![]() |
![]() | ![]() | ![]() | ![]() | ![]() | ![]() | ![]() |
| Welcome to Windows Vista Forums. Our forum is dedicated to helping you find solutions with any problems, errors or issues you are experiencing with Windows Vista. The Vista forum also covers news and updates and has an extensive Windows Vista tutorial section that covers a wide range of tips and tricks. |
| |||||||
| |
| | #1 (permalink) |
| | MediaElement Performance I've been testing the MediaElement to play some hi-def video I've recorded off my ATSC capture card. The video is in dvr-ms format (MPEG2) and using VMR9 it results in about 50% CPU utilization on my computer (Athlon XP 3200) under WinXP SP2. If I play the same video in the MediaElement it results in 100% utilization and a lot of dropped frames using the June CTP. Will the performance get significantly better in the final release? Would the performance be better under Vista? If none of the above, would I be better off getting a new video card or a new CPU to improve the playback? And what would be the minimum necessary to play hi def video smoothly using the MediaElement? |
My System Specs![]() |
| | #2 (permalink) |
| | Re: MediaElement Performance Hello Nick, Unfortunately, we don't expect performance of the MediaElement to improve a great deal. There is basically an added cost of the advanced compositing of video that WPF provides. We have seen HD videos playing very well on dual core machines with an 256MB memory video card. It is hard to say what the bottleneck on your system is, but I would make sure your video card has enough memory to store the video samples (which can get pretty large with HD videos). As a guideline, I would make sure you are not using the MediaElement inside of a VisualBrush if not strictly necessary, and also that you are displaying the video at its natural resolution. Another option to explore if you really want CPU usage to be as low as possible is to host the Windows Media Player OCX inside of WPF by means of an HwndHost. Our SDK has a sample of this under Interop. Hope this helps. Thanks, Ed "Nick Palmer" <Nick Palmer@discussions.microsoft.com> wrote in message news:44024C83-8409-4E82-AFA8-84F90B097C28@microsoft.com... > I've been testing the MediaElement to play some hi-def video I've recorded > off my ATSC capture card. The video is in dvr-ms format (MPEG2) and using > VMR9 it results in about 50% CPU utilization on my computer (Athlon XP > 3200) > under WinXP SP2. If I play the same video in the MediaElement it results > in > 100% utilization and a lot of dropped frames using the June CTP. > Will the performance get significantly better in the final release? > Would the performance be better under Vista? > If none of the above, would I be better off getting a new video card or a > new CPU to improve the playback? And what would be the minimum necessary > to > play hi def video smoothly using the MediaElement? > |
My System Specs![]() |
| | #3 (permalink) |
| | Re: MediaElement Performance Not sure if this changes anything but I did notice something after I made my initial post. I am using the Nvidia PureVideo MPEG2 codec; when I play the file using the VMR, the codec shows it is using "DirectX VA mode C (idct)" and when using the MediaElement it shows "software YUY2". I tried the intervideo codec and it didn't perform any better so I assume it was using software mode as well although I don't know how to check. So I guess my question is: Is this a bug with the MPEG2 codecs or the MediaElement or is it by design? I really want to be able to have a translucent UI overlay. I've developed a work around that uses VMR9 and a custom allocator/presentor to copy the UI (using RenderTargetBitmap) onto the D3D surface and return HTTRANSPARENT in the WM_NCHITTEST event but it’s not ideal obviously. But I guess that will have to do until I upgrade to a Dual Core processor ![]() Thanks! "Ed Maia [MSFT]" wrote: > Hello Nick, > > Unfortunately, we don't expect performance of the MediaElement to improve a > great deal. There is basically an added cost of the advanced compositing of > video that WPF provides. We have seen HD videos playing very well on dual > core machines with an 256MB memory video card. It is hard to say what the > bottleneck on your system is, but I would make sure your video card has > enough memory to store the video samples (which can get pretty large with HD > videos). > > As a guideline, I would make sure you are not using the MediaElement inside > of a VisualBrush if not strictly necessary, and also that you are displaying > the video at its natural resolution. > > Another option to explore if you really want CPU usage to be as low as > possible is to host the Windows Media Player OCX inside of WPF by means of > an HwndHost. Our SDK has a sample of this under Interop. > > Hope this helps. > > Thanks, > Ed > > "Nick Palmer" <Nick Palmer@discussions.microsoft.com> wrote in message > news:44024C83-8409-4E82-AFA8-84F90B097C28@microsoft.com... > > I've been testing the MediaElement to play some hi-def video I've recorded > > off my ATSC capture card. The video is in dvr-ms format (MPEG2) and using > > VMR9 it results in about 50% CPU utilization on my computer (Athlon XP > > 3200) > > under WinXP SP2. If I play the same video in the MediaElement it results > > in > > 100% utilization and a lot of dropped frames using the June CTP. > > Will the performance get significantly better in the final release? > > Would the performance be better under Vista? > > If none of the above, would I be better off getting a new video card or a > > new CPU to improve the playback? And what would be the minimum necessary > > to > > play hi def video smoothly using the MediaElement? > > > > > |
My System Specs![]() |