madVR - high quality video renderer (GPU assisted)

[dansrfe]

Where did you get .3204 drivers? I'm using an Intel i3 NUC and just installed 9.18.10.3165 and the seekbar is not disappearing for me and can't find the drivers you listed.

What port does madVR listen on over LAN? Also, is there some documentation on how to use it?

[kasper93]

@dansrfe: See net-protocol.txt :readguid:

[raul31]

raul31,
rename "mpc-hc.exe" to something else (e.g. "game.exe"). And you have to force mpc-hc to use dGPU, not Madhcctrl.


See vivan's reply.


thanks Vivian & mashi! (my apologies as well for late reply) who would have thought that itz that easy... cool!
when i play a film/movie, MPC-HC is now displayed in the nvidia GPU activity in the notification area, which can only mean that madVR is utilizing it. :)

i have a lot more questions for u madshi (& other nvidia videophiles) as i'm trying to get the correct set-up with every detail. i hope u can guide me in getting the correct settings with everything.

now that i have "high performance nvidia processor" selected as the preferred graphics processor for MPC-HC LITE (i'm now using the LITE version... just trying it out) ,

what exactly do i choose for the next sequential section "specify the settings for this program:" ?

these are the features that are listed:

ambient occlusion
anisotropic filtering
antialiasing - FXAA
antialiasing - gamma correction
antialiasing - mode
antialiasing - setting
antialiasing - transparency
CUDA - GPUs
maximum pre-rendered frames
multi-display/mixed-GPU acceleration
power management mode
texture filtering - anisotropic sample optimization
texture filtering - negative LOD bias
texture filtering - quality
texture filtering - trilinear optimization
threaded optimization
triple buffering
vertical sync

now what setting should be applied to each one so that madVR gains the best quality from the nvidia GPU?

[madshi (& other nvidia videophiles), i don't expect u to retype all the features when answering me, just copy & paste all the features that i typed out & then specify afterwards what setting is the best for each one]

thanks a million !!! i really appreciate it! :)

[Andy o]

madshi, have you read up on the new Haswell GT3 (Iris 5100) and GT3e (Iris Pro 5200) chips and have an idea to how they would fare with madVR's more GPU-intensive tasks? The difference between those two is that GT3e has 128MB on-die RAM. Anandtech called it (http://www.anandtech.com/show/6993/intel-iris-pro-5200-graphics-review-core-i74950hq-tested) effectively an L4 cache. I'm feeling this year is finally new laptop year, and I'm eyeing out the Asus Zenbook Infinity which will apparently come with a dual-core GT3 part (i7-4558U (http://ark.intel.com/products/75992/Intel-Core-i7-4558U-Processor-4M-Cache-up-to-3_30-GHz)), and I'm wondering how much difference the lack of this on-die memory will make on madVR specifically, since I wouldn't buy a PC that doesn't run madVR a bit more than reasonably well! Also, this ultrabook has a 2560x1440 screen :)

[bugmen0t]

I will try watching movies with madVR OSD on but I can already assure you that there is no dropped frame or presentation glitch when there is a black frame since I checked that several times after it occured (since I thought in the past that it's one of the former). So there only remains the possiblity with the empty queues.

[ryrynz]

Where did you get .3204 drivers?

http://asder00.blogspot.se/

Actually, I take it back, it's still doing it on certain content.

[sucht]

I'm fairly sure he did mean XySubFilter, as that it the version that sends the subpictures to madVR directly instead of blending them onto the video.
Unfortunately, as far as I'm aware, XySubFilter is still work in progress and a public beta for testing purposes is not yet released, so the only way to get it is to compile it yourself from the source.


sorry but where can i get the XySubFilter sources? (i'd like to play around a bit)
all i can find is xy-VSFilter :/

[kasper93]

@sucht: See this post http://forum.doom9.org/showpost.php?p=1631842&postcount=19039

[pie1394]

That doesn't look so very much different from the defaults? FWIW, I would recommend to use a CPU queue slightly higher than the GPU queue. E.g. try a CPU queue of 24 with everything else the same.


The strange random frame dropping on HD7970 is somewhat annoying. Before using that setting, I did try several combinations and couldn't prevent the frame dropping when the GPU frequency is automatically changed from 300MHz to 950 MHz during the playback. Increasing the GPU queue size to 16 somehow does help on this particular system...

This issue is never noticed on GTX260+ and ION chipset. I feel that AMD driver's GPU frequency changing logic does not seem to run as smooth as nVidia driver's...

With the latest Catalyst 13.6 beta2 driver, it seems to work better than older drivers on some very high bit-rate samples which could stress the GPU at particular scenes.

[Vyral]

The 6850 is quite fast, but even the 6850 should not be able to do Jinc3 AR chroma upsampling with 0% GPU. So I would say there's probably something wrong somewhere. Does the madVR OSD report "NV12" (YCbCr 4:2:0)?

Here, a pic with madVR OSD and GPU-Z :
http://img199.imageshack.us/img199/7749/ca0.png
Do we have to uninstall and reinstall madVR for each new version or just uncompressed the zip file on the folder where we install madVR ?
I always use the second method and maybe something happened between two versions.

[madshi]

these are the features that are listed:

ambient occlusion
anisotropic filtering
antialiasing - FXAA
antialiasing - gamma correction
antialiasing - mode
antialiasing - setting
antialiasing - transparency
CUDA - GPUs
maximum pre-rendered frames
multi-display/mixed-GPU acceleration
power management mode
texture filtering - anisotropic sample optimization
texture filtering - negative LOD bias
texture filtering - quality
texture filtering - trilinear optimization
threaded optimization
triple buffering
vertical sync

now what setting should be applied to each one so that madVR gains the best quality from the nvidia GPU?
Leave all to default. Things like anisotropic filtering and anti-aliasing should be set to "let application decide" or "application preference" or however that option is called.

madshi, have you read up on the new Haswell GT3 (Iris 5100) and GT3e (Iris Pro 5200) chips and have an idea to how they would fare with madVR's more GPU-intensive tasks?
Read up: yes. Idea how they would fare: no. Guess we need to wait for tests.

[madshi]

Here, a pic with madVR OSD and GPU-Z :
http://img199.imageshack.us/img199/7749/ca0.png
Do we have to uninstall and reinstall madVR for each new version or just uncompressed the zip file on the folder where we install madVR ?
I always use the second method and maybe something happened between two versions.
Uninstall/reinstall is usually not needed. I don't know why GPU-Z shows a 0% GPU usage. Looks like a bug in either GPU-Z or in the NVidia driver to me.

[madshi]

madVR v0.86.5 released

http://madshi.net/madVR.zip

* fixed: calibration test pattern colors were sometimes incorrect
* added support for internal VideoLUT processing
* added remote control ability for internal VideoLUT processing
Nothing important for "normal users". Just a fix and some improvements for developers of calibration software...

[QBhd]

Uninstall/reinstall is usually not needed. I don't know why GPU-Z shows a 0% GPU usage. Looks like a bug in either GPU-Z or in the NVidia driver to me.

It is the same here with a 5770... it always been that way, it will show usuage when interlaced content is played, but not 720p

I have always just chalked it up to a little quirk and nothing to worry about. I know the GPU is being used since memory usage increases as well as temps

QB

[karamancho]

raul31,And you have to force mpc-hc to use dGPU, not Madhcctrl.

how do you do that?

[dansrfe]

Backbuffer queue keeps cycling between 0/8 and 0-7/8 in windowed mode with 0.86.4/5. I've tried pausing/playing, going back and forth in the timeline, and resizing above and below 1080p (original video resolution).

[leeperry]

Having more thoughts about 3DLUT gamut mapping, would that be possible to have the "primaries" tag also activated via a dummy PS script by any chance please? Something like "// primaries=709|SmpteC|EBU|sRGB|NTSC|PAL|470M|240M|170M"?

It seems rather clear that the PS script gamut mapping days are counted and it would still be highly appreciated to have the ability to override mVR's automatic logic, because IME +75% of 24p HD movies are still SMPTE-C(they look terribly greenish in REC.709) and I like anything 25fps to be treated as EBU, even HD(following the ISF recommendations).

I know it can be rolled with a hotkey but I only have two unused buttons on my mouse, one is currently used for matrix rolling and the other for TV/PC levels.

It would also be useful for ppl with funky encodes or gamuts that aren't properly detected automatically, one automatic profile in their media player et voilà. I realize that feature requests are a no-no but this would finally nail down the whole 3DLUT management matter, hope you'll consider it :)

The key reason for this release is the added support for remote controlled calibration test patterns, in the hope that ArgyllCMS + HCFR will make use of that.
Oh wow, another dream coming true: the ability to run a test proof verification pass without the major hassle of rolling manual test patterns. If HCFR supported this, we could end up with a lot of dead-flat curves to drool at :cool:

Jinc3 (without AR) needs to read 36 input pixels to produce 1 output pixel. For each of those 36 input pixels Jinc also needs to access a Jinc 1dlut. Add to that a lot of shader math.
Ahhh, no wonder things quickly get out of hand once you reach >540p source files. That would also explain why J3AR looks so impressive, OTOH sometimes I prefer bicubic on noisy/low-res files in order to act as a poor man's noise reduction.

BTW, is there any picture viewer that would support Jinc3? I guess AR won't be possible as you seem to have come up with this feature on your own, but ACDSee's resizing looks really ugly compared to mVR :o

[zoyd]

Test version of HCFR with auto-generated madVR patterns is available here (http://www.avsforum.com/t/1393853/fork-of-hcfr-started-whats-needed/2070#post_23434733).

[andybkma]

Anything us Zoom Player users can do to help with Bug 30? My ZP constantly gets stuck on top as per the bug

[pie1394]

Read up: yes. Idea how they would fare: no. Guess we need to wait for tests.

Does the madVR code require a lot of "random" memory read access for each pixel group's processing? If not, I guess the performance is still stream-processor-count limited.

If I don't remember it wrong, nVidia's Kepler-based and AMD GCN-based (like Tahiti, Pitcairn, Cape Verde) 28nm GPUs have somewhat HUGE on-die SRAM cache than any previous generation in order to reduce the probability of external memory access. The on-die cache accessing bandwidth is at least TB scale, not just 100 ~ 200GB scale. That is why nVidia can cut GTX660 ext-mem to 192-bits from GTX560Ti's 256-bits, but still with better performance on even heavily access at big multi-texture memory fetching.

The GT3e 128MB wide-ported EDRAM definitely helps for the performance to reduce the accessing chance with main memory -- shared with CPU / peripherals. Its design is logically identical to a stand-alone GPU...

In Computex 2013, some vendors have demostrated the dual 4K output with Haswell. Although the performance is below the requirement of video playback, it still indicates Haswell capability -- up to 2 * 4K 30/60 Hz display via DP ?

[madshi]

Backbuffer queue keeps cycling between 0/8 and 0-7/8 in windowed mode with 0.86.4/5. I've tried pausing/playing, going back and forth in the timeline, and resizing above and below 1080p (original video resolution).
Is that a new problem with v0.86.4/5? Which is the last version this did *not* occur with, using the same settings?

Having more thoughts about 3DLUT gamut mapping, would that be possible to have the "primaries" tag also activated via a dummy PS script by any chance please?
Not sure what you mean.

Oh wow, another dream coming true: the ability to run a test proof verification pass without the major hassle of rolling manual test patterns. If HCFR supported this [...]
It already does, since today!

BTW, is there any picture viewer that would support Jinc3?
Viewer? I don't think so. The only other software using Jinc is ImageMagick, as far as I know. Well, ImageMagick actually had it first, thanks to Nicolas Robidoux.

Test version of HCFR with auto-generated madVR patterns is available here (http://www.avsforum.com/t/1393853/fork-of-hcfr-started-whats-needed/2070#post_23434733).
Very nice - thanks for your work! :)

Anything us Zoom Player users can do to help with Bug 30? My ZP constantly gets stuck on top as per the bug
Was waiting for feedback on that bug report, didn't notice the feedback was already there. The problem for me is that madVR seems to do everything correctly, so I'm not sure where the problem comes from. But I'll see if I can find a fix somehow...

Just for my interest: Why is this coming up now? Is it a new problem? Or did this problem always exist, in all older madVR and ZoomPlayer versions?

Does the madVR code require a lot of "random" memory read access for each pixel group's processing?
The question is how to define "random". Of course it's not random. But maybe it looks like that to the memory controller, what do I know?

[madshi]

madVR v0.86.6 released

http://madshi.net/madVR.zip

* fixed: test pattern API "BlindConnect" made problems with HCFR
* fixed: 3dlut was never applied when drawing test patterns
* fixed: HCFR running with admin rights couldn't always connect to madVR
* added new remote control API "madVR_SetBackground" for plasma users
Some more fixes, mostly aimed at making HCFR <-> madVR work well together. Which they now do, by using v0.86.6 and the latest HCFR test build available on AVSForum (http://www.avsforum.com/t/1393853/fork-of-hcfr-started-whats-needed/2070#post_23434733). Thanks, zoyd! :D

[TheShadowRunner]

Just for my interest: Why is this coming up now? Is it a new problem? Or did this problem always exist, in all older madVR and ZoomPlayer versions?

The issue has always been there for me, regardless of madVR or ZP versions.
(the issue: ZP is stuck in always on top mode when/after using madVR)
-When the issue is going on, enabling ZP's own "Stay on top" feature and disabling it right away solves the problem..
-It seems the issue (ZP becoming AOTop) occurs after madVR has gone in fullscreen exlusive mode at least once.
- Worth noting, in my case, the issue occurs WITHOUT ZP displaying OSD through madVr's OSD API.
(so in ZP; "Video covers entire display area" is checked, and "Display OSD though madVR's OSD API" is unchecked, and in madVR: "enable auto fullscreen exclusive mode" and "delay switch to exlusive mode by 3 seconds" are both checked.)
Hope it hints/helps.

[leeperry]

Test version of HCFR with auto-generated madVR patterns is available
It already does, since today!
Actually, v0.86.6 just broke HCFR. A new HCFR build will be needed. Sorry about that! After that all should be fine
OK, amazing! Will patiently wait until you guys will give the green light but this feature sounds very exciting :)

Not sure what you mean.
Well, PS scripts can have comments flagged by the "//" characters at the beginning of each line, for instance:
// DisplayLessThan16 v1.1=ps_2_0
// http://www.homecinema-fr.com/forum/viewtopic.php?t=29814317
Could you somehow recognize the "primaries=709|SmpteC|EBU|sRGB|NTSC|PAL|470M|240M|170M" tag there? or create a new internal command maybe? So we could either roll gamuts via a hotkey, a filename tag or a "dummy" PS script comment :)

It would be useful to the SMPTE-C tinfoil hat crowd(I'm not the only one ^^), those with broken encodes(29.97fps HD to SD encodes) or people working on non-standard gamuts. They would have the ability to create automatic profiles in their media player and roll gamuts just as easily as it currently is with the PS script from yesgrey & JohnAd.

I guess the code is more or less already there? You would only need to extend the rollability to PS scripts and not just filename tags.

Hope you'll consider it as I'd really crave this feature, :thanks:

Viewer? I don't think so. The only other software using Jinc is ImageMagick, as far as I know. Well, ImageMagick actually had it first, thanks to Nicolas Robidoux.
OK, thanks for the info I'll see if IM could work as a FS pictureviewer, otherwise let's hope that someday you'll find the time and motivation to make a resizing plugin for a picture viewer then :)

[madshi]

The issue has always been there for me, regardless of madVR or ZP versions.
Ok.

OK, amazing! Will patiently wait until you guys will give the green light but this feature sounds very exciting :)
"Green light".

Well, PS scripts can have comments flagged by the "//" characters at the beginning of each line
But what does gamut switching have to do with custom pixel shaders? I don't really see the connection.

[leeperry]

"Green light".
OK, sweeet!

But what does gamut switching have to do with custom pixel shaders? I don't really see the connection.
I currently have a bunch of automatic profiles based on frame rate and resolution in PotP: http://thumbnails108.imagebam.com/26085/90bea0260842791.jpg (http://www.imagebam.com/image/90bea0260842791)

Each profile uses a different 3DLUT PS script in order to map EBU/SMPTE-C/REC.709 to my native display gamut based on my own settings, but this PS script is utterly limited compared to what ArgyllCMS can now provide through a full-fledged 3DLUT.

It's possible to roll gamuts in mVR via the "primaries" filename tag or a hotkey, but IME it's a lot more convenient to create automatic profiles in your media player so you won't have to tag every damn file or keep pressing the same hotkey several time in a row for each movie....hence my request to not only recognize the "primaries" tag on filenames but on also on PS script comments, so I can keep on using automatic profiles for gamut mapping but with a mVR/Argyll 3DLUT and not a sloppy PS script anymore :o

Also, many HTPC users don't have a keyboard within hands reach so rolling gamuts via a mouse click in your media player is just much more convenient.

To make it perfectly clear, my SMPTE-C profile would not be using this 3DLUT script anymore:
// script=ps_3_0
// Code from MPC

sampler s0 : register(s0);
float4 p0 : register(c0);

static float4x4 r2r =
{
1.0298551049805,-0.0398303799731499,-3.59414612715048E-05,0,
-0.0245864381394802,1.06321188848582,-0.0344900188344141,0,
-0.0107378988239096,0.00703650256300146,0.9977150476297,0,
0, 0, 0, 0
};

float4 main(float2 tex : TEXCOORD0) : COLOR
{
float4 c0 = tex2D(s0, tex);
c0 = pow(c0, 1/0.45);
c0 = mul(r2r, c0);
c0 = saturate(c0);
c0 = pow(c0, 0.45);

return c0;
}

but this:
// primaries=SmpteC

[fairchild]

Madshi, thanks for implementing this with HCFR. I tried it out, got the latest file from that link from Zoyd, and the latest MadVR you posted but I am not sure exactly how I go about using these patterns. I've only ever used HCFR with external patterns using the DVD manual option when asked to choose a generator. I guess now I'd select view images then make sure MadVR is selected when I go to configure generator. But then I'm getting this error when I try to measure grayscale for instance:

dll found but Blind Connect failed, is MadVR running?

followed by this error:

Error during generator initialization

[madshi]

It's possible to roll gamuts in mVR via the "primaries" filename tag or a hotkey, but IME it's a lot more convenient to create automatic profiles in your media player so you won't have to tag every damn file or keep pressing the same hotkey several time in a row for each movie....hence my request to not only recognize the "primaries" tag on filenames but on also on PS script comments, so I can keep on using automatic profiles for gamut mapping but with a mVR/Argyll 3DLUT and not a sloppy PS script anymore
Well, I'm evaluating tags only when playback starts. Evaluating tags inside of shader script code is a different thing. Some tags I could put into effect in the middle of playback. But that might not work for all tags. I'd have to check for every tag what consequences changing it in the middle of playback would have. So I'll have to pass on this for now, because it's not just a minor modifications, but a "new feature".

I'm getting this error when I try to measure grayscale for instance:

dll found but Blind Connect failed, is MadVR running?
You need to start a media player and playback some video (doesn't matter which). Basically some madVR rendering needs to be happening somewhere on the PC where HCFR is running, or on any other PC in your LAN. HCFR will then take over, the video playback will be paused, and test patterns will be rendered instead.

[leeperry]

Well, I'm evaluating tags only when playback starts. Evaluating tags inside of shader script code is a different thing. Some tags I could put into effect in the middle of playback. But that might not work for all tags. I'd have to check for every tag what consequences changing it in the middle of playback would have. So I'll have to pass on this for now, because it's not just a minor modifications, but a "new feature".
Ok thanks for the fast reply, fair enough.....but it would also be useful in MPC for instance, for when you only have a wireless mouse on the couch and no keyboard around: http://thumbnails103.imagebam.com/26085/127c18260845508.jpg (http://www.imagebam.com/image/127c18260845508)

Changing primaries on-the-fly should be OK? But alright, I'll wait for mVR 1.0 and will ask again then :o

[Soukyuu]

Is the "black frame problem" referring to the screen going blank for a split second now and then? I am still at 0.86.3 and suddenly started noticing it. I don't think it happened with 0.86.2 and I haven't updated any drivers for ages. win7 x64 SP1, nVidia 306.97 (260GTX), what should I do to give you some viable info?

[bugmen0t]

@Soukyuu
Please read here (http://bugs.madshi.net/view.php?id=84).

@madshi
I was able to see several black frames with OSD on however there wasn't anything suspicious. Every queue looks fine before and after it happens, no changes at all.

[madshi]

Please post this in the bug tracker, not here.

[sucht]

@sucht: See this post http://forum.doom9.org/showpost.php?p=1631842&postcount=19039

thank you :)

[kasper93]

@madshi:

Is it by design that "disable GPU gamma ramps" change only after reseting playback? Every other options works instantly and this can be confusing. And what do you think about making it work also in windowed mode?

[Plutotype]

Hi folks,
did som tests with the new Intel HD4600. Unfortunately for some strange reason I couldnt get GPU-Z to be installed ( initial splash logo seen, but the app froze there, strange ). Fortunately the rendering times from the madVR OSD menu are very informative.

The demo unit was a NB with Intel HD4600 ( and AMD8790m ), Intel Core i7-4800MQ. Installed the Intel Haswell drivers 3186 ( couldnt install the 10th June package ).
The GOAL was to test madVR performance with Intel HD4600 by comparing its rendering performance using some samples.
Settings:
Full Screen Exclusive mode, 1080p Display.
Chroma - Bicubic 75AR
Luma up - JINC 3taps AR
Luma Down - Catmull AR, LL


720p 59.94fps, Bitrate 4,5 Mbps
Refresh 59.94128 Hz
Many dropped frames. Changing LUMA upscaling from JINC3 AR to Lanczos 3taps AR lead to zero dropped frames and full queues.
Rendering average with JINC3 2.02ms, max 90.95ms. Rendering average with Lanczos3 was 2.02ms, max 18.95ms.

1080i 25fps, Full BD, Bitrate over 20 Mbps, YADIF @1080p/50Hz ( LAVvideo outputs 50p )
Refresh 50.0010 Hz
No dropped frames, full queues. Rendering average 1.47ms, max 11.56ms.

1080p 23.976fps Full BD, Bitrate 24.7 Mbps
Refresh 23.97651 Hz
No dropped frames. Rendering average 1.57ms, max 12.56ms

720p 23.976fps, Bitrate 3.17 Mbps
Refresh 23.97651 Hz
No dropped frames. Rendering average 2.75ms, max 27.14ms

Here are some results with smooth motion ON @ 60Hz:

1080i 25fps, Full BD, Bitrate over 20 Mbps, YADIF @1080p/50Hz ( LAVvideo outputs 50p )
Refresh 60.06105Hz
No frame drops. Average rendering 4.39ms, Max rendering 14.43ms.

720p 23.976fps, Bitrate 3.17 Mbps
Refresh 60.06105Hz
Quicksync ON: Average rendering 25.04ms, Max rendering 40.43ms
Quicksync OFF: Average rendering 19.09ms, Max rendering 36.69ms
No frame drops

1080p 23.976fps Full BD, Bitrate 24.7 Mbps
Refresh 60.06105Hz
Quicksync ON: Average rendering 4.45ms, Max rendering 20.15ms
Quicksync OFF: Average rendering 4.36ms, Max rendering 16.06ms
No frame drops

I didnt have more time to check other samples (SD to 1080p upscaling for example ), but the test gave me information, that although HD4600 shows so-so performance, Im looking forward to HD5000/HD5100/HD5200 iGPUs. It could be interesting to see JINC3 capable iGPU, which would upscale SD/HD to FullHD without dropping any frames. Another positive note is, that Intel managed to offer all necessary refresh rates for 1080p and they are very accurate.
Apparently Full RGB over HDMI is still a issue for Intel iGPU = it outputs limited RGB only. I had to switch from TV-levels to PC-levels in madVR for film content to get correct dynamic range on the TV screen. I remember reading something about banding in this scenario.
Other thing I have noticed is some strange flickering when playback is started (takes 1second max). It looked like the vertical refresh takes too much to settle with the fps. Maybe Im wrong.

[mindbomb]

The deinterlacing toggle for film and video, it seems to not work sometimes, switching between video and autodetect video and not going through film mode.

edit:Oh, nvm, I was using dxva native, which doesn't work with film mode.

[zoyd]

Installable HCFR 3.0.4.3 package with the latest madVR support can be found here (https://sourceforge.net/projects/zoyd000.u/files/downloads/). Old version needs to be uninstalled first.


edit:

I wonder if anyone could help me with a video card quirk.

Symptoms:

1. Whenever there is an HDMI handshake the video card resets itself into a mode where it loads what I think are default calibration values. This will happen whenever I switch my TV to the PC input or when madVR starts up.

2. To clear the video card tables I run the ArgyllCMS tool dispwin -c (load linear tables) and this fixes the problem such that a grayscale gamma run using HCFR reproduces what I see with other pattern generators.

3. However, when I fire up madVR the same reset happens but dispwin -c no longer clears it so I assuming either madVR blocks modification of the video card tables or it does some other manipulation of it's attributes such that linear tables do not fix the problem. (checking and unchecking the loading of gamma ramps button has no effect).

4. The only way to fix 3. is to actually load a previous calibration file while madVR is running using dispwin -c calfile.cal, and then dispwin -c

I can work around it but it's very annoying, can a true "clear video calibration tables" be implemented in madVR?

[kasper93]

(checking and unchecking the loading of gamma ramps button has no effect).

This works only in full screen exclusive mode. And make sure that you reset playback after changing this switch because it will not work until you reset madvr. Hope this helps a little.

[zoyd]

This works only in full screen exclusive mode. And make sure that you reset playback after changing this switch because it will not work until you reset madvr. Hope this helps a little.

Thanks for the tip but that actually wasn't the problem. Everything works well as long as madVR does not switch refresh rates. The problem was in entering 1080p59 as an allowed mode and then it would switch to that mode on start-up and that hosed the card. If I force 1080p60 there is no reset.

[pie1394]

The question is how to define "random". Of course it's not random. But maybe it looks like that to the memory controller, what do I know?

Well... I should said if it is random external DDR3/DDR5 memory access on different memory rows / banks. The GPU's texture memory just works like CPU cache and often fetched / flushed in cacheline unit. The bottleneck of shader's register data fetching from GPU texture cache depends on the SRAM bank accessing race condition... :p

Based on Plutotype's test on Intel HD4600, it does not surprise me. the GT2/GT3's performance is still far from the requirement of Jinc3_AR scaling at 24~30fps. It looks like just around 10~12 fps for 1280x720 to 1920x1080 upscaling -- the Chroma part's scaling algorithm is not even Jinc3_AR ... I guess it will not be faster than 8 fps if both are scaled by Jinc3_AR. :D

[madshi]

Is it by design that "disable GPU gamma ramps" change only after reseting playback? Every other options works instantly and this can be confusing. And what do you think about making it work also in windowed mode?
Not every other option works instantly, if you change the list of display modes, this might also only affect playback the next time you start it. So yes, the gamma ramps option is supposed to only be executed once when playback starts. Maybe it would be nicer to have it take effect right away when changing it, but the current behaviour is not a bug.

Making it work in windowed mode is on my list of things to maybe look at.

Apparently Full RGB over HDMI is still a issue for Intel iGPU = it outputs limited RGB only.
That is pretty bad news! Is there no levels switch at all in the drivers? Or is it there, but doesn't work?

Anybody else with Haswell can confirm that?

1. Whenever there is an HDMI handshake the video card resets itself into a mode where it loads what I think are default calibration values. This will happen whenever I switch my TV to the PC input or when madVR starts up.
Interesting. Do you have an ICC profile assigned to your display? If so, removing that might fix the issue.

3. However, when I fire up madVR the same reset happens but dispwin -c no longer clears it so I assuming either madVR blocks modification of the video card tables or it does some other manipulation of it's attributes such that linear tables do not fix the problem.
Strange. Is madVR is FSE mode at that time? Does this problem occur regardless of whether you told madVR to disable the gamma ramps or not?

I can work around it but it's very annoying, can a true "clear video calibration tables" be implemented in madVR?
It's on my list of things to look at.

[huhn]

That is pretty bad news! Is there no levels switch at all in the drivers? Or is it there, but doesn't work?

Anybody else with Haswell can confirm that?

i don'tthink thi is an hawell problem i a general problem with intel driver.

the driver have a switch for YCbCr on or off that all

[nevcairiel]

That is pretty bad news! Is there no levels switch at all in the drivers? Or is it there, but doesn't work?

Anybody else with Haswell can confirm that?

It works like it always worked, it outputs over HDMI whatever the HDMI device requests. Sadly most TVs only request/offer limited RGB in their HDMI handshake, even if they allow and are configured for full RGB.
It can output full RGB just fine, for example when i connect my PC screen over HDMI, it just has no toggle to override the HDMI handshake (or well, it has one, but it doesn't work)

[huhn]

(or well, it has one, but it doesn't work)

you mean the IT content mode?

It works like it always worked, it outputs over HDMI whatever the HDMI device requests.

would be nice if that's work fine http://s3.imgimg.de/uploads/television05d4a8bapng.png

the first iiyama "television" it output limited to an pc monitor

[nevcairiel]

Sounds like your PC monitor has a wrong HDMI EDID :)
Works on my DELL screens, HDMI has full range RGB. Although i switched to DisplayPort on Haswell, but it did work on SNB ;)

[huhn]

it's only with intel this way works fine with ion and 6770 by default but i leave it at that it's not a problem in madvr. i have not tryed my new display but...

[zoyd]

Interesting. Do you have an ICC profile assigned to your display? If so, removing that might fix the issue.


no profile is loaded. It's the native behavior of the video card.


Strange. Is madVR is FSE mode at that time? Does this problem occur regardless of whether you told madVR to disable the gamma ramps or not?


yes and yes. Basically whenever anything within madVR causes an HDMI handshake the card goes to it's default (bad) behavior and I can't load linear tables unless I exit madVR.


It's on my list of things to look at.

thanks!

[bugmen0t]

I tried increasing the backbuffers for the old path FSE but the OSD is still showing the default 3 being used.
Am I missing something or is this a bug?

@toniash: It's win8.

[toniash]

I tried increasing the backbuffers for the old path FSE but the OSD is still showing the default 3 being used.
Am I missing something or is this a bug?
I think that in XP the max number is 3

[shimaflarex]

I tried increasing the backbuffers for the old path FSE but the OSD is still showing the default 3 being used.
Am I missing something or is this a bug?

@toniash: It's win8.

I reported that months ago and madshi said more than 3 backbuffers on the old path couldn't be done. I dunno why he didn't reduce the limit on the listbox.