madVR - high quality video renderer (GPU assisted)

[mr.duck]

HD7000 series will already be in production. We've been stuck at 40nm for years and the HD7000 will be a massive step to the 28nm node (i.e. expect a huge improvement in power efficiency/performance). Worth waiting for IMO.

[golagoda]

HD7000 series will already be in production. We've been stuck at 40nm for years and the HD7000 will be a massive step to the 28nm node (i.e. expect a huge improvement in power efficiency/performance). Worth waiting for IMO.

Those are the same types of things that were said about the AMD Bulldozer CPU's... and look how that turned out :sly:

[iSeries]

Hi,

Having a strange issue with 0.78. Double clicking m2ts's directly will play smooth as ever, but if I double click the index.bdmv or the mpls file then I get MASSIVE dropped frames, 3-5 a second. Going back to the previous version I have (0.75) and everything is perfect again.

Anyone got any ideas?

[mindz]

First off, great work so far on the madVR project, is there a way to donate for all your hard work?

Ive got a question about deinterlacing and getting the 2x FPS deinterlacing (im such a fan of doubling framerates ;-))!

Ive got an ATI card (5770) and i got Vector Adaptive deinterlacing enabled in CCC. When i start a PAL DVD (interlaced) in madVR i dont get the deinterlacing in double frame rate. Ive tried setting Bob (which is a auto 2x FPS, right?), but this didnt work. I used to get the double framerate when using yadif 2x in ffdshow, but that doesnt work anymore either :(. Anyone got any ideas?

Oh, and im also not getting any sound when starting a DVD (i have to switch decoder in ffdshow, libmpeg -> libav), then it works. Anyone ideas on this one? (this was a ReClock problem, its fixed!)

[Gser]

First off, great work so far on the madVR project, is there a way to donate for all your hard work?

Ive got a question about deinterlacing and getting the 2x FPS deinterlacing (im such a fan of doubling framerates ;-))!

Ive got an ATI card (5770) and i got Vector Adaptive deinterlacing enabled in CCC. When i start a PAL DVD (interlaced) in madVR i dont get the deinterlacing in double frame rate. Ive tried setting Bob (which is a auto 2x FPS, right?), but this didnt work. I used to get the double framerate when using yadif 2x in ffdshow, but that doesnt work anymore either :(. Anyone got any ideas?

Oh, and im also not getting any sound when starting a DVD (i have to switch decoder in ffdshow, libmpeg -> libav), then it works. Anyone ideas on this one? (this was a ReClock problem, its fixed!)

Yeh interlaced support seems to be a bit shaky e.g. MadVR upscales interlaced footage before the GPU deinterlaces it which screws up the picture. Doubt anything can be done about it, don't know about bobbing support though.

[dansrfe]

The pause-playback problem not only causes the video to go super fast to catch up but it also terminates the audio a couple of mins after minimizing.

[bjd]

On a Windows XP machine, what would cause the display value (ctl+j first value) to fluctuate between 45-75hz causing playback issues.

The PC resolution is 1824x1016@50hz which is a Nvidia custom resolution for a 1920x1080p HDMI TV that allows for overscan.

When the PC first starts up the the display rate shown in MadVR is 50hz +/- 0.25% and playback is rock solid however after a while this then starts to vary like mad as described above.

Reclock during this time always says the refresh rate is 50hz as does the TV and the Nvidia driver/Windows. A reboot will always cure it, but i am trying to get to the bottom of this.

@madshi - where/how does MadVR get its refresh rate timing from please ?

[sneaker_ger]

Ive got a question about deinterlacing and getting the 2x FPS deinterlacing (im such a fan of doubling framerates ;-))!

Ive got an ATI card (5770) and i got Vector Adaptive deinterlacing enabled in CCC. When i start a PAL DVD (interlaced) in madVR i dont get the deinterlacing in double frame rate. Ive tried setting Bob (which is a auto 2x FPS, right?), but this didnt work.

What happens when you use the standard setting: "Use automatic deinterlacing" = on ?

I used to get the double framerate when using yadif 2x in ffdshow, but that doesnt work anymore either :(.

ffdshow's yadif is pure software. There should be no way that madVR or the CCC settings could affect that. Are you sure you've set up ffdshow correctly?

Yeh interlaced support seems to be a bit shaky e.g. MadVR upscales interlaced footage before the GPU deinterlaces it which screws up the picture. Doubt anything can be done about it, don't know about bobbing support though.

I doubt madVR can do anything BUT bobbing.

[Gser]

What happens when you use the standard setting: "Use automatic deinterlacing" = on ?



ffdshow's yadif is pure software. There should be no way that madVR or the CCC settings could affect that. Are you sure you've set up ffdshow correctly?



I doubt madVR can do anything BUT bobbing.

Nevermind, I was using version 0.77, didn't notice madvr was updated. Any info about DXVA2 deinterlacing would be interesting.

[mr.duck]

Those are the same types of things that were said about the AMD Bulldozer CPU's... and look how that turned out :sly:
No of course not. Bulldozer is 32nm and we had 32nm already. Bulldozer is a brand new CPU architecture, not a massively die shrunk version of an existing CPU.



Ive got a question about deinterlacing and getting the 2x FPS deinterlacing (im such a fan of doubling framerates ;-))!
Try drivers from 2010 like version 10.11

[nevcairiel]

not a massively die shrunk version of an existing CPU

Thats exactly why the 7000 series won't be all that great. Their architecture is old and over-used, they need something fresh - just shrinking it won't help all that much. (Unless i missed the news on a new architecture)

On the other hand, they shouldn't take too much after Bulldozer - less performance then previous architecture is not good for selling stuff.

[sneaker_ger]

No of course not. Bulldozer is 32nm and we had 32nm already. Bulldozer is a brand new CPU architecture, not a massively die shrunk version of an existing CPU.

That was a side blow to Bulldozer, as its efficiency sucks big time, while people were expecting an Intel killer. Regardless of the production process.

[mr.duck]

Thats exactly why the 7000 series won't be all that great. Their architecture is old and over-used, they need something fresh - just shrinking it won't help all that much. (Unless i missed the news on a new architecture)
I'm surprised. 6850 performs well and is priced well for example. Imagine it with 1/2 the power consumption. "won't be all that great" doesn't come close to my expectations.

BTW, HD7000 series will be 4way VLIW architecture across the board (6800 and below is 5 way VLIW - less efficient). Apart from the high end 79xx card wich will be a new architecture called GCN

[nevcairiel]

I'm surprised. 6850 performs well and is priced well for example.

I'm not saying the cards are bad, but AMD has been milking the same base architecture for years, at some point you just hit a wall. They can certainly hide that wall with going to new processes and some jiggery pokery, but something fresh would do them good, too.

PS:
Half power consumption is rather unlikely.

[whitestar999]

I'm surprised. 6850 performs well and is priced well for example. Imagine it with 1/2 the power consumption.
power consumption for an existing architecture can't be halved without changing the architecture unless the architecture was very inefficient to begin with which is certainly not the case.

[mr.duck]

power consumption for an existing architecture can't be halved without changing the architecture
Wrong. After a die shrink you can operate on a lower voltage too. Expect noticeable efficiency gains as a bare minimum.

From one generation to the next, it has been normal to get twice the GPU for the money. 40nm to 28nm is a bigger step than normal, so it's reasonable to expect more than twice the GPU.

If my calculations are correct, the 6870 would be only 44% of it's current die size at 28nm compared to 40nm.

[6233638]

Kepler seems a lot more exciting than the 7000 series, and definitely the cards to be waiting on if you plan on gaming. (AMD drivers have been disastrous for a number of big releases recently, Rage in particular)

[madshi]

I did some performance (GPU load) testing with madVR, LAV CUVID and EVR and I'm kind of amazed how good EVR is.
For the test I used a 1080i60 TV recording of a hockey game and I had to overclock my GPU a little. To be precise I've used the P8 (Video) mode of my NVIDIA card but upped both Core/Shader and Memory clocks from 405/810 and 162MHz to 540/1010 and 192MHz.
This resulted in both madVR and LAV CUVID having about the same GPU load ~80% while EVR was around 50%.
Then I decided to lower my upscaling settings in madVR to Bilinear for both chroma and luma (even though the luma wasn't important since both video and display were 1920x1080) and this lowered madVR's usage to ~70-75% which is still ~25% above EVR.
With which renderer did you test LAV CUVID? madVR generally does consume more GPU than EVR, because madVR does many things via pixel shaders, which is more "expensive" than using hard wired video processing circuits. Interesting would be:

(1) "LAV CUVID+EVR" vs. "DXVADecoder+EVR"
(2) "LAV CUVID+madVR" vs. "SoftwareDecoder+madVR"

I would expect you to get somewhat similar GPU numbers for "LAV CUVID+EVR" and "DXVADecoder+EVR". Not sure about (2). Depends on whether video decoding is counted as GPU usage. If it's not, I would expect you to get somewhat similar GPU numbers for "LAV CUVID+madVR" and "SoftwareDecoder+madVR", too. But of course the GPU numbers should be higher for (2) than for (1).

The strange thing is that only some files seem to drop down to bob deinterlacing. (and "progressive" ones seem to disable the deinterlacer altogether when using ffdshow for decoding)
I guess those that don't look like Bob are probably soft-telecined, so that DXVA does not really do anything but let the frames pass untouched (weave). Just a guess, though.

Is there anything that can be done about this? I assume most people will want to play back film content at 24p. (either 23.976 or 24.000) I'm not sure that I have any video content—certainly nothing interlaced.
Sure, in the long run there are things that can be done. One step at a time.

A random grab from that sample.
Can you please grab an unscaled frame? We want to check out deinterlacing quality, so there's no need to scale at all.

1) MPEG2 50i with LAV Video decoder, NV12 output, EVR: works fine, looks ok, very smooth, clearly 50p
2) MPEG2 50i with LAV Video decoder, NV12 output, madVR: it does work, looks very bad, not smooth, clearly it is being deinterlaced

Radeon 5750 using latest drivers (11.10).
That's really bad. Can I have a log, please? Also, can you step through the video frame by frame with both EVR and madVR? Maybe when looking through both frame by frame you can see why it's smooth with EVR and not smooth with madVR? Maybe madVR repeats each frame twice to achieve double rate, while EVR really produces fluid unique frames?

It might be worth trying drivers 10.11. But I'd really be interested to hear your results when doing frame-by-frame stepping with your current drivers.

Since nobody will ever answer to this, where could be the issue? Should I take any options' screenshots?
Maybe creating and uploading a madVR log would help. When creating the log, please try to reproduce the problem with the first pause, then when the subtitle disappears, close the media player at once. That will help making the log nicely short. Thx.

Why 10.11 drivers? Did you mean 11.10? I'm already using 11.10 which are the latest non-beta.
No, I meant 10.11, because other madVR users already reported that these old drivers help get smooth deinterlacing results, while newer drivers seem to make problems for some users (but not all).

But please check out my reply to mr.duck (see above). I'd like to get a log file from you, too. And I'd like you to try frame stepping to get to the bottom of the non-smoothness, too. Thx.

Can you try this sample (http://www.filesonic.com/file/11998573/Lia_-_Tori_no_Uta_long_version.mkv) with my filter set.

PotPlayer 1.5.29996
(1) Haali Media Splitter 1.11.288.0
(2) LAV Video Decoder 0.38
(3) Madshi Video Renderer 0.78
(4) Audio Codec/Transform
(5) DirectSound Audio Renderer
Does it only occur with Haali but not with LAV Splitter? Does it only occur with LAV Video Decoder but not with other decoders? Does it only occur with madVR v0.78 but not with older versions? Does it only occur with PotPlayer 1.5.29996 but not with other versions? I'm asking because you're asking a lot here. Finding, downloading, installing and testing these exact versions of all the various filters/player would take me quite some time, and I don't really want to mess up my own filter configuration.

I've tried once again, this time in fullscreen exclusive mode (tested only in windowed mode before), still no problem. I'm testing in XP, though, where there's no DX11 presentation mode. What happens if you use DX9 instead of DX11 in the madVR settings? Does that change anything? The DX11 path is known to be not as stable (yet) as the DX9 path.

Seems to work mostly as intended for Intel HD3000 but I've only tested with a MPEG2 60i SD source so far, MPC-HC DXVA decoder used.
Some unexpected behavior: When i hit ctrl+shift+alt+d I see no combing yet the annotation says de-interlacing is off.
If the annotation says deinterlacing is off, then it is off. So it doesn't seem to work as expected. Either madVR thinks activating deinterlacing isn't necessary (this can happen if the movie was fully soft-telecined and all frames are marked as progressive, even though the container is 60i), or activating deinterlacing failed. Seeing that other Intel users report that deinterlacing failed to work at all for them, my current guess is that the current madVR version isn't able to get deinterlacing to work at all on Intel GPUs. I'll check this out on my Sandy Bridge laptop. To be honest, my plan was to look at Intel in one of the next versions. For now I've concentrated on ATI + NVidia, because the majority of madVR users are using ATI + NVidia. But don't worry, I'll get to Intel soon enough.

i am having an unusual problem.some 1080p videos run fine on my laptop c2d 2ghz mobilty radeon 4570 with screen resolution 1366*768(60hz) but when i try to run the same video from laptop through hdmi on a 1080p(60hz) LG LCD TV there video lags with frames dropping continuously.backbuffer & render queue are almost always 1-3/8 while decoder & other queue are full.as i understand playing 1080p video on a 1080p screen should be the easiest case scenario or am i missing something.
@namaiki,card ram is 256mb & yes there is a difference when i make video window smaller.gpu-z shows card load at ~70% when window is resized to ~480*360.how can i see the gpu-z window in full screen video playing?
The bigger the output resolution, the more GPU RAM madVR needs. It might be possible that 256MB is not enough for 1080p playback, I'm not sure. I plan to make the queue sizes adjustable soon, so you can save GPU RAM. That might make it work for you.

At least on Intel HD Graphics (1st gen Core iX-CPUs) it doesn't seem to work at all, it just says "deinterlacing failed".
Log for "deinterlacing failed" message on Intel HD graphics: http://www.mediafire.com/?n8o0jz0umd4xqw1
I've not tested deinterlacing with Intel yet. Plan to do so (and make it work, if possible) soon.

Curious, it says "VideoProcessBlt failed (8007000e)", and thats the HRESULT for Out of Memory.
Yeah, that's interesting.

I've just tried to use madVR with PotPlayer and LAVFilters, and its doing the strangest thing... Win7 64bit, 32bit apps, Nvidia GTX260 card...

PotP with LAV and EVR CP works nice, but if I select madVR as renderer I just get audio and no decoder is loaded. All colourspaces seem to be ticked in LAV, and I tried some other settings in there... What am I missing?
So do you get a picture or not? madVR has decoders on its own, but if they get used, you should see a picture. I guess you see no picture? I'd suggest that you try a different media player, just as a cross test. Also try getting the latest PotPlayer version.

I also had issues when playing back on my new 1080p projector. Using JRMC, video was unwatchable on the projector (HDMI), but played well on my 1360x768 LCD screen (connected through VGA). Tried both 24 and 60Hz on the projector. Stuttered even when video was in a window, on both 720p and 1080p files. Then I went into the MadVR settings and enabled decoding, stuttering gone and perfectly smooth playback, even with 3840x1080p side-by-side files :)
Good to hear that the internal decoders seem to work well for you. As an alternative you could also try LAV Video Decoders, if you prefer using external decoders. What you describe sounds like a problem with the external decoders you'd been using before.

Question: Will DXVA ever work with MadVR? Saw it was asked a while ago, but couldn't find an answer...
DXVA *deinterlacing* does work with madVR, just not DXVA *decoding*. Whether DXVA decoding will work in the future or not I don't want to comment on at the moment.

I report this bug (if its a bug that is) that when i use madVR and try to pause anime with mouse click in MPC-HC the subtitles dont seem to pause somtimes,
but when i press on Backspace to pause it does pause the subtitles.
What does "don't seem to pause" mean exactly? Do they continue to play, line after line?

Interesting. Any reason it's this way? I find left clicking to be highly unintuitive compared to right clicking and I would greatly prefer right clicking.
Showing different context menus for left and right click is not that unusual for tray icons. The right click context menu has more menu items. The madVR left/right context menus have a clear logic to them:

The left click menu is just a menu for the currently running media player, where you can do all kinds of things to influence the media playback. The right click menu contains all things you might ever want to do, including closing the tray icon process, changing tray icon settings etc.

The reason why the right click menu has so many sub menus is that it contains more functionality. Having all menu items of the right click menu in one level would be a usability catastrophe. The left click menu is so nicely simple because it limits itself to only control the current media player, nothing else.

[madshi]

What's the current recommendation for video playback hardware with MadVR? I currently have an ATI 5450 and older Intel 6600 dual core CPU, this is good enough for the majority of material (mainly HD), only the highest bitrate stuff causes some drops. I'm looking to build a 2nd HTPC for another display and thinking of re-using the 5450 there and upgrading the main HTPC in the theater room. Seems like both ATI and nVidia have pluses and minuses (as always). Are any of the new integrated Intel graphics worthy of consideration? Running Win7, and mostly HD material, not even sure MadVR is really needed, but I do like to go for the best solution even if it's not something I can really notice. :-) Thanks for any advice.
If you can wait, I'd suggest to wait for the new 28nm GPUs coming out early 2012. They should significantly improve the power per watt ratio.

As to ATI vs. Nvidia, that's really a tough decision. NVidia seems to have big problems with presentation glitches in exclusive mode. ATI doesn't seem to have that problem, but then ATI has problems on its own. So it's really hard to decide...

I just managed to reproduce the disappearing subtitles multiple times, sadly I only tested windowed mode there.

After restarting the player to also test fullscreen exclusive mode I couldn't reproduce it any more (neither windowed nor exclusive). The only thing I could think of that was different the first time is that Reclock was still doing its measurements (red/green switching icon). But why should Reclock cause the subtitles to disappear? Maybe the bug happens just extremly random.
Still can't reproduce it here.

Although I found one thing: The subtitles disappear in fullscreen exclusive mode as soon as the seek bar is shown. But I assume this is by design like with shown stats?
Yes.

But they are all done with the files from here (http://www.avsforum.com/avs-vb/showthread.php?t=1157287), do these work for you?
Sure, I've been testing with those, too.

is it ok to set the following options as enabled in the Video Settings for ATI cards: Use automatic deinterlacing and Pulldown detection
Yes, these are recommended.

I can reproduce this problem. In my testing, subtitle will disappear in a short time when paused for the first time since seeking. Next pauses will be fine.
Ok, tried that, but doesn't happen for me. I can seek and pause, subtitles stay. Using MPC-HC.

Any ideas why this sample (http://www.mediafire.com/?x2h4b6zjv98ej33) displayed with green bar at the bottom with default MadVR settings until I enable "use managed upload textures (XP only)"?
This doesn't happen on my PC. Also MPC-HC, XPSP3, but ATI 6850. It *could* be an NVidia driver issue. But then why does it not happen with the internal madVR decoder? That's really weird. Maybe nevcairiel can have a look at that? He's on Nvidia, but I don't think he uses XP.

How do you read the Decoder Queue, Upload Queue, Render Queue, and Present Queue in the statistics?
You open your eyes and read the white text at the top left of the screen, after having pressed Ctrl+J... :devil:

Ok, seriously. madVR has several queues where it stores frames that were already decoded, uploaded to GPU RAM, fully rendered and dithered to output resolution etc and passed on to Direct3D for presentation. Ideally, all queues should be full.

I found out why I thought 0-255 was the correct setting, it seems whatever is set on the AMD control panel will be the "correct" setting when I open MPC-HC. If I open it while I got 0-255 already set, that will be correct, and changing to 16-235 while the video is playing will get me gray blacks, and vice versa. Still using 11.10 drivers, haven't tried rolling back or switching to NV yet.
Haha. Yes, that's true. madVR analyzes the GPU behaviour once during madVR startup (so you don't have to restart the media player, just reloading the video should do the trick). If you change the GPU driver settings while madVR is already loaded, madVR will not adjust to that, until the madVR instance is destroyed and recreated.

I've aborted my deinterlacing tests for now because it seems that 0.78 broke windowed mode for me, even with progressive video.

The files 3, 4, 5 and 6 from here (http://forum.doom9.org/showthread.php?t=159486) and the "WipEout HD Trailer" (MKV, H.264, 1080p59, 170 MB) and "Devil May Cry Gameplay" (WMV, VC-1, 1080p59, 266 MB) for which I can't find download links right now give me many dropped and delayed frames each second in windowed mode. Going back to madVR 0.77 and everything is fine again. Fullscreen exclusive mode works fine in both versions.

All files are correctly identified as progressive. Also disabling deinterlacing in the settings has no affect.
The only thing that they have in common is a frame rate of 59 fps. Maybe the new version has problems with that high frame rates?
The only thing that comes to my mind right now is that maybe your GPU is downclocking with v0.78 but not with v0.77? I don't know why it should. But then I don't understand what else could be going on. Your render queue is the problem, so it seems that your GPU isn't rendering fast enough. Which is weird, since your GPU should be plenty fast enough.

If you're sure that the GPU clock is ok then please create a short log with v0.77 (no frame drops) and one with v0.78 (frame drops). Maybe 20 seconds of playback each. Thanks.

I've also seen the same thing with a Nvidia 465. I've also had this same problem using LAV CUVID with earlier madVR versions. Everything is OK in full-screen exclusive.
Have the same problem, in windows mode, every video file drops frames constantly, in Exclusive mode almost none.
"the same thing/problem" is rather relative. It could mean something totally different to you than it might mean to me. Could you please describe exactly which problem you have? Having smooth playback in exclusive mode while you have problems in windowed mode is not usual in itself. There's a reason why I implemented exclusive mode, after all. The key question for me is: Did things get noticeable worse for you with v0.78? Because that's what cremor is reporting.

HD7000 series will already be in production. We've been stuck at 40nm for years and the HD7000 will be a massive step to the 28nm node (i.e. expect a huge improvement in power efficiency/performance). Worth waiting for IMO.
I agree that the new 28nm GPUs (both ATI + NVidia) are probably worth waiting for at this point in time. Ok, we don't know exactly when which cards will be released. But I also expect a big jump in performance per watt. Which is one of the most important things for an HTPC.

Having a strange issue with 0.78. Double clicking m2ts's directly will play smooth as ever, but if I double click the index.bdmv or the mpls file then I get MASSIVE dropped frames, 3-5 a second. Going back to the previous version I have (0.75) and everything is perfect again.
Weird. Can you please try 0.76 and 0.77, just to find out which version exactly broke it for you? Also please try different splitters and decoders, just to give me more information. Thanks.

First off, great work so far on the madVR project, is there a way to donate for all your hard work?
Not until madVR v1.0 is released.

Ive got a question about deinterlacing and getting the 2x FPS deinterlacing (im such a fan of doubling framerates ;-))!

Ive got an ATI card (5770) and i got Vector Adaptive deinterlacing enabled in CCC. When i start a PAL DVD (interlaced) in madVR i dont get the deinterlacing in double frame rate. Ive tried setting Bob (which is a auto 2x FPS, right?), but this didnt work. I used to get the double framerate when using yadif 2x in ffdshow, but that doesnt work anymore either :(.
Are we talking movie content or native video content (e.g. music concerts or sports)? Double rate is the right deinterlacing method for native video content. For movies things are quite different.

Yeh interlaced support seems to be a bit shaky e.g. MadVR upscales interlaced footage before the GPU deinterlaces it which screws up the picture.
Not true. Older madVR versions required you to deinterlace before sending the frames to madVR. With older madVR versions the GPU never deinterlaced at all. With v0.78 madVR performs deinterlacing via DXVA and scales afterwards, of course.

On a Windows XP machine, what would cause the display value (ctl+j first value) to fluctuate between 45-75hz causing playback issues.

The PC resolution is 1824x1016@50hz which is a Nvidia custom resolution for a 1920x1080p HDMI TV that allows for overscan.

When the PC first starts up the the display rate shown in MadVR is 50hz +/- 0.25% and playback is rock solid however after a while this then starts to vary like mad as described above.

Reclock during this time always says the refresh rate is 50hz as does the TV and the Nvidia driver/Windows. A reboot will always cure it, but i am trying to get to the bottom of this.

@madshi - where/how does MadVR get its refresh rate timing from please ?
Ouch, this is bad. madVR has a very complicated logic in it which reads the current VSync scanline from the GPU once per millisecond. Then it looks at all the collected scanline data and tries to do some very clever math to calculate the exact refresh rate. It seems that in your case the scanline data sometimes allows different interpretations (50Hz and 75Hz). I've never heard of such a problem before. This will be hard to fix. Does this occur in windowed and/or fullscreen exclusive mode?

I doubt madVR can do anything BUT bobbing.
madVR v0.77 and older versions didn't support deinterlacing *at all*. You were required to deinterlace before sending the video to madVR. madVR v0.78 does not deinterlace itself, it lets DXVA do it. And DXVA usually does better than using Bob.

[sneaker_ger]

madVR v0.77 and older versions didn't support deinterlacing *at all*. You were required to deinterlace before sending the video to madVR. madVR v0.78 does not deinterlace itself, it lets DXVA do it. And DXVA usually does better than using Bob.

I meant "bobbing" as in "some kind of double rate deinterlacing", not some specific bob algorithm.

[nevcairiel]

"Bob" is called that way because the image bobs up and down if its used. Its not a category of algorithms that all do the right thing and use both fields to create separate images. :p

[sneaker_ger]

"Bob" is called that way because the image bobs up and down if its used. Its not a category of algorithms that all do the right thing and use both fields to create separate images. :p

Oh noez. How shall I call it from now on, then? I always used it as a synonym for deinterlacers that preserve the full temporal resolution.

[Nevilne]

madshi, you were mentioning improving 24 fps clips on 60hz screens.
May I suggest an option to make 50% frame blends?
(frame1)-(frame1)-(50%frame1+50%frame2)-(frame2)-(frame2)-
(frame3)-(frame3)-(50%frame3+50%frame4)

[madshi]

Oh noez. How shall I call it from now on, then? I always used it as a synonym for deinterlacers that preserve the full temporal resolution.
The funny thing is, in the CE (consumer electronics) world video deinterlacing pretty much *always* preserves the full temporal resolution. Because of that there's not really a word for it (that I know). I consider not preserving the full temporal resolution a bug, and it usually only happens in the HTPC world. In the HTPC world people often use "double rate deinterlacing" to name the correct solution. So maybe you want to use that instead of "bobbing". But if you talk to CE engineers, they will probably not know what you mean by "double rate deinterlacing".

(Just to avoid misunderstandings: All of the above only applies to native video content. Native movie content is a whole different thing.)

madshi, you were mentioning improving 24 fps clips on 60hz screens.
May I suggest an option to make 50% frame blends?
That is the solution I had in mind. Just haven't decided yet whether I'll blend in linear or gamma light. Linear light should be better, but it's also more expensive.

[Gser]

I read up on DXVA deinterlacing and it uses the deinterlacing your GPU supports (instead of having a native deinterlacing method). DXVA can even support 3:2 pulldown if it is implemented in the renderer. Imo best feature EVAR!

[derp]

Maybe creating and uploading a madVR log would help. When creating the log, please try to reproduce the problem with the first pause, then when the subtitle disappears, close the media player at once. That will help making the log nicely short. Thx.
How can I do this? I have no clue.

[sneaker_ger]

There's a file called "activate debug mode.bat" in the madVR folder. It will exchange "madvr.ax" with "madvr [debug].ax". MadVR will now create a log file on your desktop. Zip it and upload it to something like mediafire.

[bjd]

Ouch, this is bad. madVR has a very complicated logic in it which reads the current VSync scanline from the GPU once per millisecond. Then it looks at all the collected scanline data and tries to do some very clever math to calculate the exact refresh rate. It seems that in your case the scanline data sometimes allows different interpretations (50Hz and 75Hz). I've never heard of such a problem before. This will be hard to fix. Does this occur in windowed and/or fullscreen exclusive mode?


Thanks for the reply and info - It occurs in both. I also run TightVNC server to access the HTPC remotely from a laptop running at 1680x1050@61hz and I am just wondering whether this is causing the scanline data issue. I have just upgraded TightVNC to the latest version to see if this makes any difference and will do some further testing to see if I can isolate the issue.

[derp]

Maybe creating and uploading a madVR log would help. When creating the log, please try to reproduce the problem with the first pause, then when the subtitle disappears, close the media player at once. That will help making the log nicely short. Thx.
Ok, here it is.

http://www.mediafire.com/?hqbsevc64eebesb

There's a file called "activate debug mode.bat" in the madVR folder. It will exchange "madvr.ax" with "madvr [debug].ax". MadVR will now create a log file on your desktop. Zip it and upload it to something like mediafire.
Thanks.

[golagoda]

No of course not. Bulldozer is 32nm and we had 32nm already. Bulldozer is a brand new CPU architecture, not a massively die shrunk version of an existing CPU.




Try drivers from 2010 like version 10.11

I meant how people hyped it up so much like that, saying how much it would beat the i5 2500k and it was one of the most disappointing things ever and underperformed compared to their Intel counterparts by heaps :p Well enough of that, this is hardly the place for hardware talk.

[pankov]

With which renderer did you test LAV CUVID? madVR generally does consume more GPU than EVR, because madVR does many things via pixel shaders, which is more "expensive" than using hard wired video processing circuits. Interesting would be:

(1) "LAV CUVID+EVR" vs. "DXVADecoder+EVR"
(2) "LAV CUVID+madVR" vs. "SoftwareDecoder+madVR"

I would expect you to get somewhat similar GPU numbers for "LAV CUVID+EVR" and "DXVADecoder+EVR". Not sure about (2). Depends on whether video decoding is counted as GPU usage. If it's not, I would expect you to get somewhat similar GPU numbers for "LAV CUVID+madVR" and "SoftwareDecoder+madVR", too. But of course the GPU numbers should be higher for (2) than for (1).

For my previous tests of LAV CUVID (LAV Video in CUVID mode to be precise) I used madVR but now I used both EVR and madVR.

Here are the results of all the combinations I could think of
http://imageshack.us/g/31/lavvideocuvidadaptivehq.png/
or you can download all of them as one archive directly from here
http://www.mediafire.com/?au9ouh2w8b65z7x

From what I saw I still think that madVR's Deinterlacing is too GPU intensive - a jump from from 24% to 79% GPU Load seams a lot.
At the same time playing a progressive 1080p video (not the same one though) with LAV Video (Soft mode) and EVR takes up 28% while the interlaced video + EVR was taking 45% which is only 17% above the progressive one ... and it's far from madVR's increase of 55%. I do understand that some of the difference could be from the different video files but sadly I don't have (don't know how to do it) the same video both interlaced and progressive.

One more thing I noticed is that LAV Video (CUVID - Adaptive + HQ Proc) + madVR (Deint=OFF) uses 4-5% less GPU than LAV Video (Soft) + madVR (Deint=ON).
Does this mean that there is some room for improvement?

I hope I didn't bore you with my tests and you'll be able to improve madVR even more
:thanks: for your great work

[jmonier]

"the same thing/problem" is rather relative. It could mean something totally different to you than it might mean to me. Could you please describe exactly which problem you have? Having smooth playback in exclusive mode while you have problems in windowed mode is not usual in itself. There's a reason why I implemented exclusive mode, after all. The key question for me is: Did things get noticeable worse for you with v0.78? Because that's what cremor is reporting.

Sorry I wasn't clearer, but what I was trying to say was that I'm seeing essentially the same problem under essentially the same conditions that cremor was reporting. That is, if I wrote a description of my problem, I would be writing something very similar to what cremor wrote.

For me, windowed mode was fine (no dropped frames) prior to 0.78 with one exception: If I use LAV CUVID (either stand-alone or combined), I get the same thing with older madVR versions. So the key point seems to be that if hardware de-interlace is used either via CUVID or DXVA 2 on either AMD or NVidia I get lots of dropped frames in windowed mode only. This leads me to believe that madVR is affected somehow by the hardware de-interlace.

I have no problem with using full-screen exclusive if the hardware or driver imposes some limitation that requires it, but I thought you'd like to know about the problem in any case.

[mr.duck]

That's really bad. Can I have a log, please? Also, can you step through the video frame by frame with both EVR and madVR? Maybe when looking through both frame by frame you can see why it's smooth with EVR and not smooth with madVR? Maybe madVR repeats each frame twice to achieve double rate, while EVR really produces fluid unique frames?

It might be worth trying drivers 10.11. But I'd really be interested to hear your results when doing frame-by-frame stepping with your current drivers.
http://www.mediafire.com/?dgd5p0mt185rh59

I checked each frame. EVR looks like it's 50p and madVR looks like it's 25p x2. Each frame happens twice with madVR. EVR really does produce fluid unique frames!

Something to do with the source filter reporting 25fps? (I've switched from LAV splitter to built in MPC splitter now, and all is exactly the same).

Source material is UK digital TV in case you were wondering. MPEG2 50i.

[RBG]

Does it only occur with Haali but not with LAV Splitter? Does it only occur with LAV Video Decoder but not with other decoders? Does it only occur with madVR v0.78 but not with older versions? Does it only occur with PotPlayer 1.5.29996 but not with other versions?

This error occurs with every filter/player combination I have, including Haali, AVs, LAV as a splitter and LAV video, Potdecoder, FFdshow as a video decoder, madVR 0.77, madVR 078, LAV audio, FFdshow audio, Podecoder audio and last three stable releases of potplayer(29996, 29917, 29795)


I'm asking because you're asking a lot here. Finding, downloading, installing and testing these exact versions of all the various filters/player would take me quite some time, and I don't really want to mess up my own filter configuration.


I know, but I am not the only one who gets this annoying bug, though since you're testing madVR in XP, then I guess, there is no reason to reproduce the same filter set, because I am using Window 7 x64.


What happens if you use DX9 instead of DX11 in the madVR settings? Does that change anything? The DX11 path is known to be not as stable (yet) as the DX9 path.

I have already tried DX9, but the result is same.

[Hypernova]

Ok, tried that, but doesn't happen for me. I can seek and pause, subtitles stay. Using MPC-HC.

Just in case you're still keeping track of this, madshi. I cannot reproduce this anymore. It is true that I randomly have subtitles disappear on pause once in awhile. It's not something I'm bothered by though.

[Skwelcha]

Just in case you're still keeping track of this, madshi. I cannot reproduce this anymore. It is true that I randomly have subtitles disappear on pause once in awhile. It's not something I'm bothered by though.

Happnes to me sometimes too. but like you said, it's not that much of a bother.

[Mr Alpha]

(Unless i missed the news on a new architecture)Yes, I believe you may have. (http://www.anandtech.com/show/4455/amds-graphics-core-next-preview-amd-architects-for-compute)

[noee]

Re: the upload/render queue problems with Windowed Mode....I've found a couple of interesting things:

1. It doesn't happen on my Llano machine, Windowed and FSE mode are both maxing all queues with deint=ON. That machine is, however, using an older driver version, from July.
2. Even with that driver, my HD6570 playback machine still exhibits the starved upload/render queues in windowed mode. Again, FSE mode, the queues are maxed.
3. On the 6570 machine, if I disable the ESVP option in CCC, madVR shows a Deinterlace Failure in red warning. ESVP seems to matter not on the Llano machine.
4. While testing, I'm using ATI Tray Tools with a profile to keep the card at max clocks.

6570 machine is with CCC11.10 (or any previous). 10.xx series drivers will not support the 6xxx series cards, so testing that is not possible.

So, madshi, fwiw, if you've still got that log I sent you the other day, it shows the starved queues in Windowed mode and then the switch to FSE mode where they max out and all is well.

[nevcairiel]

Yes, I believe you may have. (http://www.anandtech.com/show/4455/amds-graphics-core-next-preview-amd-architects-for-compute)

I'm aware of those, but as i understood, it won't be used for the majority of the HD7000 series, but only for the 79xx high end models.

IMHO, having a architecture disparity in one line is really bad and will confuse alot of people (although its not the first time AMD did it). :)

[cremor]

The only thing that comes to my mind right now is that maybe your GPU is downclocking with v0.78 but not with v0.77? I don't know why it should. But then I don't understand what else could be going on. Your render queue is the problem, so it seems that your GPU isn't rendering fast enough. Which is weird, since your GPU should be plenty fast enough.

GPU is at max clock rate with both versions. Activity is also quite equal according to the CCC, 36% with 0.77 and 38% with 0.78.

If you're sure that the GPU clock is ok then please create a short log with v0.77 (no frame drops) and one with v0.78 (frame drops). Maybe 20 seconds of playback each. Thanks.

Hope this helps:
http://www.mediafire.com/?7ek3ubk645kvv4f

edit:
Just managed to play the Devil May Cry clip with full queues in windowed mode with 0.78 a few times. Activity and GPU clocks were exactly the same as previously. But then the problem was back again. So whatever the problem is, it's not always there.

[madshi]

Ok, here it is.
Argh. MPC-HC shows a little message "Pause" when you pause playback. MPC-HC uses madVR's own OSD functionality for that. However, every time that madVR tries to update the OSD, the subtitles disappear. So the "Pause" message shown by MPC-HC practically removes the subtitles.

So why does it occur only on your PC, but not on anyone else's? I've added extra code to madVR to block the "Pause" OSD message. However, you seem to be using a translated version of MPC-HC. Your MPC-HC shows "Pausa" instead of "Pause". And madVR doesn't block "Pausa". You can fix the problem by translating just the "Pausa" text back to "Pause". I'm not sure where and how to do translations, though. The next madVR build will also block "Pausa". I'm aware that this all isn't really a good solution, but that's all I can do right now.

Here are the results of all the combinations I could think of
http://imageshack.us/g/31/lavvideocuvidadaptivehq.png/
or you can download all of them as one archive directly from here
http://www.mediafire.com/?au9ouh2w8b65z7x
Thanks. These are very interesting results - well done! From what I can see:

(1) EVR progressive ~ 15%
(2) madVR progressive ~ 25%
(3) EVR deinterlacing ~ 45%
(4) CUVID deinterlacing + EVR ~ 65%
(5) CUVID deinterlacing + madVR ~ 75%
(6) madVR deinterlacing ~ 80%

If I try to interpret these numbers, I come up with the following estimates:

- EVR progressive double rate ~ 30%
- madVR progressive double rate ~ 50%
- DXVA deinterlacing ~ 25%

The surprising thing here is the low GPU consumption with EVR deinterlacing. My best guess is that EVR saves GPU power by combining all deinterlacing and chroma upsampling and color conversion and scaling into one step. I've looked into what EVR does exactly and it uses basically the same DXVA2 API calls madVR uses, just with 8bit RGB output instead of NV12 output. CUVID and madVR deinterlace first (to NV12). Chroma upsampling, color conversion and scaling are done separately. Unfortunately there's not much we can do about it, unless we want to let DXVA2 do chroma upsampling and color conversion in addition to deinterlacing.

From what I saw I still think that madVR's Deinterlacing is too GPU intensive - a jump from from 24% to 79% GPU Load seams a lot.
Deinterlacing produces double rate, so you need to double the progressive 24% to 48% for double rate handling. So DXVA deinterlacing itself "only" consumes about 25-30% GPU.

One more thing I noticed is that LAV Video (CUVID - Adaptive + HQ Proc) + madVR (Deint=OFF) uses 4-5% less GPU than LAV Video (Soft) + madVR (Deint=ON).
Does this mean that there is some room for improvement?
Probably not. madVR has to do a couple of funny conversions to make the DXVA2 output compatible to pixel shaders. That's probably the 4-5% difference you're seeing. Well, maybe I could shave off a tiny bit by using CUDA to avoid those funny conversions with NVidia cards, but I'm not sure if that's really worth it.

Sorry I wasn't clearer, but what I was trying to say was that I'm seeing essentially the same problem under essentially the same conditions that cremor was reporting. That is, if I wrote a description of my problem, I would be writing something very similar to what cremor wrote.

For me, windowed mode was fine (no dropped frames) prior to 0.78 with one exception: If I use LAV CUVID (either stand-alone or combined), I get the same thing with older madVR versions. So the key point seems to be that if hardware de-interlace is used either via CUVID or DXVA 2 on either AMD or NVidia I get lots of dropped frames in windowed mode only. This leads me to believe that madVR is affected somehow by the hardware de-interlace.
Your report is *totally* different from cremor's report... ;)

You're reporting that hardware deinterlacing makes problems with windowed mode. That is not at all what cremor has reported. He reported that even *without* deinterlacing v0.78 works noticeably worse than v0.77. So basically he's reporting a new problem in v0.78 which didn't exist in v0.77. While you've reported that all madVR versions have problems in windowed mode when hardware deinterlacing is used (either via madVR or CUVID).

Your GPU is probably not powerful enough to run hardware deinterlacing in windowed mode, together with madVR rendering. I'm not sure if there's anything I can do about it. You probably already tried toggling the "perform deinterlacing in a separate thread" switch on/off?

http://www.mediafire.com/?dgd5p0mt185rh59

I checked each frame. EVR looks like it's 50p and madVR looks like it's 25p x2. Each frame happens twice with madVR. EVR really does produce fluid unique frames!
Ah, thanks.

I've now hooked into EVR and logged all the DXVA2 deinterlacing calls it makes. And I found a couple of small differences, one of which probably explains the problem you're seeing. Could you please test this special version to check whether it fixes the problem for you?

http://madshi.net/mrduck.rar

I know, but I am not the only one who gets this annoying bug, though since you're testing madVR in XP, then I guess, there is no reason to reproduce the same filter set, because I am using Window 7 x64.
Ok, so I'll put on my to do list to test this in win7 x64. Might take a while until I get to that, though.

Re: the upload/render queue problems with Windowed Mode....I've found a couple of interesting things:

1. It doesn't happen on my Llano machine, Windowed and FSE mode are both maxing all queues with deint=ON. That machine is, however, using an older driver version, from July.
2. Even with that driver, my HD6570 playback machine still exhibits the starved upload/render queues in windowed mode. Again, FSE mode, the queues are maxed.
3. On the 6570 machine, if I disable the ESVP option in CCC, madVR shows a Deinterlace Failure in red warning. ESVP seems to matter not on the Llano machine.
What is "ESVP"?

Did you try toggling the "perform deinterlacing in separate thread" option on/off? Doesn't help? Is your Llano machine's GPU more or less powerful than the 6570? The 6570 is probably not very powerful, so that might explain the problems in windowed mode.

Hope this helps:
http://www.mediafire.com/?7ek3ubk645kvv4f

edit:
Just managed to play the Devil May Cry clip with full queues in windowed mode with 0.78 a few times. Activity and GPU clocks were exactly the same as previously. But then the problem was back again. So whatever the problem is, it's not always there.
The log doesn't show much difference, except that somehow the render queue gets empty over time with 0.78, while 0.77 doesn't dip so low. Can't really see what causes the difference.

I've created a couple of special madVR versions for you to try:

http://madshi.net/cremor.rar

Please check if any one of them fixes the issue or not. Thx.

[madshi]

From what I saw I still think that madVR's Deinterlacing is too GPU intensive - a jump from from 24% to 79% GPU Load seams a lot.
P.S: Can you please double check whether you get the same results with this test build?

http://madshi.net/mrduck.rar

I don't expect any changes, but who knows. The ATI smoothness fixes might also help NVidia GPU Load. I kinda doubt it, though.

[mr.duck]

Ah, thanks.

I've now hooked into EVR and logged all the DXVA2 deinterlacing calls it makes. And I found a couple of small differences, one of which probably explains the problem you're seeing. Could you please test this special version to check whether it fixes the problem for you?

http://madshi.net/mrduck.rar

It appears to be exactly the same :(


(ESVP = enforce smooth video playback)

[derp]

Argh. MPC-HC shows a little message "Pause" when you pause playback. MPC-HC uses madVR's own OSD functionality for that. However, every time that madVR tries to update the OSD, the subtitles disappear. So the "Pause" message shown by MPC-HC practically removes the subtitles.

So why does it occur only on your PC, but not on anyone else's? I've added extra code to madVR to block the "Pause" OSD message. However, you seem to be using a translated version of MPC-HC. Your MPC-HC shows "Pausa" instead of "Pause". And madVR doesn't block "Pausa". You can fix the problem by translating just the "Pausa" text back to "Pause". I'm not sure where and how to do translations, though. The next madVR build will also block "Pausa". I'm aware that this all isn't really a good solution, but that's all I can do right now.

So that's what it was. I simply switched to the english version and now it works perfectly. Thanks.

You know what? You could also block the "pause" translations into the major european languages in order to prevent that this kind of stuff could happen to other people.
You can exclude Italy since "pausa" is in italian and you already said that you'll fix it in the next release.

[cremor]

I've created a couple of special madVR versions for you to try:

http://madshi.net/cremor.rar

Please check if any one of them fixes the issue or not. Thx.

The "queues" version seems to fix it! All queues are equal to 0.77 and no dropped or delayed frames.
The other three versions are same as 0.78.

I noticed that this version is back to a decoder queue of 8. Interesting how that can affect the following queues when the decoder queue with 12 was always full too. Or did you change more than just the queue size in that special version?

[Thunderbolt8]

I guess why deinterlacing doenst work that well on my PC is because of that frame rate doubling. I'd say only deinterlacing itself would be ok for my card, but the additional load due to frame rate doubling is too much to handle smoothly.

[noee]

What is "ESVP"?

Did you try toggling the "perform deinterlacing in separate thread" option on/off? Doesn't help? Is your Llano machine's GPU more or less powerful than the 6570? The 6570 is probably not very powerful, so that might explain the problems in windowed mode.

"Ensure Smooth Video Playback" in CCC/Video

Yeah, I tried the thread option, no difference. And the Llano machine I have is a laptop, with a less powerful GPU than the 6570. I think my issue is similar to cremor's, I will try that build.

[madshi]

It appears to be exactly the same :(
That's really bad. madVR's DXVA2 calls are now pretty much identical to EVR's. Well, I guess I'll have to install win7 x64 and the latest ATI driver and then hope to be able to reproduce the problem on my own PC.

The "queues" version seems to fix it! All queues are equal to 0.77 and no dropped or delayed frames.
The other three versions are same as 0.78.

I noticed that this version is back to a decoder queue of 8. Interesting how that can affect the following queues when the decoder queue with 12 was always full too. Or did you change more than just the queue size in that special version?
I only changed the queue sizes, nothing else. That's really bad that the bigger decoding queue makes problems. Other users reported improvements, so I'm reluctant to change this back. What happens if you slightly dial down the scaling algorithms with v0.78? Does that help, too? My impression from your logs is that the difference between 0.77 and 0.78 on your PC isn't that big. It just adds up over time, so that 0.77 is *just* able to keep up with rendering while 0.78 is only very slightly slower.

I guess why deinterlacing doenst work that well on my PC is because of that frame rate doubling. I'd say only deinterlacing itself would be ok for my card, but the additional load due to frame rate doubling is too much to handle smoothly.
Yes, that's quite possible. Though, the 4770 should be pretty powerful. I'm surprised it has problems with deinterlacing when using madVR. You're sure you've disabled anti-aliasing, anisotropic filtering etc? Does your GPU run near 100% GPU usage?

"Ensure Smooth Video Playback" in CCC/Video

Yeah, I tried the thread option, no difference. And the Llano machine I have is a laptop, with a less powerful GPU than the 6570. I think my issue is similar to cremor's, I will try that build.
I really wonder why so many people think their issue is similar to cremor's. You're at least the 3rd person to say so, and until now nobody really shared cremor's problems. Your issue is with deinterlacing activate. cremor's issue is with progressive sources (deinterlacing off).

That said, it wouldn't hurt to try those test versions.

[cyberbeing]

cremor, have you tried playing with the Windowed Mode Tweaks?

If you haven't tried setting everything to No Flush, that would be the first thing I'd try, then proceed to test each setting by itself one by one if all No Flush didn't help. I'd also try disabling Aero if Windowed Mode Tweaks don't help.

If your ATI 6870 w/ 60fps|1920x1080@120hz is only very slightly slower with 0.78 as madshi expects, little things like above may make a difference.

[cremor]

I only changed the queue sizes, nothing else. That's really bad that the bigger decoding queue makes problems. Other users reported improvements, so I'm reluctant to change this back. What happens if you slightly dial down the scaling algorithms with v0.78? Does that help, too? My impression from your logs is that the difference between 0.77 and 0.78 on your PC isn't that big. It just adds up over time, so that 0.77 is *just* able to keep up with rendering while 0.78 is only very slightly slower.

Bilinear seems to work, queues equal to 0.77 then. But as soon as I try something with custom pixel shaders the problem is there again. (All three scaling options changed at the same time.)

If the decoder queue is really the only problem here, how about adding an option to change it (or just registry key/file in madVR folder)?

But I still don't get how the decoder queue can affect anything on the GPU. Isn't this all on CPU and system RAM (with LAV software decoding)?

cremor, have you tried playing with the Windowed Mode Tweaks?

If you haven't tried setting everything to No Flush, that would be the first thing I'd try, then proceed to test each setting by itself one by one if all No Flush didn't help. I'd also try disabling Aero if Windowed Mode Tweaks don't help.

If your ATI 6870 w/ 60fps|1920x1080@120hz is only very slightly slower with 0.78 as madshi expects, little things like above may make a difference.

Will try that tomorrow.