Universal Media Server | Forum

Personally I recommend using x264 https://www.videolan.org/developers/x264.html since it has good defaults - you can simply set a CRF (Constant Rate Factor) and leave everything at its default settings, and you'll get a high quality file. For 1080p I'd go with CRF 18-22, with 18 being highest qualit...

Please attach a debug log (instructions at viewtopic.php?f=9&t=556) after adding those two lines

I think the PS4 only supports up to level 4.2 H.264, which means that at 1080p (see reference https://planetcalc.com/3321/), files are only allowed 4 reference frames. The "bad" file is using 6 reference frames, and the person who encoded it seems to have manually set the profile flag to "High L4.1"...

Also you said 5.4.0 works, what about 5.5.0 and 6.0.0? I'm trying to determine the last version that works so we can narrow down which change broke the support for your TV

Yes that would be ideal, then we can see exactly what the difference is between what is being sent to the TV

Hi, sorry for the trouble. It would be really useful to get a trace log from the version that works, and one for 9.0.0-b2, so we can compare them and see what has changed

Make sure to enable incoming connections with your computer firewall, that is most likely the reason for this

I rebroke the path tried the old ums.jar and after seeing it not run replaced it with the new ums-9.0.0-b2-1d7b5529a. It works now with the broken path. Thanks so much for all your testing, I'll look to merge that fix and hopefully that can resolve the issue for more users And thanks for all the he...

It would be helpful if you can try to set an invalid path since I won't be at my Windows machine for a few days and it would be nice to get this wrapped up before then. Here's an attempted fix which hopefully skips over invalid sections of the path: https://www.universalmediaserver.com/uploads/ums-9...

Very nice, thanks both of you for the help and testing. Considering that it used to work, and that we seem to have a lot of people encountering the same problem, I am still interested in finding a way to support the invalid path. If I make a proposed fix, can you please break your path again and tes...