@ChuckPa
Quick test with one problem movie 1.30.2.6563 fails works with 1.31.1.6617
edit: Tested a few more “Video: hevc (Main 10), yuv420p(tv” and they now transcode fine
The transcoder settings are the same as in the previous releases; makes me think there’s actually a fix/changes in ffmpeg done?
They are still using the depreciated hq preset, see the difference and why px -tune hq was introduced preset values the default p4 -tune hq is more effective then the old hq
Looks OK so far with my initial testing, this 1.31.2.6715 release seems to be as-good or better as the previous version I was able to get working, 1.31.1.6617. I haven’t seen any files fail to play so far (HEVC and MPEG4 both appear fine).
@ChuckPa
Was able to install and test 1.31.2.6715 8bit hevc hw transcoding seems to be working properly. Seeing VAAPI errors, but hw transcoding works.
~]$ ps aux |grep "jelly" | grep -v grep; nvidia-smi; tail -10 /var/lib/plexmediaserver/Library/Application\ Support/Plex\ Media\ Server/Logs/Plex\ Media\ Server.log
plex 1580950 27.0 1.2 5340384 396048 ? Sl 13:52 0:00 /usr/lib/plexmediaserver/Plex Transcoder -codec:0 hevc -hwaccel:0 nvdec -hwaccel_fallback_threshold:0 10 -threads:0 1 -hwaccel_output_format:0 cuda -hwaccel_device:0 cuda -analyzeduration 20000000 -probesize 20000000 -i /srv/ftp/movies/test/jellyfish-30-mbps-hd-hevc.mkv -filter_complex [0:0]hwupload[0];[0]scale_cuda=w=1920:h=1080:format=nv12[1] -map [1] -codec:0 h264_nvenc -b:0 8514k -maxrate:0 11352k -bufsize:0 22704k -preset:0 hq -forced-idr:0 1 -r:0 29.969999999999999 -force_key_frames:0 expr:gte(t,n_forced*1) -f dash -seg_duration 1 -dash_segment_type mp4 -init_seg_name init-stream$RepresentationID$.m4s -media_seg_name chunk-stream$RepresentationID$-$Number%05d$.m4s -window_size 5 -delete_removed false -skip_to_segment 1 -time_delta 0.0625 -manifest_name http://127.0.0.1:32400/video/:/transcode/session/l32hgutrx5d6amo58pajpsij/0606c051-857e-4b54-8e8c-3fe5fe5cbc92/manifest?X-Plex-Http-Pipeline=infinite -avoid_negative_ts disabled -map_metadata -1 -map_chapters -1 dash -start_at_zero -copyts -vsync cfr -init_hw_device cuda=cuda: -filter_hw_device cuda -y -nostats -loglevel quiet -loglevel_plex error -progressurl http://127.0.0.1:32400/video/:/transcode/session/l32hgutrx5d6amo58pajpsij/0606c051-857e-4b54-8e8c-3fe5fe5cbc92/progress
Fri Feb 17 13:52:15 2023
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.89.02 Driver Version: 525.89.02 CUDA Version: 12.0 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 NVIDIA GeForce ... Off | 00000000:01:00.0 Off | N/A |
| 0% 48C P0 45W / 130W | 332MiB / 6144MiB | 2% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| 0 N/A N/A 1580950 C ...diaserver/Plex Transcoder 328MiB |
+-----------------------------------------------------------------------------+
Feb 17, 2023 13:51:25.620 [0x7fa8da127b38] WARN - [Req#382] QueryParser: Invalid field 'contentDirectoryID' found, ignoring.
Feb 17, 2023 13:51:43.076 [0x7fa8da127b38] ERROR - [Req#390/Transcode] [FFMPEG] - Failed to initialise VAAPI connection: -1 (unknown libva error).
Feb 17, 2023 13:51:43.237 [0x7fa8da127b38] ERROR - [Req#390/Transcode] [FFMPEG] - Failed to initialise VAAPI connection: -1 (unknown libva error).
Feb 17, 2023 13:51:45.870 [0x7fa8dadc4b38] WARN - [Req#3bb/Transcode] Got a transcode session ping without a valid session ID.
Feb 17, 2023 13:52:11.863 [0x7fa8dadc4b38] ERROR - [Req#445/Transcode] [FFMPEG] - Failed to initialise VAAPI connection: -1 (unknown libva error).
Feb 17, 2023 13:52:12.027 [0x7fa8dadc4b38] ERROR - [Req#445/Transcode] [FFMPEG] - Failed to initialise VAAPI connection: -1 (unknown libva error).
Feb 17, 2023 13:52:12.193 [0x7fa8dadc4b38] ERROR - [Req#445/Transcode] [FFMPEG] - Failed to initialise VAAPI connection: -1 (unknown libva error).
Feb 17, 2023 13:52:12.351 [0x7fa8dadc4b38] ERROR - [Req#445/Transcode] [FFMPEG] - Failed to initialise VAAPI connection: -1 (unknown libva error).
Feb 17, 2023 13:52:12.512 [0x7fa8dadc4b38] ERROR - [Req#445/Transcode] [FFMPEG] - Failed to initialise VAAPI connection: -1 (unknown libva error).
Feb 17, 2023 13:52:12.693 [0x7fa8dadc4b38] ERROR - [Req#445/Transcode] [FFMPEG] - Failed to initialise VAAPI connection: -1 (unknown libva error).
Hi @ChuckPa! Assuming this indeed is the fix whats the best way to know what publicly available version has it so I know when to update? Should I look for something specific in the release notes? Thanks for all the hard work and I hope you feel better!
I tried version 1.31.1.6617 and nearly all of my hw transcoding issues went away, the only situation I could break it - was with a 4K HDR hevc file and Chrome. It firstly would direct play (which is crazy but nice) and then I changed the playback quality to 1080p@12mbit. It loads for a second but nothing else happens. The same thing in Edge is different, Edge does not support HEVC (didn´t install the hevc windows extension) and the file would start directly as a transcoded stream just fine.
I will try the new 1.31.2.6715 version, maybe it fixes the chrome behavior.
I just got one question while reading through the thread. It’s regarding server hardware. I´m somewhat confused. Does Plex scan if the hardware is enterprise-grade? And if it is, it refuses to utilize it? Would be interesting, as I´m using an “old” Dell R730XD server and didn´t experience any “forced” slowdowns and also never ever read something in this regard. Sorry if this is off-topic…
No… Plex doesn’t scan for hardware ‘grade’. Either the hardware is capable or it isn’t.
If you use old CPUs – expect to suffer the CPU performance.
This has been a tough problem to identify and I don’t know if Engineering really understands it either because there is no single “smoking gun” which everyone agrees on.
@Mitzsch are you also able to change quality in Chrome during playback with PMS running on enterprise equipment using Nvidia, and does playback keep going after it switches quality ?
Changing the quality in Chrome does not work for me with any version of PMS running on any enterprise hardware with Nvidia GPU, I have tested so far. While it works just fine when PMS is running on a standard consumer level hardware with the same GPU. I have no idea why, or what could cause such a weird behaviour. It’s incredibly weird. But I know of several other people having the same kind of issue.
@ChuckPa I have the same issue with Chrome on the new engineering build as well.
