Not sure, a lot of that stuff got rolled into docker by default. But I think you still need an image that's based off of https://hub.docker.com/r/nvidia/cuda
EDIT: I may be wrong. Maybe all you need is --runtime nvidia and it will share the libcuda into the container
OK, so I've built a version of ah4c with vaapi support. It was pushed this afternoon as bnhf/ah4c:test. If anyone is dying to try using a Zinwell NextGen TV box with ah4c and an encoder, and has a decent Docker host with an iGPU, it should now be possible to run the encoder output through ffmpeg with hardware accleration.
The idea would be to compensate for the ill effects of HDR coming out of the Zinwell box and passing through the encoder. ffmpeg would hopefully be able to either compensate through tone mapping or tag fixes if the HDR data is still there, but flags are missing. Running ffprobe on the stream as it exits the encoder might tell the story.
So I got nvenc working, not sure how but now I need the ffmpeg command to convert hdr to sdr. thanks
ffprobe output:
Input #0, mpegts, from 'http://192.168.1.70:8090/stream0':
Duration: N/A, start: 78776.830667, bitrate: N/A
Program 1
Metadata:
service_name : HDMI-1
service_provider: Server
Stream #0:0[0x64]: Video: h264 (High) ([27][0][0][0] / 0x001B), yuvj420p(pc, bt470bg/bt470bg/bt709, progressive), 1920x1080, 60 fps, 120 tbr, 90k tbn
Stream #0:1[0x65]: Audio: aac (LC) ([15][0][0][0] / 0x000F), 48000 Hz, stereo, fltp, 130 kb/s
a little more specific info:
ffprobe -v quiet -show_streams -select_streams v:0 http://192.168.1.70:8090/stream0
[STREAM]
index=0
codec_name=h264
codec_long_name=H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10
profile=High
codec_type=video
codec_tag_string=[27][0][0][0]
codec_tag=0x001b
width=1920
height=1080
coded_width=1920
coded_height=1080
closed_captions=0
film_grain=0
has_b_frames=0
sample_aspect_ratio=N/A
display_aspect_ratio=N/A
pix_fmt=yuvj420p
level=41
color_range=pc
color_space=bt470bg
color_transfer=bt709
color_primaries=bt470bg
chroma_location=left
field_order=progressive
refs=1
is_avc=false
nal_length_size=0
ts_id=1
ts_packetsize=188
id=0x64
r_frame_rate=120/1
avg_frame_rate=60/1
time_base=1/90000
start_pts=7592831999
start_time=84364.799989
duration_ts=N/A
duration=N/A
bit_rate=N/A
max_bit_rate=N/A
bits_per_raw_sample=8
nb_frames=N/A
nb_read_frames=N/A
nb_read_packets=N/A
extradata_size=38
DISPOSITION:default=0
DISPOSITION:dub=0
DISPOSITION:original=0
DISPOSITION:comment=0
DISPOSITION:lyrics=0
DISPOSITION:karaoke=0
DISPOSITION:forced=0
DISPOSITION:hearing_impaired=0
DISPOSITION:visual_impaired=0
DISPOSITION:clean_effects=0
DISPOSITION:attached_pic=0
DISPOSITION:timed_thumbnails=0
DISPOSITION:non_diegetic=0
DISPOSITION:captions=0
DISPOSITION:descriptions=0
DISPOSITION:metadata=0
DISPOSITION:dependent=0
DISPOSITION:still_image=0
DISPOSITION:multilayer=0
[/STREAM]
edit:
I tweaked the colors a little and it looks nice
my stack looks like
services:
ah4c: # This docker-compose typically requires no editing. Use the Environment variables section of Portainer to set your values.
# 2025.09.13
# GitHub home for this project: https://github.com/bnhf/ah4c.
# Docker container home for this project with setup instructions: https://hub.docker.com/r/bnhf/ah4c.
image: bnhf/ah4c:${TAG:-latest}
container_name: ${CONTAINER_NAME:-ah4c}
hostname: ${HOSTNAME:-ah4c}
dns_search: ${DOMAIN:-localdomain} # Specify the name of your LAN's domain, usually local or localdomain
ports:
- ${ADBS_PORT:-5037}:5037 # Port used by adb-server
- ${HOST_PORT:-7654}:7654 # Port used by this ah4c proxy
- ${SCRC_PORT:-7655}:8000 # Port used by ws-scrcpy
environment:
- IPADDRESS=192.168.1.28:7654 # Hostname or IP address of this ah4c extension to be used in M3U file (also add port number if not in M3U)
- NUMBER_TUNERS=5 # Number of tuners you'd like defined 1, 2, 3 or 4 supported
- TUNER1_IP=192.168.1.71 # Streaming device #1 with adb port in the form hostname:port or ip:port
- TUNER2_IP=192.168.1.72 # Streaming device #2 with adb port in the form hostname:port or ip:port
- TUNER3_IP=192.168.1.73 # Streaming device #3 with adb port in the form hostname:port or ip:port
- TUNER4_IP=192.168.1.74 # Streaming device #4 with adb port in the form hostname:port or ip:port
- TUNER5_IP=192.168.1.75 # Streaming device #5 with adb port in the form hostname:port or ip:port
- ENCODER1_URL=http://192.168.1.70:8090/stream0 # Full URL for tuner #1 in the form http://hostname/stream or http://ip/stream
- ENCODER2_URL=http://192.168.1.70:8090/stream1 # Full URL for tuner #2 in the form http://hostname/stream or http://ip/stream
- ENCODER3_URL=http://192.168.1.70:8090/stream2 # Full URL for tuner #3 in the form http://hostname/stream or http://ip/stream
- ENCODER4_URL=http://192.168.1.70:8090/stream3 # Full URL for tuner #4 in the form http://hostname/stream or http://ip/stream
- ENCODER5_URL=http://192.168.1.70:8090/stream4 # Full URL for tuner #5 in the form http://hostname/stream or http://ip/stream
- CMD1=ffmpeg -i http://192.168.1.70:8090/stream0 -c:v h264_nvenc -preset hq -vf eq=brightness=0.2:contrast=1.5:saturation=2,unsharp=5:5:1.0:5:5:0.0 -pix_fmt yuv420p -b:v 70M -minrate 50M -maxrate 95M -bufsize 150M -c:a copy -f mpegts - # Typically used for ffmpeg processing of a device's stream
- CMD2=ffmpeg -i http://192.168.1.70:8090/stream1 -c:v h264_nvenc -preset hq -vf eq=brightness=0.2:contrast=1.5:saturation=2,unsharp=5:5:1.0:5:5:0.0 -pix_fmt yuv420p -b:v 70M -minrate 50M -maxrate 95M -bufsize 150M -c:a copy -f mpegts - # Typically used for ffmpeg processing of a device's stream
- CMD3=ffmpeg -i http://192.168.1.70:8090/stream2 -c:v h264_nvenc -preset hq -vf eq=brightness=0.2:contrast=1.5:saturation=2,unsharp=5:5:1.0:5:5:0.0 -pix_fmt yuv420p -b:v 70M -minrate 50M -maxrate 95M -bufsize 150M -c:a copy -f mpegts - # Typically used for ffmpeg processing of a device's stream
- CMD4=ffmpeg -i http://192.168.1.70:8090/stream3 -c:v h264_nvenc -preset hq -vf eq=brightness=0.2:contrast=1.5:saturation=2,unsharp=5:5:1.0:5:5:0.0 -pix_fmt yuv420p -b:v 70M -minrate 50M -maxrate 95M -bufsize 150M -c:a copy -f mpegts - # Typically used for ffmpeg processing of a device's stream
- CMD5=ffmpeg -i http://192.168.1.70:8090/stream4 -c:v h264_nvenc -preset hq -vf eq=brightness=0.2:contrast=1.5:saturation=2,unsharp=5:5:1.0:5:5:0.0 -pix_fmt yuv420p -b:v 70M -minrate 50M -maxrate 95M -bufsize 150M -c:a copy -f mpegts - # Typically used for ffmpeg processing of a device's stream
- STREAMER_APP=scripts/zinwell/livetv # Streaming device name and streaming app you're using in the form scripts/streamer/app (use lowercase with slashes between as shown)
- CHANNELSIP=192.168.1.28 # Hostname or IP address of the Channels DVR server itself
- UPDATE_SCRIPTS=true
- CREATE_M3US=false # Set to true to create device-specific M3Us for use with Amazon Prime Premium channels -- requires a FireTV device
- TZ=US/Eastern # Your local timezone in Linux "tz" format
- HOST_DIR=/data
# - NVIDIA_VISIBLE_DEVICES=all
# - NVIDIA_DRIVER_CAPABILITIES=all
restart: unless-stopped
# deploy:
# resources:
# reservations:
# devices:
# - driver: nvidia
# count: 1
# capabilities: [gpu]
volumes:
- ${HOST_DIR:-/data}/ah4c/scripts:/opt/scripts # pre/stop/bmitune.sh scripts will be stored in this bound host directory under streamer/app
- ${HOST_DIR:-/data}/ah4c/m3u:/opt/m3u # m3u files will be stored here and hosted at http://<hostname or ip>:7654/m3u for use in Channels DVR - Custom Channels settings
- ${HOST_DIR:-/data}/ah4c/adb:/root/.android # Persistent data directory for adb keys
# Default Environment variables can be found below under stderr -- copy and paste into Portainer-Stacks Environment variables section in Advanced mode
afterwords I have to go into containers/ah4c/edit and set the following
Any feel for the CPU/GPU requirements for this? Any percentage used stats or the like would be nice to know.
Also, for anybody else looking at this, the ffmpeg command shown is use with Nvidia only when the GPU is passed through.
This stack is obviously specific to @techpro2004, for anybody else I recommend using the standard ah4c stack (soon to be updated to include CMDx env vars). Also, as previously mentioned, most all of the Docker Compose YAMLs I've created are designed to be used WITHOUT hard coding anything into the stack itself.
Portainer-Stacks has a separate portion of the editor (use the Environment variables dropdown, and then select Advanced mode), which allows you to decouple the compose from the values that are specific to your installation. This way when a Docker Compose is updated, it's typically a drop-in that uses your existing list of env vars.
Nice to see we solved the ATSC 3 DRM problem. Hopefully, we'll get a network tuner on the way too (or hope and pray that DRM is outlawed)
How do I get the tonemap working with ffmpeg via software rendering. I'm running this on a qnap ts-464 but am not certain how to run ffmpeg with hardware acceleration with docker. I keep getting errors like unsharp not compatible with auto scale.
I am trying to get it working with my uray encoder.
My advice would be to try a basic ffmpeg command and then work your way to the more complex. fyi, I am using an nvidia gpu not the intel one in your qnap. You could also try the above ffmpeg comand I posted in my stack.
edit: make sure to replace h264_nvenc with h264_qsv
I can get the simple ffmpeg working but when I try to use filters with copy it doesn't load with errors like streamcopy and filtering not compatible. Furthermore when I try multiple options I get eq as an invalid input for libx264 and zscale not valid either. I tried your command and it doesn't work with h264_qsv.
Did you pass through your gpu to the container?
edit: did you remember the -vf before eq
I got a command working after trial and error now I just have to figure out how to tonemap.
ffmpeg -init_hw_device vaapi=intel:/dev/dri/renderD128 -hwaccel vaapi -hwaccel_output_format vaapi -hwaccel_device intel -filter_hw_device intel -i srt://192.168.1.141:9000 -vf format=nv12,hwupload -c:v h264_vaapi -c:a copy -f mpegts -
Very good. I had to play around with it a little to get it working also. Keep up the good work.
Is this error concerning.
Failed setup for format vaapi: hwaccel initialisation returned error.
[Parsed_tonemap_vaapi_2 @ 0x563514d21080] VAAPI driver doesn't support HDR
[Parsed_tonemap_vaapi_2 @ 0x563514d21080] Failed to configure output pad on Parsed_tonemap_vaapi_2
EDIT: not a 12th gen intel misread that. These are what I get with vainfo
ibva info: VA-API version 1.17.0
libva info: User environment variable requested driver 'iHD'
libva info: Trying to open /usr/lib/x86_64-linux-gnu/dri/iHD_drv_video.so
libva info: Found init function __vaDriverInit_1_17
libva info: va_openDriver() returns 0
vainfo: VA-API version: 1.17 (libva 2.12.0)
vainfo: Driver version: Intel iHD driver for Intel(R) Gen Graphics - 23.1.1 ()
vainfo: Supported profile and entrypoints
you have a Intel Celeron N5095 which is jasper lake from 2021. According to google it supports hdr but you may have to use quicksync instead of vaapi. I apologize but I am not an expert on intel gpus (I always use nvidia) or qnaps Please google how to get quicksync working in container station on a qnap.
I got qsv working with this command
ffmpeg -init_hw_device qsv=intel:/dev/dri/renderD128 -hwaccel qsv -hwaccel_output_format qsv -hwaccel_device intel -filter_hw_device intel -i srt://192.168.1.141:9000 -vf vpp_qsv,tonemap=hable,format=nv12 -c:v h264_qsv -c:a copy -f mpegts -
There are still 2 issues.
It seems to only playback 4 seconds before it lags and then eventually timeouts after 16 seconds.
Its still a little too dark.
I figured out this command works.
ffmpeg -init_hw_device qsv=intel:/dev/dri/renderD128 -hwaccel qsv -hwaccel_output_format qsv -hwaccel_device intel -filter_hw_device intel -i srt://192.168.1.141:9000 -vf eq=brightness=0.2:contrast=1.5:saturation=2,unsharp=5:5:1.0:5:5:0.0,format=yuv420p -c:v h264_qsv -c:a copy -f mpegts -
I don't think I can do anything further because the bitrate seems to be very low unless theres a way to increase the bitrate of the actual conversion.
-b:v 4000k
