Jump to content

Josh.5

Members
  • Content Count

    62
  • Joined

  • Last visited

  • Days Won

    3

Josh.5 last won the day on January 8

Josh.5 had the most liked content!

Community Reputation

46 Good

5 Followers

About Josh.5

  • Rank
    Advanced Member

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Subtitles are a little tricky. At the moment unmanic has real shitty support for the subtitle streams in your video files. The problem is that different containers have different implementations on how subtitles are stored and therefore ffmpeg requires different arguments depending on your combination of subtitles, src container and destination container. Removing embedded subtitle streams will fix a lot of errors when converting between containers until I expand on the settings capabilities of unmanic and how it generates tasks for ffmpeg. Sent from my ONEPLUS A6000 using Tapatalk
  2. Actually that output is enough as far as unmanic goes. It looks like ffmpeg did not like the arguments it was passed or the file needed some kind of special configuration. Can you please try running that command manually from a terminal inside the docker container. Ffmpeg will give you some output that will show what is wrong. Then can you give me that output? Sent from my ONEPLUS A6000 using Tapatalk
  3. Oh thats a different one. I'll need to grab some sources off you to debug that if you don't mind. Sent from my ONEPLUS A6000 using Tapatalk
  4. I'll finish the next lot of changes before we look further at it again. These next upcoming changes again modify ffmpeg params. Sent from my ONEPLUS A6000 using Tapatalk
  5. I've bumped Unmanic to 0.0.1-beta5 This includes the following changes: Modify some historical logging of useful stats This sets us up to start adding extra info like eta during the conversion process as well as the stats mentioned below Adds new "See All Records" screen Any suggestions for stats that you would like on this screen would be great Note that due to the changes in logging, only newly converted items will show here. Old stuff wont due to missing statistics data. Sorry Create backups of settings when saving (there were some cases where the settings were invalid but still saved). This corrupted our data and made it impossible to read. So now we test prior to committing changes. FFMPEG was causing some errors on certain files. If you have noted any conversion failures in the past, can you please re-test with this version to confirm if it is now resolved or not. Log rotation If you are debugging, you are spewing a crap ton of data to the logs. This update rotates the logs at midnight every day and keeps them for 7 days. Even if you are not debugging, this is much better. The next milestone is to add extended functionality to the settings: https://github.com/Josh5/unmanic/milestone/4 This will hopefully be the last major tidy up of core functionality. I think that once this milestone is complete we can safely pull this out of beta and look at things like HW decoding and improving on data that is displayed throughout the WebUI.
  6. This is a support thread for various Unraid docker container templates. If the template has a link to another support thread, then please post questions there. Templates will be uploaded to: https://github.com/Josh5/unraid-docker-templates Templates: TeamCity Server More Info : https://www.jetbrains.com/teamcity/ Dockerhub Page : https://hub.docker.com/r/jetbrains/teamcity-server GitHub Page : https://github.com/JetBrains/teamcity-docker-server TeamCity Build Agent More Info : https://www.jetbrains.com/teamcity/ Dockerhub Page : https://hub.docker.com/r/jetbrains/teamcity-agent GitHub Page : https://github.com/JetBrains/teamcity-docker-agent
  7. Unmanic 0.0.1-beta4 is now available. This fixes the issue that you all were having regarding files that were already processed being re-added to the queue. The major change with this release is the following: Improved job pipeline. Prior to this, the workers handled everything including the file copies and post-processing after a conversion. This was not ideal. The new pipeline works as follows: File added to job queue as a "task" object Worker picks up task object Worker configures ffmpeg Worker inits ffmpeg (monitor output) On conversion complete task object is updated by the worker Worker places update task object into a new "processed" queue New "PostProcessor" class picks up task object from "processed" queue PostProcessor tests all files and carries out the file move PostProcessor adds task object data to history This now opens doors for things like modular post-processing configurations and improved logging on historical tasks (needed for upcoming "See all records" screen). Further in the background this also enables the automated build CI server that I've setup. This build server will automatically execute regression tests which should help with my development. The next milestone on my task list is 0.0.1-beta5 (https://github.com/Josh5/unmanic/milestone/3). This will include some more bugfixes and a new "See all records" screen. I set a release date for this for next Monday, but that my end up pushing out to the following week. We'll see how things go. If you have opened issues on github, please check if I have closed them. If they are now closed, could you please test that your reported bug/enhancement now works as expected. Ta
  8. Nope. That's fine. It means the bug is not in the stuff I rewrote. I'll investigate further Sent from my ONE E1003 using Tapatalk
  9. Yea. Its not feature complete yet Sent from my ONE E1003 using Tapatalk
  10. Over the past week I've been re-writing the job pipeline. (https://github.com/Josh5/unmanic/projects) This along with some better saving of data pertaining to a completed encoding task has set us up for a new screen that will give us a detailed breakdown of converted files in our history (coming later on). I'm getting ready to merge that branch tomorrow if all tests out OK. Basically it just needs to still work for processing the pending tasks as well as adding inotify tasks. The biggest change is that all file post-processing is no longer handled by the workers, but rather it is now handled by a single separate process. If anyone else wants to help run some tests on it, just change "josh5/unmanic:latest" to "josh5/unmanic:dev-jobpipeline" in your docker container configuration. I would be especially interested to know if @Derrick and @DrAg0n141 are still seeing files added to the queue again. Last weekend I setup a TeamCity CI server to handle automated checking and building of the application. This should help heaps with the development pipeline also as it means no more manually running all the tests and building/pushing docker images to dockerhub.
  11. Thanks for submitting the issue. I'm going to first see if I can complete the job pipeline rewrite this weekend: https://github.com/Josh5/unmanic/projects/3 https://github.com/Josh5/unmanic/commit/427a55b74eec0d5112d10170a4a0afe284b0dd94 This will give us much better view of each of the files as they flow through the conversion pipeline
  12. I don't believe in "tip jars" or "shouting beers" for beta applications. Thanks tho. Perhaps if you still like it when were are closer to a release. But for now you are 100% free to use this with no guilt that you owe me a cent. Sent from my ONE E1003 using Tapatalk
  13. Hey team. I'm finally coming off my 350 hour project at work. Will finally start to see more free time to work on unmanic. I appreciate the patience. Since January I've been working overtime and have not had a chance to do any of my personal projects. I have not even logged into steam for months. But finally that's coming to an end. Over the next fortnight I'll take a look at everyone's suggestions and start coming up with a plan of attack for getting unmanic to a point where its useful for everyone. Thanks again. Sent from my ONE E1003 using Tapatalk
  14. So this is a very simple breakdown of what unmanic currently does. I'm in no way an expert with ffmpeg, so I welcome anyone to come forward and tell me this is wrong and it's not working this way.... for each video stream: re-encode: bitrate: source resolution: source codec: hevc (can be configured in settings) encoder: libx265 (locked to codec at the moment) for each audio stream: (if audio channel has more than 2 channels (not stereo):) clone source audio channel (does not re-encode) copy source audio channel re-encode to stereo (with same title where possible. Eg. English -> English Stereo): bitrate: 128k codec: aac encoder: aac (if audio channel is stereo:) re-encode stereo audio channel: bitrate: 128k codec: aac encoder: aac for each subtitle stream: (if setting for stripping subtitle streams out of mux is set) do nothing (subtitles will not be cloned to destination) (else) clone source subtitle stream (does not re-encode)
  15. Not at this point in time sorry. Sent from my ONE E1003 using Tapatalk