madVR - high quality video renderer (GPU assisted)

[Betroz]

Even if you have the performance available 256 neuron chroma upscaling is a waste of electricity/heat. I suggest you never use above 64 neurons for chroma upscaling, even that is overkill but less overkill than 256 neurons, which is just excessive.

Luma/Chroma doubling is harder than chroma upscaling because, assuming 1080p 4:2:0 video, chroma upscaling has to scale 960x540 to 1920x1080 while luma/chroma doubling has to scale 1920x1080 to 3840x2160. Also chroma doubling is twice as hard as luma doubling; chroma is two planes, U and V, while luma is only one, Y. Your results are not strange at all. :)

Use as high as you can for luma doubling but don't go above 32 neurons for chroma doubling and 64 neurons for chroma upscaling.

With a Titan X (Pascal) I have plenty of performance available but using more neurons for chroma really is pointless.

Thanks for your advice, I'll check it out :)

[Ver Greeneyes]

Strangely enough, then I could use NNEDI3 256 for Chroma Upscaling.

It's not that strange: the chroma resolution is usually a quarter of the luma resolution (half the height and half the width), but there are two chroma channels and only one luma channel - so NNEDI3 256 for chroma upscaling is comparable to NNEDI3 128 for luma doubling (as doubling the number of neurons roughly doubles the cost).

So while NNEDI3 256 for chroma upscaling still costs more than NNEDI3 64 for luma, going down to NNEDI3 64 for luma must have freed up enough resources to use NNEDI3 256 for chroma upscaling.

[Betroz]

I have a 1080 GTX as well (FTW) and Jinc with SR 4 for Chroma and Scaling with Linear Light and Anti-ringing checked looks much better than NNEDI3 256. The image looks real, not digitized. SuperXBR and NNEDI3 have a very digitized and processed look to them.

NNEDI3 is a good scaler for chroma, but it makes colors look overprocessed and in some cases they can cross.

Jinc is the cleanest scaler around and when you super-res it it only gets sharper. Other scalers like NNEDI3 and SuperXBR gets sharper too, but you start to see the artifacts easily.

From my testing Jinc AR + super-rez is not as sharp as NNEDI3 64 and higher, but I didn't try SR4, only SR2 in my testing.

[Betroz]

It's not that strange: the chroma resolution is usually a quarter of the luma resolution (half the height and half the width), so NNEDI3 256 for chroma upscaling is comparable to NNEDI3 64 for luma doubling (as doubling the number of neurons roughly doubles the cost).

So by going from NNEDI3 128 to NNEDI3 64 for luma, you freed up roughly enough resources to do NNEDI3 256 for chroma.

Yeah that makes sense. I use NNEDI3 64 for Chroma doubling too. But will try out the settings Asmodian adviced me to use.

[Ver Greeneyes]

Yeah that makes sense. I use NNEDI3 64 for Chroma doubling too. But will try out the settings Asmodian adviced me to use.

I forgot that chroma has 2 channels, updated my post above. Either way, differences in quality from chroma upscaling are often so negligible that it's worth more to put those resources into luma doubling :)

[Shiandow]

We've had discussions about the best way to handle these situations in this thread a while ago. I've also talked to Calman, Florian (displaycal), zoyd and Graeme about this.

The key is the "fix too bright & saturated pixels by" combobox in the madVR settings. Try setting that to "0% luminance reduction", and you'll get something nearer to what you expect. This may produce better looking results for this specific scene, but in other scenes it looks worse because the pixels actually become near white which means they're losing their color information.

I think I may have figured out why I didn't like that result much. It preserves the hue, which gives it a rather unnatural look. At least I've never seen a fire look anything like this (http://i.imgur.com/vdF12Db.jpg) before.

If instead you tried to pick the closest RGB colour, while preserving the luminance you'd get something like this (http://i.imgur.com/QcDrTTR.jpg), which already looks a bit more like an actual fire. Now if you were to pick the colour closest in Lab I think you'd get something like this (http://imgur.com/It9KJsk.jpg), but I'm not still not entirely sure if the method I used is 100% correct.

You're right that you'll lose some colour information this way, but you're going to lose some information no matter what you do, and I'm not convinced that hue and saturation are more important than luminance.

[madshi]

I think I may have figured out why I didn't like that result much. It preserves the hue, which gives it a rather unnatural look. At least I've never seen a fire look anything like this (http://i.imgur.com/vdF12Db.jpg) before.

If instead you tried to pick the closest RGB colour, while preserving the luminance you'd get something like this (http://i.imgur.com/QcDrTTR.jpg), which already looks a bit more like an actual fire. Now if you were to pick the colour closest in Lab I think you'd get something like this (http://imgur.com/It9KJsk.jpg), but I'm not still not entirely sure if the method I used is 100% correct.
FWIW, Lab does not really preserve hue angles well. Lab is better than gamma or linear light RGB, but still far from optimal for this kind of processing.

You're right that you'll lose some colour information this way, but you're going to lose some information no matter what you do, and I'm not convinced that hue and saturation are more important than luminance.
I've started out with the same thinking, and put luminance preservation at highest priority. But it results in serious problems in many situations. E.g. on a quick search I found this:

http://forum.doom9.org/showpost.php?p=1773692&postcount=38653

If you put priority on reproducing luminance correctly, the orange headlight of the truck becomes white. That looks really bad in that specific image.

Have you played with the "fix too bright & saturated pixels by" setting in madVR? It allows you to decide for yourself if you value luminance higher than hue/saturation.

[Shiandow]

FWIW, Lab does not really preserve hue angles well. Lab is better than gamma or linear light RGB, but still far from optimal for this kind of processing.

Well, if I did everything correctly I should be able to minimize the distance in any colour space. Unfortunately, for most spaces more advanced than Lab the difference in colour isn't measured as a simple distance, which complicates things slightly. Something like CIE94 might just be possible, although it's not symmetric, which is odd. Worth a try anyway.


I've started out with the same thinking, and put luminance preservation at highest priority. But it results in serious problems in many situations. E.g. on a quick search I found this:

http://forum.doom9.org/showpost.php?p=1773692&postcount=38653

If you put priority on reproducing luminance correctly, the orange headlight of the truck becomes white. That looks really bad in that specific image.

That headlight does seem to be over 2000 nits bright though, on a 400~600 nit screen 'white' is probably the closest you're going to get, in any colour space.

The real problem seems to be that simply picking the closest RGB colour will lead to some lost detail, regardless of how you define distances. Fixing that probably can't be done by looking at individual pixels.


Have you played with the "fix too bright & saturated pixels by" setting in madVR? It allows you to decide for yourself if you value luminance higher than hue/saturation.

Well, seems I value luminance higher than saturation, and saturation higher than hue. Which is unfortunate since the one thing your method does perfectly is preserving the hue, at the cost of luminance and saturation.

[madshi]

Well, seems I value luminance higher than saturation, and saturation higher than hue. Which is unfortunate since the one thing your method does perfectly is preserving the hue, at the cost of luminance and saturation.
Have you tried setting the "fix too bright & saturated pixels by" option to "0% luminance reduction"? That should do what you want?

[Shiandow]

Have you tried setting the "fix too bright & saturated pixels by" option to "0% luminance reduction"? That should do what you want?

That resulted in this (http://i.imgur.com/vdF12Db.jpg)image, which I think looks rather unnatural. Now maybe it's an accident that fire looks better if you don't preserve hue, but I don't really know how to test that (well an HDR screen might help, but if I had one I wouldn't be trying to figure out how to display HDR on an SDR screen).

[madshi]

One thing I found is that you can't just look at one image of one HDR demo. If you optimize for that, things will look bad in other scenes of other demos. Anyway, when dumbing down HDR to SDR, we're between a rock and a hard place. It's simply impossible to keep all the information. The only proper way to evaluate if any given HDR -> SDR algo works "correctly" would be to test with a true 4000 nits or 10000 nits display. But we don't have those yet. So it's really hard to properly test these algos.

[HillieSan]

I peacefully disagree with you.

Latest nvidia drivers are broken in every way possible, they crash the card and do more bad things...

When drivers work, madVR works. Haven't experienced problems with madVR when everything else worked fine, using windows 10 64bit, nvidia GPUs, laptop. Tried multiple configurations and all went well on most configurations.

Even on ati laptops it worked flawlessly.

Just saying, some companies break the drivers and let us walk in the darkness, had to reinstall windows + nvidia drivers 3 times to get those things working. not only madVR was broken, but everything else was...

Well, they seem stable now and everything runs well now on most configurations I've tried, so we can happily use madVR now :D

But I've had my share of mpc-hc has stopped working and mpc-hc critical errors, can understand where you're coming from...

I understand your point. Indeed broken drivers could be blamed, but sometimes people blame the drivers too easily. In this case, all renderers work well with mpc-hc and potplayer except for madvr. This points to madvr. BTW I have an AMD RX 480, not nVidia.

I have found the solution to my problem. The issue is with madvr all right. This is an issue with upscaling. With non-upscaling I experience no problems with madvr. madvr uses directx11 for upscaling and this should be enabled. This is disabled by default. Dunno why.

I've found the solution here and it works. I had to enable directx11 in madvr.

https://www.reddit.com/r/Amd/comments/52uxn6/psa_use_dx11_in_madvr_to_prevent_hangscrashes_on/

This works for the x64 versions of mpc-hc and potplayer.

Now BlueSkyFRC doesn't work anymore. SVP works fine instead.

[huhn]

have a look at this: http://bugs.madshi.net/view.php?id=431

you don't have to use DX11

[HillieSan]

have a look at this: http://bugs.madshi.net/view.php?id=431

you don't have to use DX11


This doesn’t help. I have no problems with the latest drivers and DX9 in windowed mode. My problem is with switching to EFS mode. I read that people with nVidia and Intel experience similar problems.

All other renderers I have used work fine in DX9 with DXVA2-cb. madVR is the unstable one. Looking at the different behaviors I read on forums, including my experience, I am thinking more about a race condition or a floating pointer than can cause undefined behavior , which may explain the different behaviors per driver/system. I am a software engineer and I have encountered these pathological problems a lot. My guess is that the madvr should be based on, for example, the haali implementation (they got is right). This is my feeling based on analysis, but I could be wrong. This is a nasty defect anyway.

BTW I experience issues in DX11 mode. No crashes, but weird stuttering with potplayer. I had to restart potplayer a few times.

[Asmodian]

In this case, all renderers work well with mpc-hc and potplayer except for madvr. This points to madvr. BTW I have an AMD RX 480, not nVidia.

That is not conclusive evidence it is madVR at fault. madVR uses a lot of driver features that those other renderers do not. madVR works fine with current Nvidia drivers and with older versions of AMD's drivers, nothing changed in madVR to cause this problem.

It is an issue with AMD's new drivers, they seem to have broken something for new path DX9. This had been common recently with AMD's Crimson drivers, some versions fix an issue and others introduce a new one.

I read that people with nVidia and Intel experience similar problems.

I have not had any issues with Nvidia's drivers in quite some time, madVR has been stable with all their recent drivers.

[HillieSan]

That is not conclusive evidence it is madVR at fault. madVR uses a lot of driver features that those other renderers do not. madVR works fine with current Nvidia drivers and with older versions of AMD's drivers, nothing changed in madVR to cause this problem.

It is an issue with AMD's new drivers, they seem to have broken something for new path DX9. This had been common recently with AMD's Crimson drivers, some versions fix an issue and others introduce a new one.



I have not had any issues with Nvidia's drivers in quite some time, madVR has been stable with all their recent drivers.

Indeed I have no conclusive evidence that madVR is at fault. I really hope that this is a driver issue, but the fact is that other renderers are stable (for AMD/nVidia/intel drivers) makes me wonder. This is evidence! You say that madvr is using more features of DX than other renderers (if that is true!) is also no conclusive evidence that madVR is not at fault.

I do not want to exclude the madVR implementation from the problem. I have good reasons to suspect the madvr implementation, but I do not exclude the drivers from the problem either.

BTW I am having similar issues with my GTX 660 with latest drivers.

[huhn]

DX11 is used for presentation.

madVR behaviors more like a video game than a video renderer so it's not an easy comparison.

the AMD driver are currently in a very very bad state.
just an example: http://bugs.madshi.net/view.php?id=413

and now windowed new path is completely broken on AMD.
openGL is currently pretty much broken...

i have a RX 480 myself connected to a UHD screen and i can't even get as far as you so i can't reproduce the problem sorry for that.

but you are saying the player is crashing so you should have a crash report. i would start with that.

or just use a different player and have a look at the crash report from that player.

[aufkrawall]

Does anybody else experience random player crashes when the player starts?
I'm experiencing this with MPC HC & BE x64 (latest nightlies) + latest LAV nightlies. Having this for quite some time now. Wonder if there is a connection to madVR?
Haven't tested without madVR yet since I would have to cease using it for some time to find out.
Nothing severe, opening the video file a second time usually always works.

[Georgel]

Got that part with why I would prefer to not use DSR.

what I would prefer, though, is to use a larger factor for image upsampling instead of having just chroma and luma upsampling. The algorithm for image upsampling can be better, this is why I would want an image doubling on image upsampling.

Why: Because on some anime, there is some aliasing on some edges otherwise. With everything else turned to max, maybe increasing image upsampling algorithms would yield in cleaner lines.

It could be a problem of the encoding or of the anime itself, but still wouldn't hurt to try. An anime where it's pretty proeminent is To Aru series (index, railgun, you name it). The problem is as it follows: some patters on the background will get aliasing on edges, especially for example a floor pattern that is zoomed out, gets pretty aliased. Or objects that are far in the background tend to look aliased instead of smooth (where they could look smooth). Using NNEDI chroma + JINC Image upsampling + Image doubling NNEDI + a few more bells and whistles, tried both with and without those activated.

Or maybe I haven't stricken the best combo to solve this little problem yet...



And about the drivers, madVR was stable on my ATI setup. In fact, everything was too stable on that setup. So stable that games like Dota2 run without a driver installed, but they use ATI GPU to max. Can't tell what is going on, but windows 10 drivers for ati are so bad that I get instant freeze after installing the drivers, so I run that setup (an i3 laptopy) driverless. Worked based on the theory that if it ain't broken, don't fix it..

MadVR is easily broken by bad driver support, that much is clear.

Looking forward for a madVR plugin that could be used with photoshop or other image software to enhance image quality for large pictures, if one will ever be made ^_^

[huhn]

Does anybody else experience random player crashes when the player starts?
I'm experiencing this with MPC HC & BE x64 (latest nightlies) + latest LAV nightlies. Having this for quite some time now. Wonder if there is a connection to madVR?
Haven't tested without madVR yet since I would have to cease using it for some time to find out.
Nothing severe, opening the video file a second time usually always works.

i get a one time "could not render this file" after rebooting.

not sure yet if it is constant.

@Georgel:
the to aru series is a low quality production in term of resolution. nothing unusual from J.C staff at that time.

if you sharp which you are doing with "maxing settings"
than you can easy high light aliasing.

the sharpest image is an aliased image sad but true.

[HillieSan]

DX11 is used for presentation.

madVR behaviors more like a video game than a video renderer so it's not an easy comparison.

the AMD driver are currently in a very very bad state.
just an example: http://bugs.madshi.net/view.php?id=413

and now windowed new path is completely broken on AMD.
openGL is currently pretty much broken...

i have a RX 480 myself connected to a UHD screen and i can't even get as far as you so i can't reproduce the problem sorry for that.

but you are saying the player is crashing so you should have a crash report. i would start with that.

or just use a different player and have a look at the crash report from that player.

Turn DVXA2 h/w support off in kodi. That fixed it for me with Kodi 17 beta 3. More like a workaround.

[huhn]

i don't use kodi.

[mzso]

Hi!

I had this annoyance I've been wondering for months why my ctrl+R shortcut doesn't work. I only just figured out that it's madVR's fault... I see that it's not a new thing, but I never used it so I didn't make the inference that madVR is stealing it. And it's single modifier shortcut, which makes it worse, because it's far more likely to be used by the player.

So I think it's a very bad idea that the reset statistics option doesn't have a visual feedback. I think it should display a message just like almost all the other options do. (Actually I can't think of any other hotkey function that has no visual sign.)

[neno]

This works for the x64 versions of mpc-hc and potplayer.

Now BlueSkyFRC doesn't work anymore. SVP works fine instead.

So what is the matter with BlueSkyFRC? Did you meet an issue I mentioned in #39634 and #39639?

[AngelGraves13]

So what would you suggest for upscaling to 4K?

Chroma: NNEDI3
Image: Jinc +LL +AR +SR4

Atm I use:

Chroma: SuperXBR 100, AR, SR1
Image Doubling: SuperXBR AB25
Image: Jinc AR

Chroma at Jinc SR4 as well

[aufkrawall]

Large upscaling with Jinc introduces quite some amount of aliasing, which is not the case with NNEDI3. SuperRes can be used to fight its artificial look.

[HillieSan]

So what is the matter with BlueSkyFRC? Did you meet an issue I mentioned in #39634 and #39639?

Hi Neno. I didn’t meet these issues yet.

[markanini]

SMPTE 170M is not the same as SMPTE-C, I think. madVR at least treats them differently. That also explains why madVR is using the BT.709 3dlut.


I'm confused about this. Technically 170M and SMPTE-C are identical?

Apparently they differece in black level: http://forum.doom9.org/showpost.php?p=1466457&postcount=18
MadVR output uses the same black level though.

[BatKnight]

I have been reading your opinions on scalers and have tried some combinations myself.
Since I don't understand some of the technical terms for scaling like Jinc or NEDDI3, I don't know if I am choosing the best combination for my goal.

My main goal is to watch 1080p videos at 3840x2160@23Hz.
My CPU is an Intel i5-750 with a GTX 660 with 2GB vram

My current choice is Bicubic75 AR for both chroma and image upscaling. No image doubling.
Upscaling refinement: Enhance detail 1.0
Image enhancements: 1.0 sharpen edges, 1.0 enhance detail, activated anti-bloating 100% and activated AR filter

CPU queue size 8 and GPU size 6. Windowed/Fullscreen mode both 4 frames presented in advance.
I am using D3D9 fullscreen overlay (seems to be the one providing less rendering ms)

These are giving me about 29ms at the rendering stats after the nvidia card automatically downclocks to a power saving move. It show 13ms when at the highest clock.

My question to you is, considering a 1920x1080 upscale to 3840x2160 of 4:2:0 videos:
1- how high do you think the rendering stats can go without dropped frames?
2- If I have room for improvement, which is more important PQ-wise, the image upscaling or the chroma upscaling? Why isn't there a luma upscaling choice, other than doubling?
3- In terms of image enhancements and upscaling refinements, I've noticed that some don't cost any ms, but others are quite expensive. Are there any that you might considered a must?

Sorry for the long post. Would love to read your insights on this.

Thanks
Nuno

[Warner306]

I have been reading your opinions on scalers and have tried some combinations myself.
Since I don't understand some of the technical terms for scaling like Jinc or NEDDI3, I don't know if I am choosing the best combination for my goal.

My main goal is to watch 1080p videos at 3840x2160@23Hz.
My CPU is an Intel i5-750 with a GTX 660 with 2GB vram

My current choice is Bicubic75 AR for both chroma and image upscaling. No image doubling.
Upscaling refinement: Enhance detail 1.0
Image enhancements: 1.0 sharpen edges, 1.0 enhance detail, activated anti-bloating 100% and activated AR filter

CPU queue size 8 and GPU size 6. Windowed/Fullscreen mode both 4 frames presented in advance.
I am using D3D9 fullscreen overlay (seems to be the one providing less rendering ms)

These are giving me about 29ms at the rendering stats after the nvidia card automatically downclocks to a power saving move. It show 13ms when at the highest clock.

My question to you is, considering a 1920x1080 upscale to 3840x2160 of 4:2:0 videos:
1- how high do you think the rendering stats can go without dropped frames?
2- If I have room for improvement, which is more important PQ-wise, the image upscaling or the chroma upscaling? Why isn't there a luma upscaling choice, other than doubling?
3- In terms of image enhancements and upscaling refinements, I've noticed that some don't cost any ms, but others are quite expensive. Are there any that you might considered a must?

Sorry for the long post. Would love to read your insights on this.

Thanks
Nuno

Try the guide in my signature. I'm sure you could use some form of image doubling. Downgrade chroma upscaling to Catmull-Rom and try super-xbr image doubling.

1 - 35-37 ms for 23.976 content.
2 - Image upscaling/Image doubling are most important. Chroma upscaling is not that important. Image upscaling includes both luma and chroma scaled together.
3 - SuperRes is widely regarded as the best upscaling refinement.

Try super-xbr100 + AR image doubling with SuperRes 3, ordered dithering, artifact removal - debanding.

If that is too expensive, try Jinc + AR image upscaling.

If that is too expensive, try Lanczos3 + AR image upscaling.

[Asmodian]

If you just want to know what madVR's options mean you can look at the post in my signature as well. I also cover how high the rendering time can go in the third post (more advanced topics).

[HillieSan]

So what is the matter with BlueSkyFRC?

Switching madVR to DX11 solves crashes with upscaling in FSE mode with my AMD RX 480 card. But then, blueskyFRC stops working because the stuttering/judder is back. SVP does work though. I read on a forum that blueSkyFRC does not support DX11, but this is not true.

I had contact with Bluesky and they told me

“BlueskyFRC write NV12 data in buffer allocated by the madVR.
The version of DirectX is unrelated in this process.”

Somehow the madVR connection BlueskyFRC is broken.

[har3inger]

A shot in the dark here, but have you tried enabling 1 frame per vsync under D3D11 rendering?

[ashlar42]

Can anyone point me at something if my present queue never fills up when displaying fullscreen at 2560x1440 (1080p sources)? Render times are under control but the present queue is always empty and frames get dropped. madVR 0.90.24, GTX 980, Windows 10x64, GeForce Drivers 373.06, MPC-HC x64 1.7.10.252
Happens with both D3D11 and D3D9. I use smooth motion on a 59.94Hz refresh rate monitor.

[HillieSan]

A shot in the dark here, but have you tried enabling 1 frame per vsync under D3D11 rendering?


While I was experimenting with some settings in madVR I also disabled and enabled blueskyFRC a few times. Suddenly blueskyFRC worked. I have changed the settings back and it still works. I don’t know why.

[huhn]

Can anyone point me at something if my present queue never fills up when displaying fullscreen at 2560x1440 (1080p sources)? Render times are under control but the present queue is always empty and frames get dropped. madVR 0.90.24, GTX 980, Windows 10x64, GeForce Drivers 373.06, MPC-HC x64 1.7.10.252
Happens with both D3D11 and D3D9. I use smooth motion on a 59.94Hz refresh rate monitor.

make a screen of the OSD please.

it doesn't really matter but 60p should a little little bit better than 59p with smoothmotion.

[ashlar42]

make a screen of the OSD please.

it doesn't really matter but 60p should a little little bit better than 59p with smoothmotion.The same happens with Exclusive Mode. I used Windowed for ease of screen taking (if there's a way of grabbing it with Exclusive I didn't find it quickly).

What's strange are Max Stats. In Average Stats those high values never appear but as soon as I go fullscreen the Max Stats jump to these incredibly high values.

http://i68.tinypic.com/2nbais5.jpg

[robl45]

quick question, why does madvr display composition rate of 24hz when display is 23.976 when playing with mpc-be/HC but shows composition rate of 23.976 when display is 23.976 in Jriver media center?

[huhn]

The same happens with Exclusive Mode. I used Windowed for ease of screen taking (if there's a way of grabbing it with Exclusive I didn't find it quickly).
it is maybe possible to take FSE screens with things like fraps.
but ignore this windowed is fine for now.

What's strange are Max Stats. In Average Stats those high values never appear but as soon as I go fullscreen the Max Stats jump to these incredibly high values.

http://i68.tinypic.com/2nbais5.jpg

are you using a 3D LUT?
try control+shift+3 and see if it makes a huge difference.

have a look at the CPU usages even the subtitles are not buffered.

and you haven't changed the nvidia "manage 3D settings" right?
quick question, why does madvr display composition rate of 24hz when display is 23.976 when playing with mpc-be/HC but shows composition rate of 23.976 when display is 23.976 in Jriver media center?

the composition rate is the refresh rate windows is using for the desktop.

24hz with a 23p source is pretty much save.
this type of error happens a lot with windows 7

[ashlar42]

are you using a 3D LUT?
try control+shift+3 and see if it makes a huge difference.

have a look at the CPU usages even the subtitles are not buffered.

and you haven't changed the nvidia "manage 3D settings" right?No, I'm not using 3D LUT. CTRL+SHIFT+3 didn't make an ounce of a difference.
Nvidia 3D settings are at default values for MPC-HC (both x86 and x64 versions, all at "use global settings").

Now, while checking CPU usage (which stays quite low, I'm on an i74770K, decoding with DXVA copyback h.264 content), I discovered an even stranger (to me, at least) thing. As you can see if the task manager window sits on top of MPC-HC window the present queue fills up normally. This happens even if I switch the focus to MPC-HC and away from task manager (which remains on top because I've activated the relevant function). As soon as I close task manager, present queue empties.

http://i66.tinypic.com/bf4dnk.jpg

[ashlar42]

I'm currently investigating f.lux, as it might be the reason this is happening. I've been using it for a very long time and in the past it didn't give me problems. But in July I updated to Windows 10 and later Anniversary... not using often fullscreen on my desktop PC (I usually watch movies from HTPC) I might not have noticed this before.

Activating f.lux "Safe Mode" seems to solve the problem. As it doesn't significantly impair f.lux performance or functionality, all should be good. In case it could help madshi understand, Safe Mode in f.lux does the following (per their FAQ):

1. It disables our layered window for compatibility with some older machines. 2. It disables all polling we normally do to ensure that we're the active color profile.

[robl45]

The problem is jriver media center is working correctly and mpc-hc/be aren't. I'm randomly getting 23.976 or 24 in the latter while jriver is always 23.976.



the composition rate is the refresh rate windows is using for the desktop.

24hz with a 23p source is pretty much save.
this type of error happens a lot with windows 7

[huhn]

there is nothing more i can do than pointing to this: http://bugs.madshi.net/view.php?id=204

[robl45]

there is nothing more i can do than pointing to this: http://bugs.madshi.net/view.php?id=204

Thats a little different, that says while the player is open and switching refresh and I am closing the player and madvr each time. Further that says its fixed in 8.1 and I'm using 10.

[huhn]

the point is madVR doesn't have control over it.

[Shiandow]

I'm currently investigating f.lux, as it might be the reason this is happening. I've been using it for a very long time and in the past it didn't give me problems. But in July I updated to Windows 10 and later Anniversary... not using often fullscreen on my desktop PC (I usually watch movies from HTPC) I might not have noticed this before.

Activating f.lux "Safe Mode" seems to solve the problem. As it doesn't significantly impair f.lux performance or functionality, all should be good. In case it could help madshi understand, Safe Mode in f.lux does the following (per their FAQ):

1. It disables our layered window for compatibility with some older machines. 2. It disables all polling we normally do to ensure that we're the active color profile.

Wouldn't be the first time f.lux is causing problems, two programs vying for complete control of the screen usually ends badly. It's also worth pointing out that for maximum image quality you're best of disabling f.lux during playback, as they don't bother to use dithering which results in quite horrible banding.

[sauma144]

I have always has problems with f.lux and madVR or games.
Maybe we could add a notice somewhere about it.

Is there a way for madVR to disable this kind of software when it is running?
If so why not adding an option/checkbox in madVR to do this?

[ashlar42]

I have always has problems with f.lux and madVR or games.
Maybe we could add a notice somewhere about it.

Is there a way for madVR to disable this kind of software when it is running?
If so why not adding an option/checkbox in madVR to do this?As I mentioned, putting f.lux in Safe Mode solves my problem. Although I haven't really tested if it disables dithering or not... but it's a quick test.

Let me check (goes checking...)

...

(comes back...)

In my case f.lux does not influence dithering at all. Just the present queue, problem you can solve by placing it in Safe Mode. Unless... unless Shiandow meant that the color correction done by f.lux doesn't use dithering, hence banding appearing. I tested for madVR dithering but maybe it was the wrong test.

[Dreamland]

I have a question:
Radeon RX 480 is much better than Radeon R9 380 for MadVR?
thx a lot!

[ibius]

I have always has problems with f.lux and madVR or games.
Maybe we could add a notice somewhere about it.

Is there a way for madVR to disable this kind of software when it is running?

Not exactly 'disable', but bypass, using Overlay mode, requires Nvidia GPU, and should also give slightly better performance, but it has a few cosmetic issues.