Subtitles delaied or not loaded, language issues, High CPU usage and play stopped

[alexconte]

Hi,

I'm Alex and as I didn't find answer to my problems in the forum or discord, I decided to submit a post.
First of all many thanks to offer this system, I really enjoy using it.
I'm keen to do any test to solve my issues!

Version 10.10.7
Docker on Ubuntu 20.04.6 LTS
DELL T610
CPU : 2 X Intel® Xeon® CPU  X5650  @ 2.67GHz
RAM : 2 X DIMM DDR3 Synchronous Registered (Buffered) 1333 MHz (0,8 ns)
GPU : MGA G200eW WPCM450 and GK208B [GeForce GT 710]
HDD : 8TB PERC H700 (3 X 4TO HDD in RAID 5)
Running since july 2025 behind Nginx Proxy Manager

Issues :
From the 1st day of using it
- Subtitles in advance only on smartphones 1 to 5 seconds
- Subtitles not loading on smartphones and TV (Samsung, app installed with the tyzen docker developer mode)
- Wrong language and not possible to load the good one on smartphones and TV. It may takes 3 app closing and reopening to fix

New issue since last week
- Video stops without error after few minutes on every client (web, smartphones, TV). CPU was at 1150% according to my docker manager ARCANE.
I tried several setup on transcode without success. What I should do next may be using the Geforce card but it ha sworked pretty well before so I'm afraid to do worst than better ^^

I attached my docker compose (sanitized) and the logs, can you help me?

Many tks

[34626]

Since you are using docker, please read this, mostly the status of Nvidia GPU on Linux:
https://jellyfin.org/docs/general/post-i...on/nvidia/

You have set a hardware acceleration profile in Jellyfin? (menu > dashboard > playback)

[alexconte]

I have read this doc and it may be the next step but my graphic card doesn't support NVENC so no sure it will be better :
nvidia-smi
Fri Oct 31 11:51:59 2025
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 470.256.02  Driver Version: 470.256.02  CUDA Version: 11.4    |
|-------------------------------+----------------------+----------------------+
| 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  | 00000000:06:00.0 N/A |                  N/A |
| N/A  34C    P8    N/A /  N/A |      4MiB /  2002MiB |    N/A      Default |
|                              |                      |                  N/A |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU  GI  CI        PID  Type  Process name                  GPU Memory |
|        ID  ID                                                  Usage      |
|=============================================================================|
|  No running processes found                                                |
+-----------------------------------------------------------------------------+

No hardware acceleration set, I tried but it doesn't work.

[34626]

The hardware is very old, it looks like the best performance is to disable hardware acceleration, threads to auto and if you have plenty of RAM, you can change the transcode path into a ram disk.

Im curious on what clients your are running?
What if you use Jellyfin Media Player on Linux, MacOS or Windows? (dpends on what you have)

If you have a Android smartphone or tablet, does it also stop with Findroid?

Only thing i else notice in your docker compose.yaml is "jellyfin/jellyfin" which you should set for the specific version or latest like "jellyfin/jellyfin:10.10.7" or "10.11.1" for current latest..