TranscodingOwl

Autotranscoder for TV Shows using FFMPEG with HW-Accel.
All encoders use constant quality. Supported are nvidia, amd and cpu. It will only show you avaible hardware!

You´ll need to place ffmpeg binary into ./bin/ffmpeg GLOBAL PATH ISN`T SUPPORTED!

Features

Recursive scanning of your media library (Around 8000 Files per second on my HDD array)
Simple CLI based user interface
Tracks already encoded files, so you don´t have to
Parallel file processing
Shows progress for each file in seconds
Logs all actions
Shows starting and final size at the end

Well tested!

This software already transcoded over 25TB of media files for me.

Configuration

The config file can be found in ./config/encoders.js.
The software will check if the encoder is avaible, if you got multiple encoders of the same type look at Use multiple GPUs.

Known issues

If the application crashes or is forced to exit, please make sure no ffmpeg process is alive before you restart it.

Notifications

You can use the -n or --notify flag to provide a webhook url thats called when the encoding is finished.
This string supports placeholders:

{{totalSizeBefore}} - Total size of all files before encoding
{{totalSizeAfter}} - Total size of all files after encoding
{{saved}} - Saved space in human readable format
{{saved_percent}} - Saved space in percent
{{hostname}} - Hostname of the machine
{{path}} - Path to the media folder

You can also use the -t or --tasknotify flag to provide a webhook url thats called when a task is finished.
This string supports placeholders:

{{task}} - Name of the task
{{hostname}} - Hostname of the machine
{{original_size}} - Original size of the file
{{new_size}} - New size of the file
{{saved}} - Saved space in human readable format
{{saved_percent}} - Saved space in percent
{{encoder}} - Encoder used
{{encoder_name}} - Name of the encoder used
{{encoder_quality}} - Quality of the encoder used
{{encoder_preset}} - Preset of the encoder used

Example:

node .\index.js -i "<<Path to mediafiles>>" -r -n "https://api.telegram.org/botxxxxxxx:xxxxxxxxx/sendMessage?chat_id=-xxxxxxx&text=Transcoding of '{{path}}' finished on {{hostname}} saved {{saved}}" -t "https://api.telegram.org/botxxxxxxx:xxxxxxxxx/sendMessage?chat_id=-xxxxxxx&text=Finished {{task}} on {{hostname}} {{original_size}} -> {{new_size}} ({{saved}})"

Use multiple GPUs or use the ability to do multiple streams at once

It can´t dedect multiple GPUs, so you´ll need to set it manually in the config.json
Just add the GPUs you want to use to the array.

"hevc_nvenc": {
        "cname": "hevc_nvenc",
        "name": "hevc_nvenc",
        "type": "GPU",
        "special_params": "-hwaccel 0",
        "quality": "vbr",
        "quality_command": "-rc:v vbr",
        "quality_constract": ["-cq"],
        "quality_preset_command": "-preset",
        "quality_preset_constract": ["fast", "medium", "slow"]
    },
    "hevc_nvenc_2": {
        "cname": "hevc_nvenc_2",
        "name": "hevc_nvenc",
        "type": "GPU",
        "special_params": "-hwaccel 0",
        "quality": "vbr",
        "quality_command": "-rc:v vbr",
        "quality_constract": ["-cq"],
        "quality_preset_command": "-preset",
        "quality_preset_constract": ["fast", "medium", "slow"]
    },