System Of A Down - The Pretender (Foo Fighters) Serj Tankian & Daron Malakian AI Cover

Sdílet
Vložit
  • čas přidán 7. 12. 2023
  • Here is another song that I wanted to cover for a while, and now that the angry serj is back, i was able to do it. So here is the cover of The Pretender, a song by Foo Fighters. The models used were angry Serj, normal Serj and Daron.
    AI cover tools and voice models:
    How I make Ai covers: • Video
    _________________________________________________________________________________
    RVC: works a few minutes then needs to be restarted
    colab.research.google.com/drive/1Gj6UTf2gicndUW_tVheVhTXIIYpFTYc7
    Separate vocals and music: 1 free conversion per day on this site:
    vocalremover.org
    AI HUB:
    discord.com/channels/1159260121998827560/1159548890861801553
    _________________________________________________________________________________
    VOICE MODELS:
    DARON 500 EPOCHS:
    huggingface.co/Shadow11/DaronMalakian/blob/main/DaronMalakianModel.zip
    SERJ TANKIAN TOX ERA 1K EPOCHS:
    huggingface.co/Ryanz48/RyanzRVCModels/resolve/main/SerjTankianToxRVC.zip
    SERJ TANKIAN models made by Bob90 from Discord - better selected dataset and epoch number:
    1. Generic model, 1150 epochs:
    huggingface.co/robob90/serj-tankian-toxicity-v2/resolve/main/serj-tankian-toxicity-v2-e1150.zip
    2. Angry model, 850 epochs:
    huggingface.co/robob90/serj-tankian-toxicity-v2/resolve/main/serj-tankian-toxicity-scream-v2-850e.zip
    3. Clean singing only model, 1000 epochs:
    huggingface.co/robob90/serj-tankian-toxicity-v2/resolve/main/serj-tankian-toxicity-clean-v2-e1000.zip
    4. Soft singing model, 500 epochs:
    huggingface.co/robob90/serj-tankian-toxicity-v2/resolve/main/serj-tankian-toxicity-soft-v2-e500.zip
  • Hudba

Komentáře • 24

  • @josephhinojosa4310
    @josephhinojosa4310 Před 7 měsíci +23

    It's scary, how good this is...

  • @danialonderstal3564
    @danialonderstal3564 Před 7 měsíci +3

    Its awesome hearing this stuff, its like hearing the singers are playing rockband or something and are just masters of the vocals in general, not just in their genre

  • @Artofwar3004
    @Artofwar3004 Před 7 měsíci +10

    This sounds so good hahah

  • @MultiLakmus
    @MultiLakmus Před 7 měsíci +8

    Undeservedly few views of this AI cover

  • @justspheal489
    @justspheal489 Před 7 měsíci +6

    I'll be honest, it's not as bad as the others say. The main issue I think is a bit too much Daron. Leave him to harmonies and the "Keep you in the dark" sections and I think this would be great!

    • @aronstecher1222
      @aronstecher1222 Před 7 měsíci

      Honestly... that can also be aplied to a lot of SOAD songs... simply too much Daron

  • @Ghoulascor
    @Ghoulascor Před 7 měsíci +2

    FOO FIGHTRERS

  • @snowfoxxo9480
    @snowfoxxo9480 Před 6 měsíci

    P E R F E C T I O N

  • @Allrespect420seven10
    @Allrespect420seven10 Před 6 měsíci +1

    I always thought how this would be lol

  • @serjspersonalpear
    @serjspersonalpear Před 7 měsíci +8

    Great video man! I love the balance between Daron and Serj, but I think Serj should sing more and leave Daron the harmony parts just like what the others said.

    • @gabywinner
      @gabywinner  Před 7 měsíci

      I posted another version with more Serj and less Daron. Check it out and see if you like it

    • @serjspersonalpear
      @serjspersonalpear Před 7 měsíci

      @@gabywinner Great! I will definitely do that now.

  • @BlazeBuds
    @BlazeBuds Před 7 měsíci +4

    Wow I prefer this over the original lol not a fan of the foo fighters tbh and love soad, I actually think this could pass my test. Will play it for a mate and tell him its just a cover and see if he notices

  • @gilbertosalazar5720
    @gilbertosalazar5720 Před 7 měsíci +2

    AI

  • @caiofarias8806
    @caiofarias8806 Před 7 měsíci +1

    how do you do this?

    • @gabywinner
      @gabywinner  Před 7 měsíci

      something like this: czcams.com/video/gtjCmJ_6d0k/video.html . I use vocal remover .org to split the song into vocals and music. Then I use the google colab - RVC for vocal conversion. You find the voice models on AI hub but if you don't find Daron and Serj I can paste the links for the voice models I use, here. Lately the google colab crashes after a few minutes so you have to move fast converting stuff, and when it crashes you have to reinstall it. Then you need to use an editing software like premiere pro to combine the vocals and instrumental together. I create a thumbnail in Canva and use it with the mp3 to create a one image video and post it. Note that for this particular song I searched on youtube two acapelllas - the backing vocals and the main acapella without the backing vocals - and converted them separately - because otherwise the chorus would have been a total mess.

    • @gabywinner
      @gabywinner  Před 7 měsíci

      added the voice models to the description

  • @self-discovery-official
    @self-discovery-official Před 7 měsíci +1

    Glad they stopped at Mez / Hyp if this would be the future of them hahahaha

    • @RufusShinra
      @RufusShinra Před 7 měsíci +3

      That's, like ... your opinion bro ...

    • @self-discovery-official
      @self-discovery-official Před 7 měsíci +1

      @@RufusShinra yes it is. nothing more, nothing less.

    • @rohan750
      @rohan750 Před 7 měsíci +2

      na its dumb when bands refuse to make an album but continue to tour!

  • @LionelRGuy
    @LionelRGuy Před 7 měsíci

    Please God no.

    • @mlb-yt
      @mlb-yt Před 7 měsíci

      Sounds like you need you put the phone down and just move on if you see something you don't like, snowflake.