Hey all, I’m wondering if Dynamic Audio and video are possible. I already know that there is an “Optimize” feature, and it will optimize the video and audio for the intended playback device, however, I was wondering if a far better solution would be to put in multiple video files/resolutions, and multiple audio codecs, and then let the server decide what is best for each users specific needs.
ie, lets say I want a video in my library to be at 4K HDR. Rather than relying on my PC to transcode that down to 1080 or 720 for users with TV’s that don’t support the 4K HDR content, I was considering putting in a 1080 version of the file, and a 720 version of the title. Then, when the client app makes the request, it grabs the stream most appropriate for it. This could also be done with Audio and Subtitles, so that transcoding of audio almost never happens, creating an entire library of “Direct Play” content rather than Direct Stream or Transcode…
Does this, or could this work? It’s something I’ve been mulling over, and I don’t see why it couldn’t… Plex Server already knows when content cannot be played on the other device, and subsequently transcodes what doesn’t play natively, so why couldn’t it instead just select the more approriate stream from a list of available streams?
There’s a snowballs chance in hell for you to be able to hit every possibility, but your only option, AFAIK, is to optimize for every single resolution, that your users need
lol, no, I get that… That would entail 20 different video options, and 50 different audio options, etc. But, if I could hit the most common, with say 3-4 different resolutions/codecs, do the same for the audio, and the subtitles, it wouldn’t be bad, and it would allow the server to transcode much less, saving power in the long run…
Select the library, then on top right, click on the 3 dots
There should be an option to optimize
Take care though, since that would drain you server until done!
Ref: Creating Optimized Versions | Plex Support
And yes, above talks about an individual movie/show
Dane, I think you’re missing the initial question. I already know about optimized versions… Instead of relying on Plex to transcode and create an optimized version for each device, what I would rather happen is for me to just proactively put in 4 videos
2160p HDR
2160p (non-HDR)
1080p
720p
None of which contain audio tracks, they are simply video streams.
Then say 5 Audio files
ATMOS
DTS:X
DTS
5.1 AC3
Stereo MP3
And then 3 subtitle formats
.srt
.ass
.pgs
Then, the server can intelligently decide which streams it can send to the recipient, starting in order of highest quality to lowest, while still maintaining direct play functionality.
I do know that, but if you put in the medias your self, so they act as duplicates, PMS would select one, and unsure here, but suspect it’ll simply select the first on the list
Ok, so now we’re on the same page. See, I am unsure of how it selects too, that’s why I tried to ask if it already selects based on what the client can play (Cause that’s sort of how Optimized works), or if it COULD be written in this manner. It would be very good functionality imo, especially for those of us having 15-20 transcodes going on at once… That’s a lot of power being sucked up by the processors, and could be avoided with a little pre-planning…
Sadly got a lot on my plate here already, so sadly all I can say is that this is uncovered territory.
I suggest you test, and report back, and I’ll in the mean time ask Plex.
And Plex responded with: often clients make that decision themselves; usually I'd expect them to play the higher-resolution version (which is also what PMS would prefer)
Right , but that’s only 1/2 the answer. I thank you for your chat and help by the way, I appreciate it… I guess the clarification I would need is: Will PMS play the 4K and transcode down to 1080 instead of playing the 1080 natively since the client side doesn’t have a 4K TV?, is really the question. Like what criteria does it use? If I could design it, I’d ask it to look at the client, see what it can play, and then choose the stream that closest resembles what it can play…
And adding here, after been rethinking for a while…
Your idea is brilliant, but the use case almost close to none, since what people have, are container based medias, that contains both movie and audio, and sometime subs.
For this to work, users (or PMS) would have to split it out, during like an optimize, followed by some algorithm to combine the relevant parts again, and send to the player, based on a decision engine.
To be fair, it could be accomplished in container based medias as well, MKV can have multiple video tracks if I recall correctly. But de-muxing the tracks into a single folder makes it easier since .mkv is not supported by all devices…
