3 ms·
I only watch TV through downloads (get_iplayer & youtube-dl), and I have kept it all [edit: only things I downloaded intending to watch, not all TV]. This is s
by rwmj 8y ago
I only watch TV through downloads (get_iplayer & youtube-dl), and I have kept it all [edit: only things I downloaded intending to watch, not all TV]. This is since about 2008:
fs:/volume1/media 14T 8.0T 5.6T 59% /mnt/media
The reason? Why not - disks are cheap. The hard thing is deleting something since I don't know if I might want to watch it again. And at almost no real cost I don't have to make that decision.
Edit: Some stats: 20,269 TV programmes; 11,830 radio programmes; 925 movies
- vthriller 8y agoI wish I started doing similar thing way earlier, because now I can't refer to a lot of content that vanished from the net for a number of reasons (creators getting bored, flagged out from youtube by bots that random haters set up, bogus copyright claims from impostors that upload all sorts of stolen things into contentid in order to create a quick buck, UMG taking down videos from music theory channels despite it obviously falling under the Fair Use and so on)
- rwmj 8y agoFor music on youtube, this trivial ytdl wrapper is essential: $ cat bin/youtube-mp3 #!/bin/bash - exec $HOME/d/youtube-dl/youtube-dl -t -x --audio-format=mp3 "$@" Saved to my Squeezebox library, it's available to play forever.
- qplex 8y agoDoesn't the audio quality take a hit when you transcode to MP3 though?
- rwmj 8y agoPossibly, but on the flip side neither my car nor my first-gen SLIMP3 can play other formats natively.
- asveikau 8y agoYouTube audio is usually aac. Sometimes I have saved audio as an .aac file without re-encoding. (Not using the same scripts, what I have done typically is get the mp4 and do: ffmpeg -i foo.mp4 -acodec copy foo.aac) Any audio player worth its salt should play .aac files. I think for the most common types [low-complexity] the patents are expired. (Curious if somebody has one that doesn't.)
- rhizome 8y agoProbably, but it's already down to ~150kb by the time you download it.
- mmcwilliams 8y agoDo you maintain any kind of database or additional metadata alongside all of those video files? I only ask because the default filenames generated by youtube-dl don't give you anything more than the title and ID for a particular video and I imagine it's difficult to search among tens of thousands of media files without contextual data about the files.
- rwmj 8y agoNo, just filenames and whatever metadata is stored inside the file. For get_iplayer this isn't very much unless you have "AtomicParsely" installed too, but I uninstalled AP after it segfaulted once too often. Note for BBC TV programmes you can search on the "PID" (which is part of the filename), eg: try searching for "m000427z". However I also download subtitles where available so I guess I could search those as they are just text.
- rndgermandude 8y agoyt-dl has a full-fledged output template system. It's just the default output template that is only the title+id. https://github.com/ytdl-org/youtube-dl/blob/master/README.md#output-template https://github.com/ytdl-org/youtube-dl/blob/master/README.md... Also, it has the --write-* switches, in particular --write-info-json, which allow to store metadata data. in files, and --add-metadata to embed (some) of the metadata in the file itself. Of course, the metadata is only as good as the site in question provides + the yt-dl plugin handing a particular site supports.
- mmcwilliams 8y agoI was familiar with the -o template system for filenames, but wasn't aware of the --write-info-json flag which is very useful to me. It makes a lot more sense than overloading the filenames with usernames and everything else only to parse it out after download.
- brian_herman__ 8y agoNice man!
- R_Macdonald 8y agoWould you consider donating a copy of the TV news programs and radio to the internet archive?