My sonarr has handled disenchantment fine, the files it grabbed are s5exx but they imported automatically as s3exx just fine. I didn’t have to do anything.
i don’t know too much about jellyfin compatibility, but your issue is probably ASS/SSA format subtitles, which are a complex graphical overlay subtitle, rather than simple text so device support is less guaranteed.
You could try pre-burning the subtitles in before hand by re-encoding the video, or find content with SRT subtitles which have wider support.
ASS on android clients is mostly a solved issue these days so as the other commenter suggests, the google TV chromecast with a native jellyfin app would be a more reliable solution for you.
That will be down to your GPU having hardware acceleration for certain codecs and not others, because a pure CPU encode of those codecs on anything but an Epyc or other ball tot he wall top end CPU is going to take hours.
AV1 is only just appearing in TV chipsets, and software support and stability will lag behind there for some time. if you only use your videos on a modern PC or a new-ish phone, then sure go for it, it’s pretty great.
Personally though, as good as AV1 is, I’d be avoiding it for something like a plex/emby/jellyfin library purely because while computers and phones now have decent support, many TVs and streaming boxes do not, the software on those that do is lacking support or is patchy, broken or unstable, and you can run into difficulty even transcoding those files for playback on unsupported devices because of the transcoder backends having their own support problems depending on your server hardware, operating system, and server software choice.
H265 10bit is the current best for those sorts of media libraries, just about any TV or streaming box from the last 5 years will support it just fine and it is still somewhat easy to encode with hardware acceleration.
I recently ran some of my less critical libraries through fileflows to convert a small subset of oversized H264 files to H265 10bit and with roughly 17000 files processed in those libraries I’ve saved about 5tb. that is skipping small H264 files and files already in H265, and has a few encoding tiers based on file size and some handling of reprocessing outputs that end up larger than the original, which can happen with the lower RF values that I am using.
Output quality has been perfectly acceptable, but i still have many thousands of files that I would prefer to keep in the highest quality possible, regardless of file size.
Just a tip for those of you that do cave to pressure and go with the paid premium option. or already have one but dont want to pay for more accounts or a family account.
You can set up channels or brand accounts as sub-accounts on a premium subscription and they will act like separate accounts with the advantages of premium, so if you have a large family and don’t want to pay for the full family subscription (which only has 5 slots anyway) you can set up a few sub-accounts that each get their own subscriptions, recommendations, settings and all have the premium features.
So if you want to make a premium account for a parent or child, you can do that with one single subscription if you can take the caveat of them being brand accounts rather than fully their own thing.
This works on things like Android TV or Google TV, but you need to log into the main account then switch to a sub account in the app, however, there is no authentication to switch between channel accounts this way, so it’s really only useable for families only. I use this at home to run 4 separate nvidia shield youtube apps with their own subscriptions and recommendations on one single premium payment.
I expect they will change how that works in the future to remove the loophole, probably by charging for channel accounts or having it locked behind some kind of overpriced professional usage tier, but for now, it might be a good option for some.
I dont use Tdarr because of its lack of more complex rules, but I do use fileflows to re-encode old videos on my server based on some rules considering its overall filesize, current format, and which library it is in. If the flow decides the file should be encoded it is converted to h265 10bit at a high bitrate, if it somehow ends up bigger than the original it does it again with a higher reduction factor.
The unpack process is usually single threaded, and a stock 5820k is slower than a ryzen 1600 in that case, so you’re probably just running out of CPU performance. My server runs a 6700k and it’s pretty slow to unpack even with nvme arrays