Server Version#: 1.23.6.4881, tried 4707-4881
Player Version#: Version 4.59.2
[Transcoder] [eac3_eae @ 0x7fc6107a3b40] EAE timeout! EAE not running, or wrong folder? Could not read '/transcode/Transcode/Sessions/EasyAudioEncoder/Convert to WAV (to 8ch or less)/8b71pniessvplvwb7nhtkj2w_1135-0-0.wav'
[Transcoder] Error while decoding stream #0:1: I/O error
In transcoder settings I’ve selected /transcode for pathing, and it is indeed writeable by the docker host. Same item works in the app, but refuses to play in browser and on certain devices.
Now I can’t watch anything with EAC3 without app. Even ChromeCast won’t work.
So I’ve setup an Emby server next to my PMS meanwhile, might have to switch permanently soon and do some chargebacks on my card, at least when all Plex staff do is keeping cards close to their chest, promise to fix stuff and then years later it’s the exact same condition.
I know me and a lot of other Plex Pass owners are fed up with the way staff is handling customer requests. I’ve paid at least 300$+ in Plex Pass over the years, because I thought the developers and staff intended to do the best for the customer, my patience is gone.
If you wonder why, its the simple things like:
Plex Staff has responded multiple times that they are doing something on the matter, here we are two years later, and squat has happened.
This they actually implemented, but not to the wished spec from the customer (as always)
It was actually a request for a feature that would let you disable video transcoding unless it was 100% NECESSARRY. Not flat out disable it.
TL;DR Plex Team has their own way of doing things related to maximize their profits in other segments than PMS. This is now what the customers are experiencing first hand. It seems like Plex now want to get in the pockets of bigger media companies and rent out space for ads and implementing alternative streaming services. We’ve probably made them too greedy with all our support, and they forgot who their original customer was. Someone who wanted to have control over their own media and services.
This happens because Transcode path is not being all dynamic, specifically because the EAC3 decoder/encoder decides to put the transcode in /tmp instead of the chosen /transcode path in server settings. Please fix.
I have to say though that this strikes me as quite a glaring bug, and using the search here and on Reddit appears to go back several years.
Now I am able to understand the fix you posted, but I imagine a great many users who use Plex because of its supposed simplicity, and perhaps bought a Plex Pass, will have no idea what to do.
Even the ability to change this value in the server GUI would be a no brainer.
I am using Version 1.24.3.5033 on Synology 1821+ DSM 6.2.4-25556 Update 2 (latest update before DSM 7). The NAS has 20GB of RAM installed.
I am using the latest version of the Plex client on various Android devices (Sony Android TV, Samsung Galaxy S2 tablets, 2 x Shield TV 2017, 1 x Shield TV 2019).
Whenever there is a call for Plex to transcode EAC audio, the file will not play. I see the “EAE” error in the log as per the OP.
I have edited my sysctl.conf to read fs.inotify.max_user_watches=524288 which should be “more than enough”
I have deleted my codec folder.
Naturally I have stopped and started Plex and rebooted the entire NAS in between each troubleshooting step.
Nothing, nada.
The ONLY thing that “cures” this, bizarrely, is to UNTICK the box in ther server>settings>Library section : “Scan my library automatically”.
Restarting the Plex server allows the EAC encoded video files to play properly and be transcoded as usual.
Hmm - why is transcoding allowed when this simple checkbox is disabled?
How to I fix this, as this is far from ideal as now new shows will are obviously not showing up automatically and is causing me a lot of hassle from the family members who aren’t seeing their new shows come up unless I manually scan the library.
Should I start a new topic on this? The OP posts EXACTLY what my issue is. I just cannot fathom out why my quickfix causes it to “work”.
I cannot pinpoint it exactly but I reckon this happened 6 weeks ago or so, either down to server changes (likely I feel), or the Android app.
Regarding the Android app I’ve read in places, here included, that there is some issue with Google and exoplayer, but as I can “fix” this issue by simply unticking the “scan my library automatically” box and restarting the server then I find this (IMHO as an I.T. troubleshooter by day, on Windows though, I’m pretty green in Linux) highly unlikely.
I’ve tried to troubleshoot this as best I can, but my “solution” of not having my library scanned automatically is definitely not satisfactory.
There are quite a few other threads here, some still unresolved:
edit:
Should I be changing this??
edit again:
Success, the above fixed it. I added the entries s2s_watches_max=65536 to the two “synoinfo.conf” files, rebooted and it worked. Adding the inotify info to “sysctl.conf” does nothing on the Synology.
FWIW, I recently added some old TV stuff I had stashed outside of Plex and my folder count had only recently crept above 8192 to 9087 by the looks of it.
Still, if it looks like a bug… surely there must be an easier way around this - 8192 folders really is NOT a lot in the grand scheme of things.
May I see the DEBUG logs ZIP please – captured right after a playback attempt which fails ?
The FAQ is there because Synology doesn’t do things by Linux standard – as you learned with sysctl.conf . They do most adjustments to the kernel for DSM functionality at boot time. This is why synoinfo.conf exists.
Defo related to having over 8192 folders in the end up and was fixed by applying the “sysctl.conf” fix to “synoifo.conf”.
Was having some bizarre problems with my son playing items on his Sony Android TV last night, was working one minute then spinning circle the next, I snipped some info out the logs, but rebooting the Synology (DSM 6) fixed it (or I turned off movie trailers as some others have suggested - I’ve been doing some fancy dancing to get things working).
There were a LOT of entries relating to “We’re going to try to auto-select an audio stream” when this was happening, much more than in a “normal” playback log. Once again, I never kept a copy on that occasion, but I’ll keep an eye on it should it happen again.
Something seems up with Android playback judging by the many threads bringing this up in the last month or so.
I can’t speak to anything android except that I have the Shield Pro 2019 and it’s flawless for me (which doesn’t help).
When I see log files which have a lot of “auto-select an audio stream” it’s usually because the file contains that many streams.
I curate my media very tightly with mkvtoolnix after ripping to make sure I didn’t accidentally get something I didn’t see during the rip
The Notify-Watch directories are critical for anything involving Dolby because of the Dolby decoder (Licensing). The only way to utilize it is to place a watch_notify on the output directory. For this reason, the transcoder needs 2 slots per active Dolby audio stream
