no issues on mine, tried madvr and custom presenter, plays fine. I even downloaded it myself and tried that version. Try downloading it again.
please can we restrict discussion to DXVA on this thread, please open another for other bugs, I think this is wise.
This is MD5 checksum of the file I downloaded:
You can use this tool to do it
fciv -md5 filename
I read your post and the screenshots.
I already said when you are using DXVA, the card will go into UVD clock, so it will stay at 300MHz or so on.
That's how DXVA/HW(videos) works. It use UVD clock(status).
The GPU only is solicited when you are playing games or D3D programs.
Now new RX400's card have a new option called "Power efficiency" that can let GPU solicited even you are on desktop, but once you play a video with H/W, it still will go to UVD clock.
I have the same thinking as you described until I researched the UVD clock and how H/W works.
The only problems is if you play games and play videos at the sometime what will it happen.
This is the same question if interesting:
It is ok if you want to see it as a bug, and report to AMD. I just said what I have found.
As for Fan 100% issue , there is a thread here:
I agree it.
Dude, the issue I'm having has nothing to do with Power efficiency and please stop assuming people who post/ask questions have a technical knowledge below yours. You're wasting both of our time giving a solution for something that has nothing to do with my problem.
In all your answers you missed the essential point of the issue I'm having. When I disable HW/DXVA acceleration in media players like VLC or MPC-HC and play a movie, the drivers/my card behave as if HW acceleration was enabled (ie: the GPU change to a DXVA/HW acceleration P state). Furthermore, when I do enable HW/DXVA acceleration in these video players and play a movie, the drivers/card behave as if HW/DXVA acceleration was disabled (ie: the frequency of the core and memory stay at 300/300Mhz aka Idle P state.). Things works but in a reverse way than they should normally.
I wouldn't be asking if the solution was as easy as clicking "Power efficiency" in the Radeon settings... I know what the new Power efficiency feature on the RX4x0 series is.
From your first image, DXVA is disabled and the frequency is 7xx/2000 when you play movie,is this right? If I am wrong at this point please tell me.
What I try to explain is this is not DXVA status(or UVD status)
When it was 300/300 , while you enable DXVA, this is UVD stats in the second image.
Also, if power efficiency option is disabled, GPU/memory clock will rise ,it is not only on playing movie , but also 2D desktop works, and 2D desktop task is no DXVA(DirectX Video Acceleration) related.
I am sure I know what you are saying but we have different thinking on this topic.
As for your scenario ,you are not along , I see the same behavior on my RX460.
But as for it is a bug or not, I think I will open a new thread to discuss this topic.
As for this thread ,this is for DXVA corruption issue, so let us discuss in another thread.
well......no the new driver yet again does not fix the dxva matroska problem as i am now referring to it.
Thats Great ! ! ! Go Go RED Team
1 of 1 people found this helpful
H.264 content may experience blocky corruption when streaming using P2P content players on some Radeon RX 400 series graphics products.
So at last h.264 problem gets a mention.
Not sure why Matroska is being quoted here. My problem can be created while streaming directly from terrestial or satellite diretly to player (DVBviewer). Also I don't use an MKV container for storage but .ts
matroska is how this problem manifests itself on video playback, you get identical corruption on matroska MKV encoded movies see my above posts, encoding your movies as MP4's using identical output modes (high profile 4.1 / 4 reframes / 30mb/sec + ) fixes the problem so its defo only matroska.
I do not understand where Matroska is used in my scenario
Terrestial or Satellite broadcast DVB-S2 or DVB-T2 , then player plays stream. DXVA H.264 does not work, H.265 works.
DVB-T Standard is published as EN 300 744, Framing structure.
The frames use OFDM frame structure.
OFDM frame structure
Clause 4.4 specifies the OFDM frame structure to use for the 2K and 8K transmission modes. The OFDM frame
structure parameters for the 4K transmission mode shall be compliant with the specifications provided in the annex F.
The transmitted signal is organized in frames. Each frame has a duration of TF, and consists of 68 OFDM symbols.
Four frames constitute one super-frame. Each symbol is constituted by a set of K = 6 817 carriers in the 8K mode and
K = 1 705 carriers in the 2K mode and transmitted with a duration TS. It is composed of two parts: a useful part with
duration TU and a guard interval with a duration Δ. The guard interval consists in a cyclic continuation of the useful
part, TU, and is inserted before it. Four values of guard intervals may be used according to table 5.
The symbols in an OFDM frame are numbered from 0 to 67. All symbols contain data and reference information.
Since the OFDM signal comprises many separately-modulated carriers, each symbol can in turn be considered to be
divided into cells, each corresponding to the modulation carried on one carrier during one symbol.
In addition to the transmitted data an OFDM frame contains:
- scattered pilot cells;
- continual pilot carriers;
- TPS carriers.
The pilots can be used for frame synchronization, frequency synchronization, time synchronization, channel estimation,
transmission mode identification and can also be used to follow the phase noise.
The carriers are indexed by k ∈ [Kmin; Kmax] and determined by Kmin = 0 and Kmax = 1 704 in 2K mode and 6 816 in
8K mode respectively. The spacing between adjacent carriers is 1/TU while the spacing between carriers Kmin and Kmax
are determined by (K-1)/TU.
The numerical values for the OFDM parameters for the 8K and 2K modes are given in tables 4 and 5 for 8 MHz
channels, in annex E for 6 MHz and 7 MHz channels, and in annex G for 5 MHz channels.
none of that means anything to me, obviously, whatever is breaking DVB is breaking matroska playback, I find it unlikely these two problems are totally unconnected as they both use DXVA h264.
Let's wait and see if we get any feedback from AMD. Even if they fix this we will probably not get any clues as to what the issue was.
got 20x 720p videos .MKV running 10 in media player classic rendering in dxva and 10 in potplayer average gpu 20% cpu around 50% no issues at all reference rx480
Got the same issues that you guys have at 1080p 8 frames high lvl4 . In pot player everything is ok unless I move slide to move through video. In MPC the video starts flashing then corrupts . Noticed the parts of the video that do not move are ok its the parts that change that corrupt. These video are animated films and this behaviour really stands out.
Looks more like the decoder not getting its timing information correct from the video stream have seen this happen with fault encodes.