Modular Tdarr Flows

The idea is that the entire system is 'modular.' That meaning that I can tweak a handful of library variables to best suit the media I want to transcode. For example, some pieces of media I might only want to clean the subtitles on and that's it, and others I might want to go ham on and run a full transcode. A few notes:

• I separate my libraries out per TV show. Some shows I don't want to transcode because I have great downloads for, and others I want a different quality level. i.e RuPaul's Drag Race is going to get a 24 ql vs Arcane getting a 19
• The end goal is to have my entire library be x265. I don't care what Trash says- 1080p h265 looks fine.
• I have 15 users on my plex server all streaming from a range of different devices, so compatibility is key. For example, each audio transcode gets a 2ch AAC version.
• The server is ran on my Mac Studio via docker. I have a node on the Mac Studio and a node on my PC. Since my Mac Studio is also my plex server, the only thing I have that node do is move/save files.
• NVENC logic isn't fully integrated yet
• PC has a RTX 4080 Ti and a AMD Ryzen 5 5600X in it
• This is all trial and error and I'm sure I'm doing things wrong.
• I still need to route general animation through this (kids shows, Midnight Gospel, etc)
• I like really detailed logs, so I try to add as many comments as possible to my flows to help get a more clear picture of what's happening. todo: requires more emojis.
• Tuning for anime content is:

-tune animation -bf 3 -maxrate 2000k -bufsize 4400k -x265-params "frame-threads=6:deblock=-2:-2:psy-rd=1:psy-rdoq=0.5:aq-mode=1:aq-strength=0.80:ipratio=1.4:pbratio=1.3:qpmax=69:qpmin=10:no-cu-lossless=1:no-amp=1:no-rect=1:rskip=1:ctu-info=0:limit-refs=1:no-aq-motion=1:scenecut=40:scenecut-bias=0.1:no-sao=1:strong-intra-smoothing=0:no-sao=1"

• Tuning for non-anime content is:

-x265-params "frame-threads=6:deblock=-2:-2:psy-rd=1.1:psy-rdoq=1:aq-mode=1:aq-strength=0.80:ipratio=1.4:pbratio=1.3:qpmax=69:qpmin=10:no-cu-lossless=1:no-amp=1:no-rect=1:rskip=1:ctu-info=0:limit-refs=1:no-aq-motion=1:no-dynamic-refine=1:scenecut=40:scenecut-bias=0.1"

Variables

Global Variables

Tools > Global Variables

I use global variables as a defacto dotenv and store any sensitive or easily forgettable info here. I have a few groupings: Plex library IDs, my Plex token, volume mapping paths, Arr URLs, and Arr Tokens. The Tdarr paths vs plex paths are because Tdarr uses bind mounts via docker and Plex uses system paths.

You can find instructions on how to get your Plex library IDs here.

Variable	Key Example	Category	Arg
plex_path_movies	/Volumes/Drive_01/data/media/movies/	Plex Path	{{{args.userVariables.global.plex_path_movies}}}
plex_path_tv	/Volumes/Drive_02/data/media/tv/	Plex Path	{{{args.userVariables.global.plex_path_tv}}}
plex_path_anime	/Volumes/Drive_03/data/media/anime/	Plex Path	{{{args.userVariables.global.plex_path_anime}}}
plex_libraryKey_movies	1	Plex Library Key	{{{args.userVariables.global.plex_libraryKey_movies}}}
plex_libraryKey_tv	2	Plex Library Key	{{{args.userVariables.global.plex_libraryKey_tv}}}
plex_libraryKey_anime	3	Plex Library Key	{{{args.userVariables.global.plex_libraryKey_anime}}}
tdarr_path_movies	/mnt/01/media/movies/	TDARR Volume Mapping	{{{args.userVariables.global.tdarr_path_movies}}}
tdarr_path_tv	/mnt/02/media/movies/	TDARR Volume Mapping	{{{args.userVariables.global.tdarr_path_tv}}}
tdarr_path_anime	/mnt/03/media/movies/	TDARR Volume Mapping	{{{args.userVariables.global.tdarr_path_anime}}}
plex_token	xxxxxxxxxxxxx	Plex Token	{{{args.userVariables.global.plex_token}}}
url_plex	192.168.1.xx	Plex IP	{{{args.userVariables.global.url_radarr}}}
url_radarr	http://192.168.1.xx:8989	Arr URL	{{{args.userVariables.global.url_radarr}}}
url_sonarr	http://192.168.1.xx:8989	Arr URL	{{{args.userVariables.global.url_sonarr}}}
url_sonarrAnime	http://192.168.1.xx:8989	Arr URL	{{{args.userVariables.global.url_sonarrAnime}}}
api_key_radarr	xxxxxxxxxxxxx	API Key	{{{args.userVariables.global.api_key_radarr}}}
api_key_sonarr	xxxxxxxxxxxxx	API Key	{{{args.userVariables.global.api_key_sonarr}}}
api_key_sonarrAnime	xxxxxxxxxxxxx	API Key	{{{args.userVariables.global.api_key_sonarrAnime}}}
api_key_tmdb	xxxxxxxxxxxxx	API Key	{{{args.userVariables.global.api_key_tmdb}}}

Library Variables

Variable	Key Example	Used By Flow	Arg	Note
name	MOVIES	All	{{{args.userVariables.library.name}}}	Must be all uppercase
enable_audio_cleaning	true	Controller & Audio Cleaning	{{{args.userVariables.library.enable_audio_cleaning}}}	lowercase, true/false
enable_subs_cleaning	true	Controller & Subs Cleaning	{{{args.userVariables.library.enable_subs_cleaning}}}	lowercase, true/false
enable_audio_transcoding	true	Controller & Audio Transcoding	{{{args.userVariables.library.enable_audio_transcoding}}}	lowercase, true/false
enable_video_transcoding	true	Controller & Video Transcoding	{{{args.userVariables.library.enable_video_transcoding}}}	lowercase, true/false
enable_notifications	true	Controller & Notification	{{{args.userVariables.library.enable_notifications}}}	Not yet implemented
enable_control_flow	true	All	{{{args.userVariables.library.enable_control_flow}}}	Prevents returning to the controller
quality_level	21	Video Transcoding	{{{args.userVariables.library.quality_level}}}	18 to 25 recommended. lower = higher quality
use_nvenc	false	Video Transcoding	{{{args.userVariables.library.use_nvenc}}}	lowercase, true/false
ffmpeg_preset	slower	Video Transcoding	{{{args.userVariables.library.ffmpeg_preset}}}	lowercase. Options: slower, slow, medium, fast, faster. NVENC cannot use 'slower'
use_foreign	false	Subtitle	{{{args.userVariables.library.use_foreign}}}	Optional
use_checkpoints	false	Controller	{{{args.userVariables.library.useCheckpoints}}}	Optional. Should we overrwrite the source file after each flow?
check_hardlinks	true	Controller	{{{args.userVariables.library.check_hardlinks}}}	Optional. Filter check to see if video is hard linked anywhere.

Audio

I have several plex users that have older TVs, so each video should have a base aac 2ch regardless of quality. Additional qualities might be added, depending on the source audio. A few example scenarios:

• Source: 8ch FLAC | Outcome: 8ch AC3, no 6ch, 2ch AAC
• Source: 6ch AAC | Outcome: 6ch AAC, 2ch AAC
• Source: 2ch AC3 | Outcome: 2ch AAC
• Source: 2ch AAC | Outcome: 2ch AAC

The only allowed audio formats are AAC and AC3, all other formats will be discarded. The following are discarded: dca,dts,flac,mp2,mp3,truehd,vorbis,wav,wma,eac3. Commentary is not removed.

If {{{args.userVariables.library.is_remux}}} is set to true, we will only transcode a single AAC audio channel at the highest channel count. Commentary is not removed.

Customizing for Your Personal Setup

Since I take advantage of three nodes (two mac, one PC), I have a few logic flows specific to my setup. If you have a different setup or only use one node, you'll have to adjust these flows accordingly. I've added comments on the flows to help pinpoint which ones need to be adjusted.

Controller Flow

• Bypass 🧠 | 🤖 Tags: Worker Type (MEDIA SERVER) at the top of the flow.
• Bypass 🧠 | 🤖 Tags: Worker Type (PC) at the top-middle of the flow after the 'copy to working directory' plugin.

Subtitle Flow

• No adjustments needed.

Audio Cleaning Flow

• No adjustments needed.

Audio Transcoding Flow

• No adjustments needed.

Video Transcoding Flow

• Bypass Tags: Worker Type GPU after the NVENC decision making plugin towards the middle.
• Bypass Tags: Worker Type CPU after the NVENC decision making plugin towards the middle.
• Bypass Tags: Worker Type ANY in the NVENC sub-flow area towards the bottom.

Cleanup Flow

• Bypass 🧼 | 🤖 Tags: Worker Type (ANY) at the bottom of the flow.

Notification Flow

• No adjustments needed.

ToDo

• Add support for commentary filtering (low priority)
• Add optional library variables for FileSizeLowerBoundsCheck and FileSizeUpperBoundsCheck (high priority)
• Add additional emoji indicators unique to each flow state (high priority)
• Add a way to specify the target audio format (e.g., AC3, AAC) (low priority) or something like KeepAc3
• Add variable to set the maximum number of audio channels to keep (low priority)
• Add optional variable to enable Loudnorm processing (medium priority)
• Add 'checkpoint' file overwrites. If enabled, the controller flow will overwrite the original file after each module. For example, if audio transcoding is successful but video transcoding fails, the audio transcoding will have already overwritten the original file. When we retry the flow, it will bypass all the work needed for audio transcoding and go directly to video transcoding, saving time/resources. ✅ Done!
• Change variable names to camelCase (low priority)
• Replace images in /docs with updated screenshots (high priority)
• Add additional documentation for checkpoint and hardlink variables/logic