Announcement

Collapse
No announcement yet.

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

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

  • Well "embedded" could be bluray players. But even with those content i have examples with wrong color - vdpau shows em correctly. Maybe ATI wants to hide the Sony connection with that sdk - there are so many function names with deinterlacers or filters that it could not be just an coincidence.

    Comment


    • Originally posted by Kano View Post
      Well "embedded" could be bluray players. But even with those content i have examples with wrong color - vdpau shows em correctly. Maybe ATI wants to hide the Sony connection with that sdk - there are so many function names with deinterlacers or filters that it could not be just an coincidence.
      are you playing from a bluray player or from your pc?

      Comment


      • I play bluray content (m2ts) with mplayer vaapi - using xvba or vdpau backend (or pure vdpau in case or errors).

        Comment


        • Originally posted by Kano View Post
          I play bluray content (m2ts) with mplayer vaapi - using xvba or vdpau backend (or pure vdpau in case or errors).
          Well, then the case of embedded vs pc still stands.

          Comment


          • What's your definition of embedded? Basically it is a pc, it just does not look like one. xvba can be used with opengl output via xvba-video wrapper - that's nice to try, but it shows directly the limits.

            a) The only known driver is 9-10 with working OpenGL output - and that driver is definitely broken.
            b) OSD does only work without OpenGL output - but how to get that as long as the SDK is not available. Without OSD no subtiles are possible too - a very stupid restriction.

            On a standard pc there are more possibilities than in a bluray player - there it only has to decode mpeg2 (which is not possible with mplayer vaapi), vc1 (works most of the time) and h264 l4.1 (it works in maybe 70% of all cases at max).

            If you try other/non-standard h264 files then you can get everything from just rendering errors to complete driver lockup. That happended with vdpau too in the beginning, but now this is a really rare case. But compared even to the first vdpau enabled drivers it was much better than this 1 year old xvba!

            Comment


            • Originally posted by Kano View Post
              What's your definition of embedded? Basically it is a pc, it just does not look like one.
              As someone who works with embedded systems, I stopped reading your comment right there. Embedded systems are very different to a pc.

              Comment


              • Following this thread for a while now...

                I've been following this thread for a while now and it seems to show both the best and worst of OSS. First, I think thanks are in order to gbeauchesne for doing the work. I personally am excited about getting some HW acceleration of video, although, I don't really need it with a okayish CPU and MB and no Blueray to play. It's a nice technical achievement, although my 9.10 64 bit Ubuntu on a HD4670 the colors are a bit off sometimes (green and red in particular)
                Second, Kano did a nice job IMHO with the mplayer-vaapi script. Thanks for that.

                On the whole NVIDIA vs AMD flamewar: come on guys, that's not necesarry. I feel the sentiment, that if you buy some hardware which doesn't work directly or as nice as under Windows, it might be a little frustating. Then again, new hardware and linux only play nice every now and then, and it has been like that for some time . The point being: if you desperately need something to work, buy something that has been shown to work and not the newest thing you can find. If you decide to buy new stuff and have problems.... well you knew that could happen, right?

                Comment


                • Well as i do no know of fglrx for something else than x86 then tell me what you think is so different?

                  Comment


                  • If you're of the opinion that programming for a pc (x86) is the same as, for example an embedded PPC, or microblaze or msp430, then I suggest you find some cheap devices and play around with them.

                    Comment


                    • Originally posted by Dewni View Post
                      On the whole NVIDIA vs AMD flamewar: come on guys, that's not necesarry. I feel the sentiment, that if you buy some hardware which doesn't work directly or as nice as under Windows, it might be a little frustating. Then again, new hardware and linux only play nice every now and then, and it has been like that for some time . The point being: if you desperately need something to work, buy something that has been shown to work and not the newest thing you can find. If you decide to buy new stuff and have problems.... well you knew that could happen, right?
                      You know, its funny.
                      In the past, I've had *serious* trouble with ATI hardware, so I stuck exclusively with *anything but* ATI... had all kinds of crazy things (read TRIDENT) back in the day since the ATI drivers *never* worked on ANY platform. Once in a while, I got stuck with an ATI for one reason or another, and every time, it was a nightmare -- right up to the R300 IGP. So for my stuff, it was 3dfx (voodoo3-2000) back when they still existed -- worked like a dream (and that card is still in use and works great -- for what it is anyhow...), and then nvidia.

                      But a funny thing... the nvidia experience wasn't without any bumps. There have been some nightmarish problems with their stuff too -- obscure hardware problems, driver problems, general lack of support.

                      However, a major turning point was met when ATI got gobbled up by AMD. A neat thing is that even hardware that *NEVER* worked with ANYTHING slowly started coming in to shape -- that R300 IGP that was a previous cause of sleepless nights now works like a dream -- no tinkering required. And the way that R6xx drivers have shaped up... well my previous workhorse with an nvidia card has now been relegated to server-duty and my primary use-it machine is running with an RHD-3650 on open source drivers. And even with the currently in-development state of the drivers, it STILL is the best experience with a graphics card I've EVER had -- *just works* (at least for what I use it for... which is all that matters to me.)

                      Now I'm guessing that the back and forth flamewars (both sides are guilty of) is due to a few factors;
                      1) When someone makes a decision to acquire/use some particular thing, THEY THEMSELVES did so because they believe that their choice is the best. The unfortunate thing is that some people are very INSECURE and need constant reinforcement that they made the best choice, to the extent that they take OFFENCE when someone says something bad about their choice. It is, unfortunately, human nature. Another unfortunate thing is that what is the best choice for one person MAY NOT BE the best choice for someone else -- and this needs to be understood by those involved in the flame wars.
                      2) One's experiences are affected by their experiences. There can be YEARS of bad (or good) experiences that can affect one's perceptions. My own personal bad experience with old ATI *junk* could have (and did in fact) keep me from using ATI hardware. However, it is hard to miss everything that has happened (and is still happening) on this front. I'm sure that I was, to some degree, influenced by very very *GOOD* experiences with AMD hardware, and so I decided to try a radeon GPU again (although this time with the understanding that it is a work in progress), and have, thus far, been extremely impressed. Perceptions have momentum, and it usually takes a LOT more positive energy to improve someone's perceptions than it takes negative energy to destroy it.

                      The real cool thing for those interested in AMD Radeon GPUs is this; as you've mentioned, "new hardware and linux only play nice every now and then". Well from the *looks* of things (which really CAN change at a moment's notice), this situation seems to be improving drastically. There really is a lot of work that goes into building up the infrastructure that has been missing for quite a long time -- there were missing 3D drivers for R3xx-R5xx series, and for R6xx+, and the whole architecture was a mess, which means that there has been quite a bit of catch-up work going on and development of new cool stuff, like gallium3d. If (actually, lets say WHEN) the catch-up work is done and all the pieces are in place, I predict that it will be much easier/faster for new graphics hardware to be supported in a timely manner than it ever was before, i.e., the resources are spread out all over the place -- on multiple generations of hardware support, on infrastructure, etc., when all these pieces are in place, the same resources may go straight into support for new hardware, which has the potential to greatly accelerate support for new hardware. I *hope* that we can get to the point of having same-day support for new hardware in open source... and it most certainly *is possible*.

                      Comment

                      Working...
                      X