SVT-AV1 vs AV1 NVENC Quality Comparison | Multiple Bitrates #01

แชร์
ฝัง
  • เผยแพร่เมื่อ 15 ม.ค. 2025

ความคิดเห็น • 25

  • @ExtremelySimplified
    @ExtremelySimplified  ปีที่แล้ว +4

    Watch in the highest possible quality. Feel free to pause at any moment in the video to be able to distinguish the differences.
    Watch more comparisons:
    th-cam.com/play/PLRW-minR44q8CXc2M4Jf1GwxQGdtboHOc.html&si=l20yLvZ9648lU9lo

  • @elmacmado
    @elmacmado 4 หลายเดือนก่อน

    Hey! I just discovered your channel and i´m really really really impressed and gratefull with your videos and your descriptions! No one gives the time anymore to write usefull descriptions. My job is asking me for testing 7900XTX series with streaming, recording, etc... and all your videos just helped me a lot, infinite thanks!
    Hope you can monetize you channel soon for a super thanks (i know how much time and effort it takes) you really have impressive content... keep it this way!

    • @ExtremelySimplified
      @ExtremelySimplified  4 หลายเดือนก่อน

      Thank you so much for your kind words! My goal was to create comprehensive comparisons to help people understand the differences between these codecs. Regarding monetization, my channel is already monetized, but Super Thanks isn't available in my country. Nonetheless, your support truly means the world to me, and I deeply appreciate it. 🙏🏻
      Keep it up on your TH-cam channel. 💟

    • @elmacmado
      @elmacmado 4 หลายเดือนก่อน

      ​@@ExtremelySimplifiedi´m glad to hear that! Gonna keep my bell on for new content!!
      I would like to ask you something.... in other video (Best OBS Recording Settings for Beginners! (Highest Quality) you used CPQ at 20
      And here i see you record your source video with variable bitrate at 100-100k.
      ¿Wich one do you recommend more CPQ at 12-16 o VBR at 110k? i have a dual setup my Streaming PC (7900XTX) is capturing Gaming PC with NZXT signal4k at 1080p60fps. (sorry for too much info).

    • @ExtremelySimplified
      @ExtremelySimplified  4 หลายเดือนก่อน

      No problem!
      I personally use CQP 20 all the time with all my different resolutions (4K, 1440P, 1080P).
      it gives me the best balance between file sizes and bitrate. The bitrate usually goes up the higher the resolution is.
      But CQP for me is intended for using the recorded clips to be compressed later on when I edit them for example.
      As to why I chose VBR with 110K, it's simply because I needed a somewhat predictable file size so I can put the information of it in the description box. Otherwise, I don't use VBR at all.
      To give you the best solution:
      For streaming use CBR with a predetermined bitrate.
      For recording for editing later on, choose CQP 20 or even lower like 18 or 16.
      It's funny you mentioned it; I also have the NZXT Signal 4K30, and I'm planning to review it soon! 😄

    • @elmacmado
      @elmacmado 4 หลายเดือนก่อน

      @@ExtremelySimplified Oh i see! I´m gonna take your advices! Guess the best birate is gonna be for the main video source and then try all codec and bitrates variations.
      Also Signal4k30 it´s pretty cool! for 60 fps it´s amazing, sometimes a little fps drops with 120 fps (i know few people use 120fps), but guess that´s because cable data limitation. Don´t know if your using that in your video but would be nice if you show us how to use it properly for best quality!

    • @artischeff
      @artischeff 15 วันที่ผ่านมา

      @@ExtremelySimplified I'm a bit of a layman but I'd like to know more. I would like to record and transmit with the best possible quality in 1920x 1080 at 360hz/540hz, thank you in advance

  • @Renan-cg2ox
    @Renan-cg2ox 18 วันที่ผ่านมา

    awesome video, the greatest feature is the vmaf. ty for the info.

  • @WebLeoTV
    @WebLeoTV ปีที่แล้ว +3

    Great job!!! Like this comparison. thought minutes were hours! but i know you have a nice PC (mine is a ryzen 3600, do some test but take too much time). Thanks again for the video. Conclusions for me. 3 and 5.5 mbps CPU encoding looks better, but at 10 mbps, looks the same as GPU.

    • @ExtremelySimplified
      @ExtremelySimplified  ปีที่แล้ว +8

      You're welcome.
      The video took minutes to encode because I used a powerful CPU and also the source video is just 1 minute long. If it was 30 minutes or an hour long, it would take forever to encode.
      AV1 NVENC is a good encoder but I think that it needs high bitrate alongside it. I'd say 15 mbps for 1080p60 or somewhere around 35-40 for 4K60.
      SVT-AV1 is still superior to AV1 NVENC with lower bitrates since it's using more time to encode. And P3 is not useful compared go P6! The extra time it takes is not worth it in my opinion.

  • @roideus
    @roideus ปีที่แล้ว +4

    I tried AV1 streaming and was left disappointed with NVENC performance. The VMAF numbers are really misleading.

  • @Rocky712_
    @Rocky712_ ปีที่แล้ว +2

    Wow, at 5 mbps and higher the NVenc seems pretty strong, especially considering almost identical file size. For low bitrates the vmaf gap is quite big while there is roughly 10-20% file size difference (preset 3 vs nvenc at 3mbps)
    Nonetheless, very promising. I am wondering how well AMD's AV1 implentation would compare to this? I'd love to see that, especially since AMD has a strong price/performance ratio in more general cirumstances.

    • @ExtremelySimplified
      @ExtremelySimplified  ปีที่แล้ว +1

      I agree. NVENC is a much better option especially for bitrates that are 10 mbps and above. it's not efficient to compress the videos using SVT-AV1 P6 or P3 due to the slow time it takes to encode.
      Unfortunately I don't have AMD GPU to test their new AV1 encoder. however, from what I've read online, I think it's a little bit inferior to NVIDIA's NVENC. I could be wrong though.

  • @MehdiSattari
    @MehdiSattari 9 หลายเดือนก่อน +1

    What tool(s) have you used to encode?

    • @ExtremelySimplified
      @ExtremelySimplified  9 หลายเดือนก่อน +2

      Hi. I mentioned all the details in the description.

  • @hashtag9990
    @hashtag9990 4 หลายเดือนก่อน

    it is better to upload the actual videos in some cloud storage like gdrive and link them, so people can just play them by themselves to see the difference, here in youtube is almost pointless

  • @TheLiddokun
    @TheLiddokun ปีที่แล้ว +1

    Yikes vmaf is really bad. Even at 10,000kbps thr mountains were much clearer eith svt av1 rven tho vmaf was lower. Sv1 av1 looked almost transparent whereas nvenc was a muddy mess

    • @ExtremelySimplified
      @ExtremelySimplified  ปีที่แล้ว +1

      Indeed! Thank god I wasn't the only one who noticed that.

  • @Diegonando64
    @Diegonando64 4 หลายเดือนก่อน

    Impressive

  • @user-hv5jv9gb6c
    @user-hv5jv9gb6c 9 หลายเดือนก่อน +1

    Sooo... AV1 Nvenc is useless.

    • @ExtremelySimplified
      @ExtremelySimplified  9 หลายเดือนก่อน +1

      I'd suggest you watch this comparison as well:
      th-cam.com/video/x6qeG8KJ1Vw/w-d-xo.html

    • @user-hv5jv9gb6c
      @user-hv5jv9gb6c 9 หลายเดือนก่อน +2

      @@ExtremelySimplified It's nearly exactly the same as H265. I don't play games but Transcoding I'll take Av1-AOM over H265, AV1-Nvence, or AV1-STV any day of the week.

    • @artischeff
      @artischeff 15 วันที่ผ่านมา

      @ExtremelySimplified Do you think the same thing?