all

It doesn't seem to contain any errors, just screen output. Please tell me if I used the log syntax wrong and I'll try again. Anyways, this didn't happen in 4.0.

logs: http://pastebin.com/nPf7gymp

[rednoah]

And? What behaviour do you expect? Is there a question in this last post somewhere?

[f3bruary]

Well, isn't -get-missing-subtitles supposed to see if there are existing subtitles first ? and if not, it will then fetch them ? I already have subtitles in those folders, yet it's fetching them again.

In 4.0 this didn't happen. It would only fetch where needed.

[rednoah]

The logs don't contain info about subtitle files that already exist so it's hard to find the issue... it doesn't happen if you already have tur subtitles, but if you already have eng and tur subtitles.

Fixed with r2232.

[f3bruary]

I'm running FileBot 4.1 (r2218) and fetch en, nl and tr subs. Every now and then I do another pass for missing subs and this version just ignores existing subs and fetches new ones.

I'll try out the r2232 version. Is there an IPK available yet ? I'm running it on a Synology NAS

[rednoah]

Nope, for these minor updates you can just replace the .jar

[f3bruary]

I have, thanks a lot