Announcement

Collapse
No announcement yet.

AMD's UVD2-based XvBA Finally Does Something On Linux

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

  • I works a bit better for VC1 and some H264 L4.1 work too. But compared to vdpau it is definitely crap in the current state.
    Last edited by Kano; 11-03-2009, 08:48 PM.

    Comment


    • Originally posted by Kano View Post
      I works a bit better for VC1 and some H264 L4.1 work too. But compared to vdpau it is definitely crap in the current state.
      the opensource driver will be better than nvidia's vdpau ....

      you know it!

      in my point of view i never touch a Nvidia again 8600gt RIP 7900GT RIP Nvidia is a complete FAIL!

      FAIL in the hardware side not software all my ati/amd cards are still alive up to my 8500 and X800 and so on.. all my nvidia's RIP...

      Comment


      • The OSS driver is not even working correctly for years old hardware - will update my snapshot soon to have got a few new ids, but basically the ati drivers are the most problematic ones. Until we can see working video accelleration with ati cards you will have to wait more likely years than months. Nvidia's vdpau works NOW.

        Comment


        • Originally posted by Kano View Post
          The OSS driver is not even working correctly for years old hardware - will update my snapshot soon to have got a few new ids, but basically the ati drivers are the most problematic ones. Until we can see working video accelleration with ati cards you will have to wait more likely years than months. Nvidia's vdpau works NOW.
          "Nvidia's vdpau works NOW."

          no on my hardware vdpau does NOT work!

          because all my nvidia hardware are broken !

          geforce 7900 RIP geforce 8600 RIP...

          working driver on an failing hardware is no driver!

          Comment


          • Originally posted by Qaridarium View Post
            the opensource driver will be better than nvidia's vdpau ....

            you know it!

            in my point of view i never touch a Nvidia again 8600gt RIP 7900GT RIP Nvidia is a complete FAIL!

            FAIL in the hardware side not software all my ati/amd cards are still alive up to my 8500 and X800 and so on.. all my nvidia's RIP...
            I've got about 6 dead ATI cards for ya. No nvidia's.

            Comment


            • Originally posted by tuxdriver View Post
              http://dl.getdropbox.com/u/132004/Pl...mple_16ref.mkv (The well-known Bird documetary , 1080p H.264)
              Okay this one is the first I've seen that actually makes this thing (UNpatched mplayer...i.e. w/o VAAPI) sweat As soon as the birds appear the a-v sync just gets worse and worse and it begins to visibly drop frames, heh. Unfortunately an assert stops the VAAPI-patched mplayer:

              Code:
              mplayer: libvo/vo_vaapi.c:1436: get_surface: Assertion `mpi->number < va_num_surfaces' failed.
              Corrupt video window when this happens. Tried seeking to various starting points (mplayer -ss) with no luck.

              http://showcase7.divx.com/ElephantsDream[DivX7].mkv (Open Source animated video , 1080p H.264)
              This one I had tried before along with Big Buck Bunny; same results as most of the others: average usage w/o accel, negligible use with.

              http://www.andyqos.ukfsn.org/BBCHD-P-40.ts (BBC TV clip , 1080p H.264)
              Ditto

              http://video.fraps.com/sunrise.zip (Game Video ,720p 60fps! , VC-1)
              This wasn't too bad to begin with, actually...only about 50% usage peak w/o accel.

              Thanks for the links.

              Cheers,
              Jesse
              Last edited by avsa242; 11-03-2009, 10:02 PM.

              Comment


              • Originally posted by monraaf View Post
                Okay I just had my 3rd reboot due to a X Server lock up today , I'm done testing. F*ck UVD, I'll wait for the stable open source shader based solution.
                Whoops, spoke too soon. Even without the -vo vaapi option the patched mplayer crashes. Looks like the checkout-patch-build.sh script from mplayer-vaapi-latest.tar.bz2 build me a borked version of mplayer

                Comment


                • Originally posted by deanjo View Post
                  I've got about 6 dead ATI cards for ya. No nvidia's.
                  i found more apple notebooks broken bei nvidia's cpus than ati ones...

                  nvidia fail!

                  http://www.electronista.com/articles...o.nvidia.risk/
                  Last edited by Qaridarium; 11-03-2009, 10:10 PM.

                  Comment


                  • @monraaf

                    when you use the script i posted you get a fully updated mplayer (it just calls svn up) + a m2ts audiofix for truehd which is not mainline - i do not like to post things on mailinglists.

                    Comment


                    • Originally posted by Qaridarium View Post
                      i found more apple notebooks broken bei nvidia's cpus than ati ones...

                      nvidia fail!

                      http://www.electronista.com/articles...o.nvidia.risk/
                      Really? Because I can tell you as a matter of fact even including bumpgate effected cards the ATI's have a higher failure rate then the Nvidia cards. For every Nvidia card that was replaced ATI cards were replaced 10 x more. These include the 7500, 2600HD, X1900, X850, 9000 Pro.

                      In fact there is a whole section in Apple engineering notes to agents under the heading "ATI distorted video" that all say "Replace the card". In fact most of the extended beyond warranty coverage on some machines such as the eMac had to do with ATI graphics failing.
                      Last edited by deanjo; 11-03-2009, 11:07 PM.

                      Comment

                      Working...
                      X