madVR - high quality video renderer (GPU assisted)

[madshi]

Would you please increase this to 2 sec at least? I have a rather slow rig where some seeks get into this interval without a black screen, whereas some don't and I see a black screen flashing. So it's a tad irritating to have such fickle seek experience.
Ok, will do.

By the way, are there any known issues with "use Direct3D 11 for presentation" or "use separate device for presentation" with HD4000 IGPs? On my laptop using either of these options leads to major presentation issues (frames jumping about all over the place). Not a big deal since I can disable both options, just wondering if you knew.
This is a known problem. Probably a GPU driver issue. At the moment there's not much I can do about it because I can't reproduce it on my PC. However, I'll probably get a Skylake laptop soon, then hopefully I can reproduce it, and then *maybe* do something about it.

the problem does not seem to be related to D3D11 supported refresh rates. It happens no matter the refresh rate, from 23.976 to 60Hz. As I said, the issue only happens when madVR sends 10bit per color channel output, and the TV receives it (well, it says it receives 12bit per channel). And that happens only when using D3D11 + fullscreen, and the TV set configured as "10bit or more" in the madVR settings.
Well, what can I say. I don't think this is likely to be madVR's fault. I'm not doing something completely different when outputting 10bit. It's only a small difference in my code. So there's probably nothing I can do about it.

Just a follow up. The bug I reported has been fixed in the latest release.
Great, good to know, thanks. However, in low latency mode this problem will likely be back. In any case, it was never madVR's fault.

I need help to understand what I'm going to set the MPC-HC and Madvr.
I have a HTPC that is connected with HDMI to AV Marantz 6900sr and play with EPSON Tw7200.
Should I set something on Madvr or MPC, or fix Marantz everything.
Your question is very unspecific. I wouldn't know where to start. It might be best for you to either read one of the many madVR guides/tutorials. Or alternatively ask a more specific question.

Is there any advantage to using smooth motion when your refresh rate is an exact multiple of the video frame rate or is it just wasted resources?
I would disable it in that case, but then you will want to fine tune the refresh rate as much as possible so you don't get any (or at least not more than 1-2) frame drops/repeats during the duration of a movie.

http://abload.de/image.php?img=353.62x5pbz.png
OMGosh. I thought they would at least have replaced "cl_nv_d3d9_sharing" with "cl_khr_dx9_media_sharing", which would have been the proper way to handle the situation. After all, they seem to claim to support OpenCL 1.2 now. But no, they simply disabled OpenCL <-> D3D9 interop *completely*. Are they bonkers??

And no, I'm not going to be bullied by NVidia to drop OpenCL. If you madVR users want to continue using NNEDI3 with NVidia GPUs with newer NVidia drivers, I'd kindly ask you to complain to NVidia. Anybody got any contacts to NVidia? I think there were some users with NVidia contacts in the past. Basically what NVidia should do is bring D3D9 <-> OpenCL interop support back, either through "cl_nv_d3d9_sharing" or "cl_khr_dx9_media_sharing". Or do they really want to be the only GPU company without any sort of D3D9 <-> OpenCL interop support!?!?

(I may add support for PS5.0 NNEDI3 at some point in the future, but not soon, and I won't change my time schedule because NVidia is stupid enough to remove D3D9 OpenCL interop.)

[har3inger]

It's not motion blur. It's ghosting. It may not even be noticeable ghosting if your display already has slow response time.

Smooth motion also costs more than it seems. Normally 24 fps content needs to have total render/present times under 42ms or queues start emptying. With smooth motion on with a 60 fps display, queues start emptying around 38ms or so until you start dropping all your frames. This is because smooth motion renders the 24 fps source frames as well as additional combined frames when a frame transition fails to coincide with a vsync (which depends on ratio of source to display fps).

[aufkrawall]

And no, I'm not going to be bullied by NVidia to drop OpenCL.

Well, it never worked for me on powerful AMD GPUs with PCIe 2.0 either.
Shiandow's SM5.0 implementation is not much slower than OpenCL, so probably everyone would profit if you'd replace the OpenCL implementation with it (of course I don't know how much work this integration would be).


If you madVR users want to continue using NNEDI3 with NVidia GPUs with newer NVidia drivers, I'd kindly ask you to complain to NVidia. Anybody got any contacts to NVidia?

I'm already trying to find out why this happened. Can't promise anything, though.

[huhn]

OMGosh. I thought they would at least have replaced "cl_nv_d3d9_sharing" with "cl_khr_dx9_media_sharing", which would have been the proper way to handle the situation. After all, they seem to claim to support OpenCL 1.2 now. But no, they simply disabled OpenCL <-> D3D9 interop *completely*. Are they bonkers??

And no, I'm not going to be bullied by NVidia to drop OpenCL. If you madVR users want to continue using NNEDI3 with NVidia GPUs with newer NVidia drivers, I'd kindly ask you to complain to NVidia. Anybody got any contacts to NVidia? I think there were some users with NVidia contacts in the past. Basically what NVidia should do is bring D3D9 <-> OpenCL interop support back, either through "cl_nv_d3d9_sharing" or "cl_khr_dx9_media_sharing". Or do they really want to be the only GPU company without any sort of D3D9 <-> OpenCL interop support!?!?

(I may add support for PS5.0 NNEDI3 at some point in the future, but not soon, and I won't change my time schedule because NVidia is stupid enough to remove D3D9 OpenCL interop.)

maybe there way to fix this issue: http://bugs.madshi.net/view.php?id=250

i have a response from a "Customer care" so nothing world moving: https://forums.geforce.com/default/topic/856761/geforce-drivers/windows-10-display-driver-feedback-thread-7-16-15-/post/4616215/#4616215

feel free to say you have the same issue but please don't over do it just say you have the same issue and you don't like it. maybe 10 page will speed things up.

[madshi]

Well, it never worked for me on powerful AMD GPUs with PCIe 2.0 either.
Shiandow's SM5.0 implementation is not much slower than OpenCL, so probably everyone would profit if you'd replace the OpenCL implementation with it
Last time I checked, the OpenCL implementation was noticably faster with AMD, and slightly faster with NVidia. Of course with AMD it depends on PCIe speed. But a lot of AMD users with PCIe 3.0 would be unhappy if I suddenly made NNEDI3 speed quite a bit slower for them.

Anyway, as I said, I might look into PS5.0 NNEDI3 as an option to OpenCL at some point in the future. But definitely not soon.

I'm already trying to find out why this happened. Can't promise anything, though.
Thanks.

i have a response from a "Customer care" so nothing world moving: https://forums.geforce.com/default/topic/856761/geforce-drivers/windows-10-display-driver-feedback-thread-7-16-15-/post/4616215/#4616215
Thanks.

[mogli]

Not really. But different users had a couple of different weird issues. Try v0.88.21.
It wasn't the new version that fixed it, but switching once from D3D11 to D3D9 (no problems there) and then back. Now it doesn't happen anymore. :confused:

More weird stuff:
Default settings, but flush after copy to backbuffer in FSE. With D3D9 everything's fine but with D3D11 the render queue doesn't fill completely. It is pushed down by the present queue: Increasing the present queue (or decreasing the render queue) will give back the render queue space to live.

A lot of funny things happening in the last few releases. :D

[madshi]

It wasn't the new version that fixed it, but switching once from D3D11 to D3D9 (no problems there) and then back. Now it doesn't happen anymore. :confused:

More weird stuff:
Default settings, but flush after copy to backbuffer in FSE. With D3D9 everything's fine but with D3D11 the render queue doesn't fill completely. It is pushed down by the present queue: Increasing the present queue (or decreasing the render queue) will give back the render queue space to live.

A lot of funny things happening in the last few releases. :D
Is that extra flush needed? Did you always have it in there? And can you double check if any of these problems are really new in the latest version, compared to v0.88.16 (which was the last build before all the OSD related changes).

[Sunset1982]

what is the last nvidia driver working with NNEDI3? And is there a working version for windows 10?

[mogli]

Is that extra flush needed?No.
Did you always have it in there?No, never.
Just found this stuff by accident while testing and thought I report just another strange D3D11 behaviour.
And can you double check if any of these problems are really new in the latest version, compared to v0.88.16 (which was the last build before all the OSD related changes).This seems to be a D3D11 issue. Why should it have anything to to with the OSD changes?

[aufkrawall]

what is the last nvidia driver working with NNEDI3? And is there a working version for windows 10?
353.49 hotfix driver is best for this purpose. It's not WHQL, but unlike 353.30, it contains important fixes to prevent driver crashes on the desktop/while surfing.

Edit: Fixed driver numbers.

[huhn]

https://forums.geforce.com/default/topic/856761/geforce-drivers/windows-10-display-driver-feedback-thread-7-16-15-/post/4616450/#4616450

he messed the quote up but this part is from him: "I'll file a bug for this"

i guess this is the limited a customer care worker at nvidia can do.

[madshi]

he messed the quote up but this part is from him: "I'll file a bug for this"

i guess this is the limited a customer care worker at nvidia can do.
Yeah, it's worth a try.

Just found this stuff by accident while testing and thought I report just another strange D3D11 behaviour.
This seems to be a D3D11 issue. Why should it have anything to to with the OSD changes?
Well, you said "A lot of funny things happening in the last few releases", and many of the weird things lately had to do with the OSD changes. If it's just some strange D3D11 behaviour, and if it was always there since D3D11 was introduced, then it's got nothing to do with the OSD changes, of course.

[Akeno]

I would disable it in that case, but then you will want to fine tune the refresh rate as much as possible so you don't get any (or at least not more than 1-2) frame drops/repeats during the duration of a movie.

How would I fine tune it? I'm on a 60hz locked laptop display so does that necessitate ReClock's framerate adjustment feature?

there is no general answer if you get an repeated or dropped frame every 4 h or lower than yes it is still pretty useful.

if you have an very unlikely perfect synced audio and video clock no it just wastes some system performance in this case.

Could you explain to me what by synced clocks? I read something about mismatched clocks in the ReClock thread but no real explanation as to what that means. I always set ReClock to play back at the original speed rather than its auto function but does that mean I'm negating the benefits of ReClock in the first place?

[huhn]

How would I fine tune it? I'm on a 60hz locked laptop display so does that necessitate ReClock's framerate adjustment feature?

you are on a 60 hz device so you should use smooth motion and reclock is not really useful in this case.
fine tuning is a complicated thing about custom refresh rates. i don't see a real use for this on a 60 hz only display.

Could you explain to me what by synced clocks? I read something about mismatched clocks in the ReClock thread but no real explanation as to what that means. I always set ReClock to play back at the original speed rather than its auto function but does that mean I'm negating the benefits of ReClock in the first place?

audio and video have different clocks so they get more and more async over the time so the video renderer is dropping or repeating frames from time to time to fix this. dropping audio "frames" is not a viable option. reclock can kind of fix this issue by changing the speed of audio by resampling it but a usually 24 hz movie can't be fixed this way on a 60 hz screen but smooth motion can do this easily.

[Akeno]

audio and video have different clocks so they get more and more async over the time so the video renderer is dropping or repeating frames from time to time to fix this. dropping audio "frames" is not a viable option. reclock can kind of fix this issue by changing the speed of audio by resampling it but a usually 24 hz movie can't be fixed this way on a 60 hz screen but smooth motion can do this easily.

So ReClock is a waste with smooth motion then? I usually play 23.976hz material and always kept it playing at original speed rather than the ReClock preferred 24hz.

[huhn]

So ReClock is a waste with smooth motion then? I usually play 23.976hz material and always kept it playing at original speed rather than the ReClock preferred 24hz.

yes it is. you can't do 24 hz anyway.

you could use reclock to use wasapi but there are other better way to use wasapi. and reclock is 32 bit only...

for me 64 bit is the way to go can be a lot faster.

[Arm3nian]

I have a 980 ti and a 4k monitor on the way, dis gon b' gud.

[6ari8]

Ok, please try again with v0.88.21.


The stuttering issue with all OSD elements (excluding the seekbar) is fixed in v0.88.21.

[Sunset1982]

Is GPU-RAM a big point using madvr?

I'm thinking of getting a MSI R9 380 and I dont know if its worth the extra 50 € for the 4GB version instead of the 2 GB version.

[ryrynz]

Is GPU-RAM a big point using madvr?

I'm thinking of getting a MSI R9 380 and I dont know if its worth the extra 50 € for the 4GB version instead of the 2 GB version.

It isn't, but if you want to use it for gaming too then by all means get the extra.

[nevcairiel]

Is GPU-RAM a big point using madvr?

I'm thinking of getting a MSI R9 380 and I dont know if its worth the extra 50 € for the 4GB version instead of the 2 GB version.

If you plan on doing anything with 4K, then you will want the extra memory.

[Sunset1982]

If you plan on doing anything with 4K, then you will want the extra memory.

It isn't, but if you want to use it for gaming too then by all means get the extra.

Thanks guys! I ordered the 4GB version. ;)

[Thunderbolt8]

It isn't, but if you want to use it for gaming too then by all means get the extra.Id say it can be, but primarily not because of the amount, but because of the type. you really want DDR5 RAM (or anything faster if newer stuff is out, Im not up to date here) and not the slower DDR3 RAM or anything like that.

[ryrynz]

Id say it can be, but primarily not because of the amount, but because of the type. you really want DDR5 RAM (or anything faster if newer stuff is out, Im not up to date here) and not the slower DDR3 RAM or anything like that.

Didn't answer it as a generic question as it was purely about RAM size in this particular case.

I believe unless it's 4K >24FPS you'll be fine with 2GB.. but things always move forward, 2GB is getting a bit pokey these days so it's worth getting the extra.
I chew up about 1.2GB+ with 4K content.

[edigee]

If you plan on doing anything with 4K, then you will want the extra memory.
My GTX 960 is 2Gb of memory. It handles any 4k content with no problems

[metadethrax]

Guys, I really need your expertise with a problem I am encountering.
First, lots of thank you's to madshi for his continuous work, now a few words about my setup: I am using a GTX 970, i5-760, 8 GB RAM for watching 1080p-only high quality mkv encodes. I am also using mpc-hc x64 with latest madvr and LAV Filters. I am sending the video signal through HDMI from my computer to an Onkyo receiver that is connected to a 46" Samsung TV set.
madvr's settings are as follows: chroma upscaling super-xbr (sharpness: 100, anti-ringing and SuperRes filters checked), image downscaling Catmull-Rom (anti-ringing and scale in linear light checked), image doubling always super-xbr sharpness:100, image upscaling Lanczos (3 taps, activate anti-ringing filter checked), upscaling refinement SuperRes (strenght 4, radius 0.66, refine the image only once after upscaling is complete and apply SuperRes first checked). I don't know if these are the best settings for 1080p encodes and for my system, any recommendation is welcomed, but without ReClock I am having a lot of repeated frames, one frame at 40+ seconds. ReClock is not an option for x64 and for my setup in particular, as LAV Audio bitstreams the audio to the receiver, therefore ReClock won't work, even when using x86 versions of mpchc and madvr. I have attached a screenshot, maybe you will be able to tell me what to do with the repeated frames problem.
Thank you for your help!

[Ver Greeneyes]

Id say it can be, but primarily not because of the amount, but because of the type. you really want DDR5 RAM (or anything faster if newer stuff is out, Im not up to date here) and not the slower DDR3 RAM or anything like that.Note that graphics cards tend to use a special type of memory with a 'G' prefix, so GDDR3, GDDR4, GDDR5. Some cheaper cards do use regular DDR3 memory to save on costs, but DDR5 memory doesn't exist yet - and when it does, its primary target will be standalone sticks of RAM, not GPUs :)

[Akeno]

Guys, I really need your expertise with a problem I am encountering.
First, lots of thank you's to madshi for his continuous work, now a few words about my setup: I am using a GTX 970, i5-760, 8 GB RAM for watching 1080p-only high quality mkv encodes. I am also using mpc-hc x64 with latest madvr and LAV Filters. I am sending the video signal through HDMI from my computer to an Onkyo receiver that is connected to a 46" Samsung TV set.
madvr's settings are as follows: chroma upscaling super-xbr (sharpness: 100, anti-ringing and SuperRes filters checked), image downscaling Catmull-Rom (anti-ringing and scale in linear light checked), image doubling always super-xbr sharpness:100, image upscaling Lanczos (3 taps, activate anti-ringing filter checked), upscaling refinement SuperRes (strenght 4, radius 0.66, refine the image only once after upscaling is complete and apply SuperRes first checked). I don't know if these are the best settings for 1080p encodes and for my system, any recommendation is welcomed, but without ReClock I am having a lot of repeated frames, one frame at 40+ seconds. ReClock is not an option for x64 and for my setup in particular, as LAV Audio bitstreams the audio to the receiver, therefore ReClock won't work, even when using x86 versions of mpchc and madvr. I have attached a screenshot, maybe you will be able to tell me what to do with the repeated frames problem.
Thank you for your help!

Do you notice the repeated frames? I often don't so if you can't notice it, don't let it bother you. The other option is to enable Smooth Motion to blend frames and mitigate the problem.

[ashlar42]

yes it is. you can't do 24 hz anyway.

you could use reclock to use wasapi but there are other better way to use wasapi. and reclock is 32 bit only...

for me 64 bit is the way to go can be a lot faster.How to use wasapi without reclock on 64 bit? Pardon my ignorance and the offtopic.

[huhn]

How to use wasapi without reclock on 64 bit? Pardon my ignorance and the offtopic.

looks at mpc-be or mpc-hc newest nightly builds

[huhn]

My GTX 960 is 2Gb of memory. It handles any 4k content with no problems

this was more about gaming. but 1080p nnedi3 -> uhd can eat a lot too.

[Sunset1982]

@metadethrax

you have to create a 23,976 custom resolution within the nvidia drivers to get as close as possible to 23,976.

[chros]

As for me, it's not important if this gets fixed or not, I am not using fullscreen mode, as it's too annoying that lag which happens when I try to display the context menu in MPC-HC to switch the audio or subtitle track.
Have you ever tried a remote app for MPC-HC like Remoti.co?
https://play.google.com/store/apps/details?id=com.kdombrovsky.mpcremote
I can't live without it :) Once you started MPC-HC you can do a lot of things with it, like rotating through audio and subtitle tracks and even more. And it won't affect FSE playback.

[XMonarchY]

So NVidia is completely ditching the necessary OpenCL extension for NNEDI3 in Windows 10 drivers? Or is it just early/first official drivers since official Windows 10 release date?

[aufkrawall]

We don't know. Maybe it will appear in a few weeks again (I doubt this), in a few months or never.

[XMonarchY]

What's the exact info do I need to submit to NVidia to properly inform them of this problem? I communicated with NVidia a while back regarding madVR and I know NVidia is aware of madVR's importance to HTPC world. They can't ignore it.

[aufkrawall]

Either complain in the forums:
https://forums.geforce.com/default/topic/860152/geforce-drivers/official-windows-10-353-62-game-ready-display-driver-feedback-thread-7-29-15-/

Or submit a bug report via support form:
http://nvidia-submit.custhelp.com/app/ask

[baii]

windows 10 WU amd driver seem to fix the bugs in preview drivers. brief test show no artifact, flicker or scalerer not working.

[edigee]

Either complain in the forums:
https://forums.geforce.com/default/topic/860152/geforce-drivers/official-windows-10-353-62-game-ready-display-driver-feedback-thread-7-29-15-/

Or submit a bug report via support form:
http://nvidia-submit.custhelp.com/app/ask
http://forums.guru3d.com/showpost.php?p=5131647&postcount=14
As you can see something really weird is going on.
I really hope this is not NVIDIA playing dirty.

[KoD]

Have you ever tried a remote app for MPC-HC like Remoti.co?
https://play.google.com/store/apps/details?id=com.kdombrovsky.mpcremote
I can't live without it :) Once you started MPC-HC you can do a lot of things with it, like rotating through audio and subtitle tracks and even more. And it won't affect FSE playback.
Thanks, I'll have a look! :)

[aufkrawall]

http://forums.guru3d.com/showpost.php?p=5131647&postcount=14
As you can see something really weird is going on.
I really hope this is not NVIDIA playing dirty.
Maybe, but maybe it's just guru3d staff doing things by mistake.
They were also very incompetent when that GTX 970 issue came up. Well, even their reviews suck. But their forums are highly frequented by users and people were complaining there that NNEDI3 stopped working.
My post which disappeared was information about the issue and asked users to complain to Nvidia about it.

[MSL_DK]

My render, upload and present queue doesn't get filled (1-3/16) when using 1080p60 no problem if i choose 1080p23 (FSE) in win 10 using AMD card no problem in windowed mode.

[bcec]

353.49 hotfix driver is best for this purpose. It's not WHQL, but unlike 353.30, it contains important fixes to prevent driver crashes on the desktop/while surfing.

Edit: Fixed driver numbers.

Is that for Win 10 or Win 7? Is this issue with Win10 only or is Win7 also affected? (sorry, late to the party)

[Anima123]

nVidia driver 353.62 now available at nVidia's website. Is the same issue of NNEDI3 not working still there?

[Ver Greeneyes]

A new GeForce driver is out to accompany the release of Windows 10, only available on www.geforce.com at the moment:

32-bit GeForce Game Ready Driver 353.62 for Windows Vista, 7, 8 and 8.1 (http://www.geforce.com/drivers/results/87813)
64-bit GeForce Game Ready Driver 353.62 for Windows Vista, 7, 8 and 8.1 (http://www.geforce.com/drivers/results/87831)
32-bit GeForce Windows 10 Driver 353.62 (http://www.geforce.com/drivers/results/87921)
64-bit GeForce Windows 10 Driver 353.62 (http://www.geforce.com/drivers/results/87939)

Has anyone tried these yet to see if they have D3D9 OpenCL interop?

[aufkrawall]

Is that for Win 10 or Win 7? Is this issue with Win10 only or is Win7 also affected? (sorry, late to the party)
You can easily find this out by yourself:
http://nvidia.custhelp.com/app/answers/detail/a_id/3698/~/geforce-hotfix-driver-353.49

nVidia driver 353.62 now available at nVidia's website. Is the same issue of NNEDI3 not working still there?
It's the very same driver as provided via Windows Update (just sorts out a control panel issue on x64) and so of course NNEDI3 OpenCL isn't working.
Miracles can happen, but better don't expect a fix anytime soon.

[XMonarchY]

You can easily find this out by yourself:
http://nvidia.custhelp.com/app/answers/detail/a_id/3698/~/geforce-hotfix-driver-353.49


It's the very same driver as provided via Windows Update (just sorts out a control panel issue on x64) and so of course NNEDI3 OpenCL isn't working.
Miracles can happen, but better don't expect a fix anytime soon.

In case NVidia ditches the needed extension, is there a chance madshi will create a work-around via Direct Compute or something similar? I mean its obviously NVidia's problem, but that means Windows 10 will not be viable for madVR for NVidia users, which is like watching a cute red panda die :( , but we all kow madshi would never let a red panda die :) .
http://blog.chron.com/momhouston/files/legacy/toby.jpg

[huhn]

Guys, I really need your expertise with a problem I am encountering.
First, lots of thank you's to madshi for his continuous work, now a few words about my setup: I am using a GTX 970, i5-760, 8 GB RAM for watching 1080p-only high quality mkv encodes. I am also using mpc-hc x64 with latest madvr and LAV Filters. I am sending the video signal through HDMI from my computer to an Onkyo receiver that is connected to a 46" Samsung TV set.
madvr's settings are as follows: chroma upscaling super-xbr (sharpness: 100, anti-ringing and SuperRes filters checked), image downscaling Catmull-Rom (anti-ringing and scale in linear light checked), image doubling always super-xbr sharpness:100, image upscaling Lanczos (3 taps, activate anti-ringing filter checked), upscaling refinement SuperRes (strenght 4, radius 0.66, refine the image only once after upscaling is complete and apply SuperRes first checked). I don't know if these are the best settings for 1080p encodes and for my system, any recommendation is welcomed, but without ReClock I am having a lot of repeated frames, one frame at 40+ seconds. ReClock is not an option for x64 and for my setup in particular, as LAV Audio bitstreams the audio to the receiver, therefore ReClock won't work, even when using x86 versions of mpchc and madvr. I have attached a screenshot, maybe you will be able to tell me what to do with the repeated frames problem.
Thank you for your help!

movies are usually 24000/1001 hz (it is called 23p) not 24 hz, you should use 23p not 24p for this file in question.

In case NVidia ditches the needed extension, is there a chance madshi will create a work-around via Direct Compute or something similar? I mean its obviously NVidia's problem, but that means Windows 10 will not be viable for madVR for NVidia users, which is like watching a cute red panda die :( , but we all kow madshi would never let a red panda die :) .

http://forum.doom9.org/showpost.php?p=1731949&postcount=32210

wasn't he clear enough?

[XMonarchY]

movies are usually 24000/1001 hz (it is called 23p) not 24 hz, you should use 23p not 24p for this file in question.



http://forum.doom9.org/showpost.php?p=1731949&postcount=32210

wasn't he clear enough?


In the past there was another similar issue that was NVidia's fault, but madshi found a way around it and NVidia never fixed it...

[baii]

Can't you like dump the dlls into the player's folder for it to work? noob question~