Hi Lemmy!

This post is a way for me to announce one of my apps that I made for my personal usage.

Azul box

It is a front end for yt-dlp and ffmpeg. It isn’t just a yt-dlp downloader; it’s more like a utility app that does anything that I need. The “special” part about this software is probably the ability to download YouTube subtitles and then embed them into the audio file as synced lyrics. Well, that is the only “unique” thing about it. As I’m still quite new to programming, there may be some bugs, and I appreciate your understanding. I’m also learning how to package it as a deb/rpm and plan to dedicate time to this during the summer. For now, the only way to download it will be to build from source with the bash install script in my repo.

If you have some time to try the app, I would love to hear your thoughts. Thank you for taking the time to read this! :)

      • webghost0101@sopuli.xyz
        32·
        5 days ago

        Dont give up on it! Its very feasible that yt starts attacking yt-dlp. Invidious is already obfuscated. The more alternatives the better.

          • cheese_greater@lemmy.world
            3·
            4 days ago

            I make a point of always being on the lookout for different versions or offerings exactly because its so debillitating to me when I lose the use of a tool I rely on. Keep on fighting the good fight. I remember how much it sucked when I ProTube stopped working and Ivory stopped working, those were both bad times for me

            • rhabarba@feddit.org
              3·
              4 days ago

              It’s still rather hard to maintain a software that can download from as many sources as possible on my own. It’s often a game of cat&mouse, sadly.

              • cheese_greater@lemmy.world
                3·
                4 days ago

                No no, thats not what I meant. I mean just that I usually keep my eyes peeled for apps that do the same or a similar function as a redundancy backup plan type thing. Like I collect apps that do the same thing but maybe have slightly different or novel implementations cuz I dont want to be caught with my pants down when they wise up to whatever way I was getting the needful done

                Lots of respect for what you’re doing and it must be very difficult to try and ensure that level of compat. Hope you’re kind to yourself about it all, thats alot, yo.

  • N0x0n@lemmy.ml
    12·
    5 days ago

    Cool stuff, thank you for sharing your project with the community ! It’s very cool and useful for people who just want a set and forget situation !

    I stoped using yt-dlp frontends the moment I saw youtube actually serving upscaled opus media files (very visible line on a spectrogram). Also their metadata is totally fucked-up and not very well organized and full of shit (comments with huge spaces and non useful metadata…).

    Sure, the metadata part is easily fixable with Picard MusicBrainz, but the quality downgrade was a huge no for me. Nowadays, I use nicotine, rutracker and private trackers to download FLAC quality files and transcode them to opus 192k to serve them in my Navidrome library with a well curated metadata structure !

    Yes it takes way more time and some dedication but it’s worth it :) specially If you are some kind of perfectionist and like everything neatly organized ! 😁

    More power to you for keeping the opensource community thriving !! Thank you !

    • Kiuyn@lemmy.mlOP
      2·
      5 days ago

      Thanks for the nice comment! I’m actually planning to maybe integrate the Picard metadata into my music downloader. It is still a plan though. I have no working prototype yet.

    • Jiří Král@discuss.tchncs.de
      2·
      5 days ago

      I stoped using yt-dlp frontends the moment I saw youtube actually serving upscaled opus media files (very visible line on a spectrogram). Also their metadata is totally fucked-up and not very well organized and full of shit (comments with huge spaces and non useful metadata…).

      Wow really? Are you sure it applies to all audio files? YouTube gathers music records from different companies so they could be of varying quality. To me the opus quality from YouTube was always decent and personally I cannot hear any compression in the audio. The metada is not perfect, but I usually use some tag editor to complete what’s missing. YTDLnis on Android does a great job of scraping as much usable metada from YouTube Music as possible.

      • N0x0n@lemmy.ml
        2·
        5 days ago

        Not sure if every audio file is upscaled, however I had a few files who clearly were 128kbs mp3 upscaled to ~192kbs opus !

        personally I cannot hear any compression in the audio.

        100% agree on that xD !! However, I’m some kind of strange audio freak hoarder. If the spectogram of an audio doesn’t reflect my standards (not upsacles audio files…) I won’t keep/archive it !

        Do note that I didn’t used a paid YouTube account… Maybe they don’t serve HQ audio files to free users? Dunno, but after some digging into the spectograms I was not happy seeing how bad they looked !

    • Kiuyn@lemmy.mlOP
      2·
      4 days ago

      Yes, it work perfectly for youtube. In fact, it is the main way I download and consume music now days!

  • cheese_greater@lemmy.world
    2·
    4 days ago

    Hey, while we have you, can I ask what is the best way to join or merge all the downloaded videos still using the command line. Like do I cd to the download directory and have it do something, something somethint ffmpeg?

    Also, is downloading the mp4 significantly different in sound quality than bestaudio (which tends to output a webm rather than mp4)?

    • Kiuyn@lemmy.mlOP
      3·
      4 days ago

      For the first question, I assume that you want to put all video downloaded into one big video in CLI. I guess you can do something like this with ffmpeg: ffmpeg -i "concat:input1|input2|input3|etc" -codec copy output.mkv If you want to work it out with yt-dlp you can try this:

      yt-dlp "a link" Assuming you get mp4 format ls -1 *.mp4 > file_list.txt then ffmpeg -safe 0 -f concat -i list.txt -c copy output.mp4

      Also, is downloading the mp4 significantly different in sound quality than bestaudio (which tends to output a webm rather than mp4)?

      For your second question, I am not sure my self to be honest. I never actually look at it. That is why for my app this is the arg I use to get best video and also best audio: -f "bestvideo[ext=mp4]+bestaudio[ext=m4a]/best[ext=mp4]/best"

      You can probably play a bit more with the format, if you feel like it.

      Hope this help!