Announcement

Collapse
No announcement yet.

Intel Sandy Bridge VA-API Performance

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Originally posted by Shining Arcanine View Post
    Try a trailer for a high profile film like Avatar:

    youtube-dl http://www.youtube.com/watch?v=nTtBXpw8db4

    It is considered fair use.
    You tube stuff isn't even moderately close to being demanding. Their HD is really crappy and poorly encoded. It's optimized for web streaming. Also movie trailers are NOT fairuse.
    Last edited by deanjo; 03-07-2011, 08:45 PM.

    Comment


    • #17
      Originally posted by deanjo View Post
      You tube stuff isn't even moderately close to being demanding. Their HD is really crappy and poorly encoded. It's optimized for web streaming.
      Did you watch the trailer? The HD quality is not bad and the bit rate is something like 6Mbps. What do you consider to be a good bitrate?

      Comment


      • #18
        Originally posted by Shining Arcanine View Post
        Did you watch the trailer? The HD quality is not bad and the bit rate is something like 6Mbps. What do you consider to be a good bitrate?
        Yes, I even have the 1080p file.

        General
        Complete name : Avatar- Extended Collector's Edition Trailer- On Blu-ray_DVD November 16.mp4
        Format : MPEG-4
        Format profile : Base Media / Version 2
        Codec ID : mp42
        File size : 55.5 MiB
        Duration : 1mn 22s
        Overall bit rate : 5 655 Kbps
        Encoded date : UTC 2010-10-04 22:58:43
        Tagged date : UTC 2010-10-04 22:58:43

        Video
        ID : 1
        Format : AVC
        Format/Info : Advanced Video Codec
        Format profile : High@L4.0
        Format settings, CABAC : Yes
        Format settings, ReFrames : 1 frame
        Codec ID : avc1
        Codec ID/Info : Advanced Video Coding
        Duration : 1mn 22s
        Bit rate mode : Variable
        Bit rate : 5 522 Kbps
        Maximum bit rate : 12.0 Mbps
        Width : 1 920 pixels
        Height : 1 080 pixels
        Display aspect ratio : 16:9
        Frame rate mode : Variable
        Frame rate : 29.970 fps
        Minimum frame rate : 29.412 fps
        Maximum frame rate : 200.000 fps
        Color space : YUV
        Chroma subsampling : 4:2:0
        Bit depth : 8 bits
        Scan type : Progressive
        Bits/(Pixel*Frame) : 0.089
        Stream size : 54.2 MiB (98%)
        Tagged date : UTC 2010-10-04 22:58:46

        Audio
        ID : 2
        Format : AAC
        Format/Info : Advanced Audio Codec
        Format version : Version 4
        Format profile : LC
        Format settings, SBR : No
        Codec ID : 40
        Duration : 1mn 22s
        Bit rate mode : Variable
        Bit rate : 127 Kbps
        Maximum bit rate : 170 Kbps
        Channel(s) : 2 channels
        Channel positions : Front: L R
        Sampling rate : 44.1 KHz
        Stream size : 1.25 MiB (2%)
        Title : (C) 2007 Google Inc. v08.13.2007.
        Encoded date : UTC 2010-10-04 22:58:45
        Tagged date : UTC 2010-10-04 22:58:46
        It's not very demanding at all, Blu-ray spec has a max of 48mbps combined audio and video, of which 40 can be video, and 20 audio at any one time. This is also a CABAC encoding where a more demanding encode uses CAVLC.

        **** Correction, CABAC is the more demanding on decoding.
        Last edited by deanjo; 03-07-2011, 09:29 PM.

        Comment


        • #19
          Originally posted by cl333r View Post
          Test not fair for Nvidia imo,
          Sandy Bridge is the latest greatest from Intel while GeForce 9500 is a (much) older Nvidia product.

          I'd vouch at least for Sandy vs GeForce 460+ to call it a fair comparison, not to mention taking a "more real-world" x264 video as other folks implied.
          http://www.anandtech.com/show/4083/t...-2100-tested/8

          You just lost your ability to vouch for two months

          Michael, try this http://www.sitasingstheblues.com/watch.html.
          I've not downloaded the 1080p version but it is CC.

          Comment


          • #20
            Originally posted by Michael View Post
            Where is any other high quality sample content that's widely available on HTTP/FTP mirrors and under a free license? I haven't seen such links.
            what , you dont know where to get and use the generic HD park_joy parkrun etc HD samples as used in many a x264 test everywhere, or con cat them togetehr to make longer sequences

            http://media.xiph.org/video/derf/
            http://media.xiph.org/video/derf/y4m..._joy_2160p.y4m etc

            OC there's also the killasample.mkv Here is a link http://rapidshare.com/files/82525583....x264.mkv.html

            and BTW when are you finally going to reset the edit time as you sead you would a long time ago now...

            Comment


            • #21
              I'm curious about the Arrandale CPU/GPU combos

              My interest lies in what the difference might be between the Arrandale setups and these newer ones. This article referenced the previous article about the Clarkdale performance, but it's how well the GPU can do even with a low-power CPU that I'm curious about. I tried digging around a bit on openbenchmarking.org but I couldn't seem to actually find any, for example, i3-330UM tests.

              I mean, it's all well and good that the latest and greatest Intel chipsets perform so well for video playback, but in cases like for http://www.system76.com/product_info...roducts_id=106 I'd like to know if an ultraportable that's purely Intel (and thus can be purely blob-free) is up to the task.

              I mean, I'd hope so; I play 1080p video all the time on my PIII 600mHz and all it has in it is a GeForce 8400GS, heh, but that's all that computer needs to do; for something like a laptop carried around every day, I'd rather not have to deal with closed-source drivers.

              Comment


              • #22
                Originally posted by deanjo View Post
                Well I started downloading the raw pngs for ED. I'll see what I can come up with for some samples encoded at various levels. I'll also see about creating a VC1 version as well as possible a WebM version.
                make sure you grab and use the very latest x264/ffmpeg git pull's though not some antiquated distro package from last year

                http://media.xiph.org/video/derf/y4m/1080p/riverbed.y4m is good

                Comment


                • #23
                  Originally posted by cl333r View Post
                  Test not fair for Nvidia imo,
                  Sandy Bridge is the latest greatest from Intel while GeForce 9500 is a (much) older Nvidia product.

                  I'd vouch at least for Sandy vs GeForce 460+ to call it a fair comparison, not to mention taking a "more real-world" x264 video as other folks implied.
                  VDPAU uses the internal Nvidia fixed function ASIC video decoder and that is the same in all of them even the GeForce 460 AFAIK so makes not difference to the result

                  Comment


                  • #24
                    Originally posted by popper View Post
                    OC there's also the killasample.mkv Here is a link http://rapidshare.com/files/82525583....x264.mkv.html
                    Although a really nice demanding clip the legality of it is questionable as it is a clip from the BBC's documentary living planet and redistribution puts the fairuse into question.

                    Comment


                    • #25
                      Originally posted by popper View Post
                      make sure you grab and use the very latest x264/ffmpeg git pull's though not some antiquated distro package from last year
                      Not to worry, I'm usually using daily builds.

                      Comment


                      • #26
                        Originally posted by popper View Post
                        VDPAU uses the internal Nvidia fixed function ASIC video decoder and that is the same in all of them even the GeForce 460 AFAIK so makes not difference to the result
                        Actually the decoders performance does vary from generation to generation and performance differences are even noticed with faster larger cards. This is easily verified with vdpauinfo.

                        Comment


                        • #27
                          Whats the problem taking ANY h264 clip or movie?

                          In the end, you will not be broadcasting the clip itself, but will share the benchmarking results. The clip itself will be played on your machine at time of test, without leaving it.

                          Comment


                          • #28
                            Originally posted by crazycheese View Post
                            Whats the problem taking ANY h264 clip or movie?

                            In the end, you will not be broadcasting the clip itself, but will share the benchmarking results. The clip itself will be played on your machine at time of test, without leaving it.
                            I won't do any test that is not in the Phoronix Test Suite tree otherwise it cannot be easily reproduced, etc.
                            Michael Larabel
                            http://www.michaellarabel.com/

                            Comment


                            • #29
                              Originally posted by crazycheese View Post
                              Whats the problem taking ANY h264 clip or movie?

                              In the end, you will not be broadcasting the clip itself, but will share the benchmarking results. The clip itself will be played on your machine at time of test, without leaving it.
                              To get A vs B comparisons you need the same clip. To do this you also have this clip freely available without any potential legal whiplash from the content creator/owner.

                              Comment


                              • #30
                                Originally posted by deanjo View Post
                                To get A vs B comparisons you need the same clip. To do this you also have this clip freely available without any potential legal whiplash from the content creator/owner.
                                You will not, granted you possess the clip.
                                So if A/B==gfx cards of same owner, owning the clip and only publishing the results - its not a problem.
                                But if A/B==different people with different machines they will all be required to license the clip, its logical. Yes, thats what Michael told me.

                                And if we create such clip - from big buck bunny lower bitrate h264 as source or multiple big buck bunny screens rendered from one lower bitrate formats into new file especially for phoronix test? It won't be commercial encode, the source is also free - should be no problem. Of course it will require some location to be buffered on the internet.
                                What do you think?

                                Comment

                                Working...
                                X