well from my understanding the problem with vc-1 is that is single threaded, so I would assume that it is the individual core passmark that matters, not the aggregate, for vc1.
single core 1350 really isn’t much for todays cpu, but could be an issue if you are reusing old hardware as a plex server.
https://www.cpubenchmark.net/singleThread.html
rounding to an even 1400 @ Intel Core i5-2450M @ 2.50GHz 1,400 $88.02*
minimum 4k passmark is still up in the air, as far as I know a safe minimum would be ~14000 passmark > per 4k transcode.
Thanks Otto for the reply. I guess my question boils down to what bitrate was used for that test that calculated it at 1350? Given how poorly transcoding performs on a cpu with 1290 single thread passmark, it is hard to imagine that 1350 is the magic number for transcoding high bitrate streams. To state my question another way - how confident should I be that a cpu with 1518 single thread performance can transcode a 1:1 rip of a high quality blueray (e.g., LOTR or Star Wars) with the client set to max quality?
You can only try it yourself. The sample pool of test files was pretty small.
x5687 vs x 5690 are close enough together that it is probably not gonna make that much of a difference regarding vc1, I’d go for the bigger total since that will help out more with the more common multithreaded stuff.
either way, if you continue to have problems perhaps you should try the plex ‘optimize’ function to pre-transcode problem files (it keeps the original, creates an extra smaller more compatible version).
if you are not direct playing, it doesn’t really matter what the transcoded quality is as long as it is watchable, eh?
you could also try setting the plex transcoder quality to ‘prefer higher speed encoding’
