Jump to content

SubRemover – remove unwanted embedded subtitle tracks (new plugin, SubKiller replacement)


Recommended Posts

snake98
Posted (edited)

Hi Emby users, 

SubKiller no longer works on current Emby, so I wrote a new plugin from scratch that does the same job: SubRemover. It removes the embedded subtitle tracks you don't want from your video files and leaves everything else alone. Video and audio are copied as-is, with no re-encoding.

🆕 Latest version: 1.0.3. What changed is under Revisions at the bottom.

✅ What it does

  • Choose the languages to keep: every other embedded subtitle track is removed.
  • Test mode is on by default: nothing is changed; it only logs what it would do.
  • Backups: the original is kept next to the video (movie.mkv.subremover.bak) or copied to a folder you choose. Old backups can be deleted after a number of days.
  • Undo: Restore originals from backups lists every video that has a backup, newest first, and can show only video files, only subtitle files, or only one reason ("Why"). Pick the ones to put back and confirm.
  • Moved videos are found again: move a video, a folder or a whole library on your share and the next run finds it by its file name. Its backup, its "already done" record and the subtitle files SubRemover made for it follow it, so it isn't redone and can still be restored.
  • Safety check: every new file is compared with the original (video and audio tracks, subtitle tracks, length) before it replaces it. If anything doesn't match, the original stays.
  • OCR picture subtitles (PGS / DVD) to text files like Movie.pgs.eng.srt, using the OCR built into Emby's own ffmpeg. Blu-ray subtitles get proper end times, English misreads are cleaned up, and "Both" reads every line with two OCR models and keeps the better one (about 1 word in 100 wrong on a real library).
  • OCR in Emby's subtitle search: "OCR of the built-in subtitles" shows up when you search for subtitles in the player. OCR files remember which OCR made them, and "Redo if made by an older OCR" brings them up to date when the OCR improves.

🆕 New in 1.0.2 and 1.0.3: Fix subtitle file timing

SubRemover can now check the .srt files next to your videos against the actual speech in the video (Subtitle files tab, off by default):

  • Off by the same amount all the way through? Every line is moved to fit, e.g. "off by +13.56 s, moving every cue −13.56 s". It never stretches a file.
  • Several copies in one language? One that's out of sync is deleted when another copy is in sync. When none are in sync, the best one can stay and the worse ones go.
  • Way off, and moving it can't fix it? That usually means a file made for another cut or frame rate. Optionally, English files like that are deleted, even the only one, so Emby can fetch a better one:
    Setting Deletes an English file when…
    Off never (the default)
    Careful its start, middle or end is more than 5 s off
    Recommended its start, middle or end is more than 2 s off
    Most as Recommended, plus files that don't match the speech at all
    Videos whose audio has no language tag can count as English for this (…count audio with no language tag as English, off by default).
  • Scan run measures everything and changes nothing, so you can read the log first. Fix timing now then works from that scan without reading the audio again, and a stopped run carries on where it left off.
  • Every changed or deleted file is backed up first and shows up on the restore list.

🆕 New in 1.0.0

  • Download hash-matched subtitles before changing a video (on by default): subtitle sites like OpenSubtitles find exact-match subtitles by the hash of the original file, and changing a video changes that hash. So before a video is first changed, SubRemover asks Emby's subtitle downloaders for perfect matches in your languages and saves them. Since 1.0.2 it checks the release name against the title and year first, so another film's or a sequel's subtitle is turned down. A daily limit (default 10) leaves the rest of your provider's allowance for Emby. If the search fails or the limit is used up, the video is left unchanged for a later run. Needs a subtitle downloader (such as OpenSubtitles) set up in Emby. Count and Download buttons on the Subtitle files tab do the whole library in one go, without changing any video. AudioRemover follows this setting too.
  • Ads in subtitle files (off by default): takes ad lines like "Support us and become VIP member…", "Advertise your product or brand here" and "Synced and corrected by … www.addic7ed.com" out of the .srt files next to your videos, wherever they are in the film, and numbers the rest again. The list of what's removed is editable (words, sites or regular expressions). Subtitle credits ("Subtitles by…", "Sync and corrections by…") can go too, and TV captioning credits only if you ask. Each changed file is backed up first and is on the restore list, with its own "Delete subtitle file backups older than" setting.
  • Test run (changes nothing) button for the ads: lists every file and the exact lines it would take out, with a link straight to its log, and suggests text that turns up in several films or shows but isn't on the list yet. Remove ads now does it for real.
  • Subtitles tagged with the wrong language: SubRemover reads the start of a text subtitle and calls its tag wrong only when it clearly is. Leave them just logs them; Fix the tag writes the real language into the video, so your keep and remove choices go by what the subtitle really is.
  • Continue an interrupted run where it stopped: a run that was stopped, or cut off by an Emby restart, skips the videos it already checked (and can start again by itself when Emby starts).
  • Videos to check at once (1–8): checking runs side by side at low priority, so a full-library run is much faster; changing videos still happens one at a time.

⚙️ Options

  • Keep subtitles with no language set
  • Guess the language of subtitles with no language set: reads the text (common words, or the alphabet for Greek, Hebrew, Thai, Korean, Japanese and Chinese) and treats a track it can tell as that language. Can be limited to naming extracted files only, so the video itself never changes.
  • Subtitles tagged with the wrong language: leave the tag, or fix it
  • Keep forced subtitles in any language
  • Remove SDH / CC tracks (titles with "CC" or "deaf" count too), only when a regular track in the same language stays, so no video loses its only English subtitles
  • Remove commentary tracks
  • Make the regular subtitle track the default: the default moves off SDH, CC and commentary tracks. Or choose a Default subtitle language.
  • Extract kept text subtitles to .srt files that Emby picks up. Optionally take them out of the video, and name the regular one Movie.eng.default.srt so Emby turns it on.
  • Prefer subtitle files that are already there (e.g. downloaded by Emby): SubRemover makes no file of its own in that language, and Emby turns your file on.
  • Download hash-matched subtitles before changing a video, with a daily limit
  • Remove ads from subtitle files, with an editable list and switches for subtitle credits and TV captioning credits
  • 🆕 Fix subtitle file timing, with switches for deleting out-of-sync copies, worse copies and files that are way off, and how many videos are measured per run and at once
  • 🆕 With Test mode on, still download hash-matched subtitles and clean up subtitle files: the videos stay untouched, but the downloads, ad removal and timing fixes still happen
  • OCR picture subtitles (PGS / DVD) to text files like Movie.pgs.eng.srt, using the OCR built into Emby's own ffmpeg
  • OCR in Emby's subtitle search: OCR of the built-in subtitles shows up when you search for subtitles in the player

🖥️ Easy to use

  • Its own settings page with tabs: Run, Subtitles, Subtitle files, Backups and Logs. One Save covers every tab.
  • Each tab explains its own settings in a Help section at the bottom.
  • Run now with a live status line, and a log viewer on the Logs tab: open any run log in a new window or download it. 🆕 The log viewer can filter by what an entry is about (video tracks, subtitle files, hash-matched downloads, OCR), and each button on the Subtitle files tab shows its last result, even after Emby restarts.
  • Process new videos as they're added: when Emby adds videos, SubRemover waits until none have arrived for 3 minutes, then checks just those in one run (a whole season is one run with one log).
  • Quick setup button: English + subtitles with no language
  • An entry in the server sidebar under Advanced (can be turned off)
  • Limits for testing: Max videos per run, and Only process paths containing to try it on a single title (several titles separated with ;)
  • Every run log starts with your settings and ends with totals of what was (or would be) done.
  • Works alongside AudioRemover, its twin for audio tracks: the two never rewrite the same file at the same time, and either one gets the hash-matched subtitles first, so the order they run in doesn't matter.

📥 Install

  1. Download Emby.Subremover.dll from the GitHub Releases page (take the newest one):
    https://github.com/Snake98bm/Emby.Subremover/releases
  2. Stop Emby Server and copy the DLL into Emby's plugins folder:
    • Windows: %AppData%\Emby-Server\programdata\plugins
    • Linux: /var/lib/emby/plugins
    • Docker (emby/embyserver or linuxserver/emby): /config/plugins inside the container. On the host, that's the plugins folder inside the folder you mapped to /config, e.g. /docker/emby/config/plugins. Stop the container (docker stop emby), copy the DLL there, and make sure it has the same owner as the other files in that folder (your PUID/PGID).
      Or copy it straight into the container: docker cp Emby.Subremover.dll emby:/config/plugins/
      (Replace emby with your container's name.)
  3. Start Emby (Docker: docker start emby) and press Ctrl+F5 in your browser.
  4. Go to Settings → Plugins → SubRemover. On the Run tab pick your libraries and leave Test mode on. On the Subtitles tab pick your languages (or click Quick setup). Click Save.
  5. Click Run now, then read the run log on the Logs tab.
  6. When the log looks right, turn Test mode off, set Max videos to change per run to a few, and run it again. Check those files in Emby before letting it do the rest.

Updating: same as installing. Stop Emby, replace the DLL, start Emby and press Ctrl+F5. Your settings carry over. All the timing options start off. To try them, turn on Fix subtitle file timing on the Subtitle files tab, click Scan run, and read its log before you click Fix timing now.

🧪 Tested on

  • Emby Server 4.10.2 on Windows, with full-library test runs and real runs (8,000+ videos, 2,800 subtitle files checked for ads, 2,700 subtitle files of 2,300 videos checked for timing)
  • The build and tests also pass on Linux, but I haven't run it on a real Linux or Docker server yet. Feedback welcome!

📝 Revisions

  • 1.0.3 – Way-off subtitle files: one bad part (start, middle or end) is now enough, so a file that's right until a jump near the end (another cut) goes too. New option …count audio with no language tag as English (off by default). Files left alone before are checked again on the next Fix.
  • 1.0.2 – New: Fix subtitle file timing (moves .srt files that are off by one amount, deletes out-of-sync and worse copies, optionally deletes English files that are way off; Scan run, Fix timing now, Videos per timing run, Videos measured at once). With Test mode on, hash-matched downloads and subtitle file cleanup can still run. Hash-matched downloads check the title and year first. Restore list "Why" filter, log viewer filter, and the last result of each button task kept across restarts. OCR is no longer marked experimental.
  • 1.0.1 – The restore list has a Show choice: All backups, Video files only, or Subtitle files (.srt) only.
  • 1.0.0 – New: Download hash-matched subtitles before changing a video (on by default, with a daily limit and Count / Download buttons for the whole library), Ads in subtitle files (editable list, credits, Test run button with a link to its log, suggestions, backups on the restore list), Subtitles tagged with the wrong language (leave or fix the tag), Continue an interrupted run where it stopped, and Videos to check at once. Fixed: the list of subtitle files SubRemover made no longer grows huge in its saved state.
  • 0.7.8 – MP4 / MOV videos keep the default flag on their last subtitle (the MP4 format always turns one on, so taking it off made the safety check fail every run); unwanted tracks are still removed. The restore list says in plain words what a default change did.
  • 0.7.7 – OCR files remember which OCR made them; new "Redo if made by an older OCR" redoes only outdated ones, including files from Emby's subtitle search. The restore list says under each video what SubRemover changed. The OCR model help shows your server's Tesseract version.
  • 0.7.6 – Much better OCR: Blu-ray subtitles end when the disc clears them (they used to stay on screen), English misreads are fixed, and the new OCR model setting (Fast, Accurate or Both) can read every line with two models and keep the better one. Words wrong on a real library: 3.5% → 1.0%.
  • 0.7.5 – New settings page with tabs (Run, Subtitles, Subtitle files, Backups, Logs), each with its own help, and one Save for all of them. Restore is now one list with tick boxes. Videos moved on your share are found again by file name, and their backups and records follow them.
  • 0.7.4 – The Help section on the settings page no longer shows old or repeated entries after an update.
  • 0.7.3 – The run log names each library ("Movies (1234)") instead of just its number. A language tag of "unknown" or "null" counts as no language. A video AudioRemover is rewriting at that moment is left for the next run.
  • 0.7.1 – New: Process new videos as they're added. Open or download any run log from the settings page. Tracks that are removed anyway are no longer checked for SDH.
  • 0.7.0 – New: default subtitles (the regular track becomes the default, Default subtitle language, .default file names), Guess the language of subtitles with no language set, OCR in Emby's subtitle search, and Prefer subtitle files that are already there. Restore now lists every backup so you pick which videos to restore. Run logs start with your settings and end with totals. Help section on the settings page.
  • 0.6.8 – Titles with "CC" or "deaf" count as SDH. A commentary track is never treated as the regular track, so Remove SDH can't leave a video with only commentary subtitles. OCR skips commentary tracks.
  • 0.6.x – First release.

Source code, issues and the MIT license are on GitHub:
https://github.com/Snake98bm/Emby.Subremover

🔊 Its twin for audio, AudioRemover (removes foreign-language audio tracks, optionally checking the language with Whisper):
GitHub: https://github.com/Snake98bm/Emby.AudioRemover
Forum thread: AudioRemover – remove foreign-language audio tracks

⚠️ Please keep Test mode on for your first run and keep backups on. If something looks wrong in the log, post it here or open an issue on GitHub.

 

 

Emby.Subremover.dll

Edited by snake98
update to 1.0.3
  • Like 2
Dickydodah!
Posted

@snake98This is really good news. I will be testing it out very soon. This will be a godsend for older Samsung TV owners (the TV's old not necessarily the owner 🤣) as they suffer greatly with too many subtitles causing major playback issues. I spent a lot of time testing the origional pluging with @Cheesegeezerand we ironed out a fair few bugs so be prepared 😄

  • Like 1
snake98
Posted

Glad this is useful for someone else, just send the example and log files, i'll be happy try to fix them.  Make sure to run the test first, it should show you any errors, and you do run it i suggest using the filter only process path containing to start out with, so your backups don't get huge.

  • Like 1
Dickydodah!
Posted

I've installed the dll and ran it in test mode on one of my libraries. I will be adding a new library to actually do a real test. Everything looks good so far 👍

I have a few questions/observations 😀

I really like the option to extract to external SRT. Does this option remove the internal subtitle or do you end up with both? I would be interested to hear your opinion on why you would want to have external SRT files. Also does the extraction have much of an overhead that may be an issue on a low powered machine?

The OCR option looks really cool but I doubt I personally would use it.

The backup option is again a really great addition.

The "max videos to change per run" made me wonder why?? Other than testing what do you use this for?

The logs are really well laid out and informative. A nice to have would be to be able to "open in a new window" button as there is one in the default Emby logs section. Maybe even a download button? I know that the SubRemover logs can be viewed in the Emby logs page but it would be nice.

snake98
Posted

🆕 SubRemover 0.7.0 is out

Download: https://github.com/Snake98bm/Emby.Subremover/releases (previous version: 0.6.8)

What's new

  • Default subtitle: the default moves off SDH, CC and commentary tracks to the regular one; you can pick a default language; optional .default file names
  • Language guessing: works out the language of subtitles with no language set, and spots SDH from the text
  • Wrong language tags: catches tracks tagged English that are really another language
  • Smarter OCR: only runs when there's no text subtitle in that language
  • OCR from Emby's "Search for Subtitles": optional, with a message on the player while it works
  • Prefer subtitle files that are already there, e.g. ones Emby downloaded
  • Restore list on the settings page: pick exactly which videos to put back
  • Logs list all your settings at the top and end with totals
  • Only process paths containing takes several titles, separated by ;
  • Cleaner settings page with a Help section at the bottom

Upgrading

  1. Stop Emby and replace Emby.Subremover.dll in your plugins folder with the new one.
  2. Start Emby and press Ctrl+F5 in your browser.
  3. Your settings carry over. The new options start off, except Make the regular subtitle track the default, which is on.
  4. Because the settings changed, the next run checks every video again. Do a run with Test mode on first and read the log under Log files.

Feedback welcome, especially from Linux and Docker users.

snake98
Posted
1 hour ago, Dickydodah! said:

I've installed the dll and ran it in test mode on one of my libraries. I will be adding a new library to actually do a real test. Everything looks good so far 👍

I have a few questions/observations 😀

I really like the option to extract to external SRT. Does this option remove the internal subtitle or do you end up with both? I would be interested to hear your opinion on why you would want to have external SRT files. Also does the extraction have much of an overhead that may be an issue on a low powered machine?

The OCR option looks really cool but I doubt I personally would use it.

The backup option is again a really great addition.

The "max videos to change per run" made me wonder why?? Other than testing what do you use this for?

The logs are really well laid out and informative. A nice to have would be to be able to "open in a new window" button as there is one in the default Emby logs section. Maybe even a download button? I know that the SubRemover logs can be viewed in the Emby logs page but it would be nice.

Reply to the questions in your SubRemover thread. Click inside the box below, press Ctrl+A then Ctrl+C, and paste into the reply box with Ctrl+V (use the forum's Quote button on their post first if you like).

Thanks for testing it and the feedback

Extract to external SRT: do you end up with both?
By default, yes: the track is saved as a .srt file next to the video and stays inside the video. If you only want the file, turn on "...and remove the extracted tracks from the video" (it shows under the extract option once that's on).

Why external SRT files? 

Two main reasons, Emby can hand an external text file straight to the player. With embedded subtitles it sometimes has to pull them out of the video first, which can delay the start or add to a transcode. Choosing what comes on: with "Mark the regular subtitle file as default" the file is named Movie.eng.default.srt, so Emby turns that one on.

If you keep both, the language shows twice in the subtitle menu (the file and the embedded track). That's the main reason to use the "remove" option.

Overhead on a low-powered machine?
Very little CPU is needed,  OCR is the only CPU-heavy part.

"Max videos to change per run":

Mainly used for testing, or the first big run so you can process for example 20 files a day.  The backups would double your space needed, this allows you to run 20 a day, and set 7 day backup, so you have time to spot check files and logs.

Logs: open in a new window / download
I'll try to put the in the next release.

if you have problems, direct message me the log file and I"ll try to see if i can fix it.

Dickydodah!
Posted

Thanks for the explanation, all makes sense. I'll probably leave the embeded sub to start with as I don't think I've ever really had any issues apart from too many embedded subs screwing up playback on my older Samsung TV.

I have some other questions about the general usage and how I can achieve my quite simple needs. Keeping it really simple I just need to remove all subtitles except English forced. I never watch foreign language shows or movies as they drive my wife nuts 🤣 I can see that your plugin can achieve this quite easily except that the English non forced would also be retained but thats not an issue.

Currently I use fileflows which uses filesystem events to "see" a new file has been added and then add it to the processing queue which runs imediatley. This is better than a schedule as processed files are available to watch much quicker. Where fileflows is falling down is when a file is replaced with a better quality but the same filename (upgraded) it doesn't get processed so I have to manually intervene.

From what I've seen so far your plugin needs to be run manually or on a schedule. Not a big problem but I assume it would need to rescan every file especially if they have the same name as the origional one. This may not be an issue as the overheads are small but with 5000+ files it may turn into one. What are your thoughts?

 

snake98
Posted
5 minutes ago, Dickydodah! said:

Thanks for the explanation, all makes sense. I'll probably leave the embeded sub to start with as I don't think I've ever really had any issues apart from too many embedded subs screwing up playback on my older Samsung TV.

I have some other questions about the general usage and how I can achieve my quite simple needs. Keeping it really simple I just need to remove all subtitles except English forced. I never watch foreign language shows or movies as they drive my wife nuts 🤣 I can see that your plugin can achieve this quite easily except that the English non forced would also be retained but thats not an issue.

Currently I use fileflows which uses filesystem events to "see" a new file has been added and then add it to the processing queue which runs imediatley. This is better than a schedule as processed files are available to watch much quicker. Where fileflows is falling down is when a file is replaced with a better quality but the same filename (upgraded) it doesn't get processed so I have to manually intervene.

From what I've seen so far your plugin needs to be run manually or on a schedule. Not a big problem but I assume it would need to rescan every file especially if they have the same name as the origional one. This may not be an issue as the overheads are small but with 5000+ files it may turn into one. What are your thoughts?

 

I just check, Emby allows the plugs to run with new files, i'll add that option next week. 

The second, Subremover tracks the path, file size and modified time, if any of those change it sees it as a new file.

I'll have to think about the force option, you are depending a lot on it being marked right in the file.  I ran across english subtitles marked as french, etc or other flags that where completly wrong.  Since it deletes it, there no going back after the backups are deleted.  

Does that answer your question?

Dickydodah!
Posted

I wouldn't worry too much about the forced only subs. At the moment I end up with English, Forced English and English SDH. I have Emby set to always show forced subs and it works very well. I too have come across badly named subs but so far its been very rare so not an issue. I was just thinking aloud when I saw that you have the option to delete SDH.

I will get a proper test setup going very soon and see if I can break it 🤣

I did notice something in the logs that you may be able to explain. It's not important I was just wondering.

Subremover indexes subs with a numer as below;

remove: #4 eng subrip SDH "SDH"

However if I look at the same file with mediainfo I get different info, in this case SDH is ID 5

Text #1
ID                                       : 5
Format                                   : UTF-8
Codec ID                                 : S_TEXT/UTF8
Codec ID/Info                            : UTF-8 Plain Text
Duration                                 : 1 h 45 min
Bit rate                                 : 40 b/s
Frame rate                               : 0.173 FPS
Count of elements                        : 1100
Stream size                              : 31.2 KiB (0%)
Title                                    : SDH
Language                             : English
Default                                  : Yes
Forced                                   : No

snake98
Posted
On 10/29/2026 at 3:40 PM, Dickydodah! said:

I wouldn't worry too much about the forced only subs. At the moment I end up with English, Forced English and English SDH. I have Emby set to always show forced subs and it works very well. I too have come across badly named subs but so far its been very rare so not an issue. I was just thinking aloud when I saw that you have the option to delete SDH.

I will get a proper test setup going very soon and see if I can break it 🤣

I did notice something in the logs that you may be able to explain. It's not important I was just wondering.

Subremover indexes subs with a numer as below;

remove: #4 eng subrip SDH "SDH"

However if I look at the same file with mediainfo I get different info, in this case SDH is ID 5

Text #1
ID                                       : 5
Format                                   : UTF-8
Codec ID                                 : S_TEXT/UTF8
Codec ID/Info                            : UTF-8 Plain Text
Duration                                 : 1 h 45 min
Bit rate                                 : 40 b/s
Frame rate                               : 0.173 FPS
Count of elements                        : 1100
Stream size                              : 31.2 KiB (0%)
Title                                    : SDH
Language                             : English
Default                                  : Yes
Forced                                   : No

Okay thanks for the feedback.

  • SubRemover uses ffmpeg’s stream index, which starts at 0. The video is usually #0, audio #1, and so on. Emby uses the same numbering in its media info.
  • MediaInfo’s “ID” for an MKV is the track number stored in the file, which starts at 1.
Dickydodah!
Posted

Thanks for the explanation, I thought it would be down to differences in the way software indexes it. Theres's nothing like standardization and this is nothing like a standard 🤣

snake98
Posted

🆕 SubRemover 0.7.7 is out: much better OCR

Download: https://github.com/Snake98bm/Emby.Subremover/releases (previous version here: 0.7.5)

OCR of picture subtitles (0.7.6)

  • Blu-ray subtitles leave the screen again. OCR'd Blu-ray (PGS) subtitles used to stay on screen until the next line, or for the rest of the movie. They now end exactly when the disc clears them.
  • Common misreads are fixed in English: 1, l, / or ! read as "I" ("1'll" → "I'll"), lost or misread brackets ("[ Door Opens /" → "[ Door Opens ]"), music notes (♪), missing apostrophes ("dont"), and stray capitals.
  • New setting, OCR model: Fast, Accurate, or Both (the default). Both reads every subtitle with Tesseract's fast and best models at the same time, on two CPU cores, and keeps the better reading of each line, checked against a built-in English word list.
  • Measured on a real library (22 movies and episodes with both a picture and a text subtitle track, about 11,000 words): words wrong went from 3.5% to 1.0%.
OCR Words wrong
Before (fast model, no cleanup) 3.5%
Fast + cleanup 1.7%
Accurate + cleanup 1.5%
Both + cleanup (default) 1.0%

Redo older OCR files (0.7.7)

  • Every OCR file now remembers which OCR made it. New choice under If the OCR subtitle file already exists: Redo if made by an older OCR. It redoes only files from an older OCR (or a weaker model than the one you chose) and leaves good ones alone. Files from before 0.7.7 count as older, so the first run with it on brings your existing OCR files up to date.
  • It also covers OCR you picked in Emby's subtitle search: SubRemover remembers it and redoes the file Emby saved, unless you deleted or edited it.

Restore list says what changed (0.7.7)

  • Under each video, one or two lines say what SubRemover did, for example "Removed 3 subtitles: French, Spanish (2) · default now English". Backups made before 0.7.7 say their changes weren't recorded.
  • The OCR model help on the settings page shows the Tesseract version your server's ffmpeg has.

Upgrading

  • Stop Emby and replace Emby.Subremover.dll in your plugins folder with the new one.
  • Start Emby and press Ctrl+F5 in your browser.
  • Your settings carry over. OCR model starts on Both; nothing else new is turned on. To bring your existing OCR files up to date, choose Redo if made by an older OCR and run.

Feedback welcome, especially OCR results in other languages (the cleanup is English only; the timing fix and the Both model work for every language).

Dickydodah!
Posted

I have installed the latest version V1.0.1 and you have been very busy. There is a massive range of options now but they are well explained. This is very much a comprehensive tool to manage Subs in Emby.

Currently I'm running it on my library and I've set it to one file at a time as my server is not the most powerful. I may have been a bit conservative as there seems to be very little CPU activity so I'll probably up it to 2 or 3 once the whole library is completed.

I think I may have made a mistake by leaving the backup option on as that may be slowing it down a bit and also using a huge amount of space so I cancelled the run and changed the backup to off, I'm not processing any unreplaceable files so I'm not concerned. I did notice that when I manually deleted the *.bak files SubRemover still thinks they are still available even if I restarted Emby.

Posted

Send over your log files in chat, and I"ll  see if any errors.  I'm going to add 1 more options(always 1 more) to clean up bad srt files, that the timing are off.  I'm using this plugin to cleanup my media also.

Dickydodah!
Posted

It's still running and is at 89% so I'll wait until its finished and grab any logs and a screenshot and send them. 

Dickydodah!
Posted

It's finished now and I can see that the GUI shows that the backup files are missing.

image.png.24109c242ee09f6c9b73aa0896c24542.png

If I try to restore the file it of course fails and the entry is removed from the list. The entry in the log is as below;

2026-11-16 16:15:11.115 Info  SubRemover: restoring 1 backups. Details: subremover-restore-20261116-161511.txt
2026-11-16 16:15:11.116 Warn  SubRemover: backup Z:\Movies\Hangman (2017) [ImdbId=tt3011960]\Hangman (2017) {tmdb-323368} [Bluray-1080p][AC3 5.1][x264]-VietHD.mkv.subremover.bak is missing, forgetting it
2026-11-16 16:15:11.161 Info  SubRemover: restored 0 of 1 backups. Details: subremover-restore-20261116-161511.txt

It seems it is working correctly but there is a bit of a hole in the logic as there seems to be no way to manually delete an individual backup. I've set the backups to be deleted after 1 day now and I expect that will delete the ghost files I have. However some people might like to keep specific backups, not me though as I have far to many files to keep the backups.

 

Posted

hmm, never though about delete the backups one at a time, i'll add that in the next release

 

Dickydodah!
Posted

It may be a bit superfluous but I'm sure that it could be useful at times, such as making a mistake like I did 🙁 I have noticed that my change to delete backups after 1 day has cleared all the ghost files. I was slightly surprised that there was no error in the log as the file didn't actually exist but I suppose it would be a bit pointless to record it.

I shall monitor the logs every day for a while and maybe play around with a few more settings but it does seem to be doing exactly what I needed with some great extras available.

snake98
Posted

🆕 SubRemover 1.0.3 is out: fixes the timing of your subtitle files

Download: https://github.com/Snake98bm/Emby.Subremover/releases (previous version: 1.0.1)

New: Fix subtitle file timing (Subtitle files tab)

SubRemover now checks the .srt files next to your videos against the actual speech in the video:

  • Off by the same amount all the way through? Every line is moved to fit, e.g. "off by +13.56 s, moving every cue −13.56 s".
  • Several copies in one language? One that's out of sync is deleted when another one is in sync. When none are, the best one stays and the worse ones go (optional).
  • Way off, and moving it can't fix it? That usually means a file made for another cut or frame rate. Optionally, English files like that are deleted, even the only one, so Emby can fetch a better one.
  • Scan run measures everything and changes nothing, so you can read the log first. Fix timing now then works from that scan without reading the audio again.
  • Every changed or deleted file is backed up first and shows up on the restore list.

New in 1.0.3: "...delete subtitle files that are way off"

Setting Deletes an English file when…
Off never (the default)
Careful its start, middle or end is more than 5 s off
Recommended its start, middle or end is more than 2 s off
Most as Recommended, plus files that don't match the speech at all
  • Changed: one bad part is now enough. Before, a file that was fine until a jump near the end (another cut) was kept.
  • New option: "...count audio with no language tag as English". Lots of videos have audio tagged und or nothing, and those were always skipped. Turn this on if your untagged audio is English. It's off by default, so leave it off if you have untagged foreign-language videos.
  • Small drifts (under 2 s) are still left alone, and so are forced, commentary and audio-description files.

Also since 1.0.1

  • Log filter on the Logs tab, and each button task shows its last result, even after Emby restarts
  • The restore list has a "Why" filter: ads, timing, way off and more
  • Hash-matched downloads check the release name against the title and year first, so a wrong film or a sequel is turned down
  • With Test mode on, the new "...but still download hash-matched subtitles and clean up subtitle files" lets runs download hash matches, remove ads and fix timing while the videos stay untouched

Upgrading

  1. Stop Emby and replace Emby.Subremover.dll in your plugins folder with the new one.
  2. Start Emby and press Ctrl+F5 in your browser.
  3. Your settings carry over, and all the new timing options start off.
  4. Turn on Fix subtitle file timing on the Subtitle files tab, click Scan run, and read its log before you click Fix timing now.
  5. If you already use the way-off setting, the next Fix checks the files it left alone before again.

Feedback welcome, especially from Linux and Docker users.

Emby.Subremover.dll

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...