Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > Hardware & Software > Software players

Reply
 
Thread Tools Search this Thread Display Modes
Old 26th April 2014, 16:27   #26241  |  Link
Anime Viewer
Troubleshooter
 
Anime Viewer's Avatar
 
Join Date: Feb 2014
Posts: 333
Nvidia using OpenCL 1.1 AMD & Intel using 1.2

seidweise over on the KCP Codec forum put out a program that detects OpenCL on systems. Looks like AMD and Intel have the lead over Nvidia when it comes to the OpenCL version implemented on their devices. It makes me wonder if the Nvidia bug is related to Nvidia using the 1.1 version of OpenCL. Anyone feel like installing one of the old drivers that correctly functioned with OpenCL/NNEDI3, and then run seidweise's program to see if maybe the old drivers were using 1.0 compared to the new drivers which are using 1.1?
__________________
System specs: Sager NP9150 SE with i7-3630QM 2.40GHz, 16 GB RAM, 64-bit Windows 10 Pro, NVidia GTX 680M/Intel 4000 HD optimus dual GPU system. Video viewed on LG notebook screen and LG 3D passive TV.

Last edited by Anime Viewer; 26th April 2014 at 16:39.
Anime Viewer is offline   Reply With Quote
Old 26th April 2014, 16:40   #26242  |  Link
huhn
Registered User
 
Join Date: Oct 2012
Posts: 5,723
Quote:
Originally Posted by Anime Viewer View Post
seidweise over on the KCP Codec forum put out a program that detects OpenCL on systems. Looks like AMD and Intel have the lead over Nvidia when it comes to the OpenCL version implemented on their devices. It makes me wonder if the Nvidia bug is related to Nvidia using the 1.1 version of OpenCL. Anyone feel like installing one of the old drivers that correctly functioned with OpenCL/NNEDI3, and then run seidweise's program to see if maybe the old drivers were using 1.0 compared to the new drivers which are using 1.1?
you are aware that this bug doesn't matter for MadVR at all anymore?
huhn is offline   Reply With Quote
Old 26th April 2014, 16:53   #26243  |  Link
wolfman2791
Registered User
 
wolfman2791's Avatar
 
Join Date: Jan 2014
Posts: 55
I ran that check on my NVIDIA GTX 660... and with my driver version (337.50), the OPEN CL version is 1.1. Interestingly, the CPU is running version 1.2 lol. OPEN CL's latest version is 2.0...
wolfman2791 is offline   Reply With Quote
Old 26th April 2014, 16:54   #26244  |  Link
wolfman2791
Registered User
 
wolfman2791's Avatar
 
Join Date: Jan 2014
Posts: 55
Quote:
Originally Posted by huhn View Post
you are aware that this bug doesn't matter for MadVR at all anymore?
It doesn't. But i'm sure newer versions of OPEN CL run more efficiently regardless of MadVR, no? Just like updating your driver makes MadVR run faster too typically.
wolfman2791 is offline   Reply With Quote
Old 26th April 2014, 21:58   #26245  |  Link
Asmodian
Registered User
 
Join Date: Feb 2002
Location: San Jose, California
Posts: 3,597
Quote:
Originally Posted by Anime Viewer View Post
Anyone feel like installing one of the old drivers that correctly functioned with OpenCL/NNEDI3, and then run seidweise's program to see if maybe the old drivers were using 1.0 compared to the new drivers which are using 1.1?
From what I understand based on madshi talking about the workaround it has nothing to do with the OpenCL version but more likely an optimization or an unintended consequence from something Nvidia changed.

The interop would report finished so madVR cleared the memory and started on the next frame but the interop wasn't really done. Not clearing the memory until later solved the black/green screen. This isn't something that would change in the specs between 1.0 and 1.1. I suppose it could be a bug present in Nvidia's implementation of OpenCL 1.1 that wasn't present in their implementation of OpenCL 1.0 but even if true I am still not sure what benefit knowing this would give us.

Quote:
Originally Posted by wolfman2791 View Post
It doesn't. But i'm sure newer versions of OPEN CL run more efficiently regardless of MadVR, no? Just like updating your driver makes MadVR run faster too typically.
I don't think newer versions of OpenCL would run faster only due to the new version. New versions have more features, but if madVR doesn't use these new features they do not change anything. The reason new drivers are sometimes faster is due to optimizations not features sets.

A new feature sets might offer a faster solution for a specific task compared to a previous version but the developer would need to implement them for them to do anything.
Asmodian is offline   Reply With Quote
Old 26th April 2014, 22:47   #26246  |  Link
Soukyuu
Registered User
 
Soukyuu's Avatar
 
Join Date: Apr 2012
Posts: 169
Quote:
Originally Posted by Anime Viewer View Post
seidweise over on the KCP Codec forum put out a program that detects OpenCL on systems. Looks like AMD and Intel have the lead over Nvidia when it comes to the OpenCL version implemented on their devices. It makes me wonder if the Nvidia bug is related to Nvidia using the 1.1 version of OpenCL. Anyone feel like installing one of the old drivers that correctly functioned with OpenCL/NNEDI3, and then run seidweise's program to see if maybe the old drivers were using 1.0 compared to the new drivers which are using 1.1?
Ha. I still have the old 327.23 driver installed. You're right, the old one says "OpenCL 1.0 CUDA" for me.

Quote:
Originally Posted by huhn View Post
you are aware that this bug doesn't matter for MadVR at all anymore?
It's still a bug. nVidia's stance of "if there is a workaround, don't bother fixing" is bound to lead to this bug reapppearing somewhere else again.
__________________
AMD Phenom II X4 970BE | 12GB DDR3 | nVidia 260GTX | Arch Linux / Windows 10 x64 Pro (w/ calling home shut up)

Last edited by Soukyuu; 26th April 2014 at 22:50.
Soukyuu is offline   Reply With Quote
Old 27th April 2014, 01:44   #26247  |  Link
Ceremony
Registered User
 
Join Date: Jan 2014
Posts: 94
iGPU of an A10-7850K is performing remarkably well:
using Jinc as the upscaling algorithm, it can play any 1080 or below resolution flawlessly.

Unfortunately, 48fps or higher content is out of question (at these settings), as well as any NNEDI upscaling or the smooth motion feature.

Hopefully, it'll become e
Ceremony is offline   Reply With Quote
Old 27th April 2014, 02:47   #26248  |  Link
Mangix
Audiophile
 
Join Date: Oct 2006
Posts: 353
if Nvidia truly supports OpenCL 1.1, shouldn't this line of code be changed to use || ?

Code:
#if defined(__GPU__) && defined(__AMD__) && __OPENCL_VERSION__ >= 110
Mangix is offline   Reply With Quote
Old 27th April 2014, 16:53   #26249  |  Link
James Freeman
Registered User
 
Join Date: Sep 2013
Posts: 919
madshi, I think we're ready for a high quality Frame Interpolation algorithm now...
__________________
System: i7 3770K, GTX660, Win7 64bit, Panasonic ST60, Dell U2410.
James Freeman is offline   Reply With Quote
Old 27th April 2014, 17:03   #26250  |  Link
wolfman2791
Registered User
 
wolfman2791's Avatar
 
Join Date: Jan 2014
Posts: 55
I get the sense that the NVIDIA OpenCL bug will likely get fixed in the driver that upgrades from version 1.1 to 1.2 since it seems that the bug occured from 1.0 to 1.1.
wolfman2791 is offline   Reply With Quote
Old 27th April 2014, 17:04   #26251  |  Link
QBhd
QB the Slayer
 
QBhd's Avatar
 
Join Date: Feb 2011
Location: Toronto
Posts: 452
Quote:
Originally Posted by James Freeman View Post
madshi, I think we're ready for a high quality Frame Interpolation algorithm now...

No I don't think so... Frame Interpolation is just bad VooDoo

QB
__________________
QBhd is offline   Reply With Quote
Old 27th April 2014, 17:22   #26252  |  Link
James Freeman
Registered User
 
Join Date: Sep 2013
Posts: 919
Quote:
Originally Posted by QBhd View Post
No I don't think so... Frame Interpolation is just bad VooDoo
Bad Frame Interpolation is bad.
HFR is the future.
__________________
System: i7 3770K, GTX660, Win7 64bit, Panasonic ST60, Dell U2410.
James Freeman is offline   Reply With Quote
Old 27th April 2014, 18:09   #26253  |  Link
aufkrawall
Registered User
 
Join Date: Dec 2011
Posts: 1,708
Personally, I think "proper" frame interpolation with madVR would be a dream.

Different topic:
Does anyone know if DXVA scaling with Maxwell is still bilinear + sharpening?
aufkrawall is offline   Reply With Quote
Old 27th April 2014, 19:20   #26254  |  Link
QBhd
QB the Slayer
 
QBhd's Avatar
 
Join Date: Feb 2011
Location: Toronto
Posts: 452
So you guys want to take up GPU resources to do something that is not really the job of a renderer. Interesting concept when we are all trying to squeeze every last drop of GPU performance for the best PQ image scaling.

QB
__________________
QBhd is offline   Reply With Quote
Old 27th April 2014, 19:29   #26255  |  Link
aufkrawall
Registered User
 
Join Date: Dec 2011
Posts: 1,708
What's the difference to current smooth motion?
aufkrawall is offline   Reply With Quote
Old 27th April 2014, 21:44   #26256  |  Link
Asmodian
Registered User
 
Join Date: Feb 2002
Location: San Jose, California
Posts: 3,597
Smooth Motion is a mathematically exact way to display a frame rate at a different refresh rate while frame interpolation is trying to guess what a middle frame might have looked like and cannot be done without artifacts (in the near future at least). There is no "correct" way to do the interpolation and lots of situations where the best course is simply not to attempt an interpolation. This is a very complex problem with no right answer, a much bigger development effort, and no way to reach near perfection so I can see why madshi would not implement or would wait to implement something.
Asmodian is offline   Reply With Quote
Old 27th April 2014, 22:10   #26257  |  Link
Ceremony
Registered User
 
Join Date: Jan 2014
Posts: 94
smooth motion blends frames occasionally, in case the video framerate isn't a multiple of the display refresh rate to avoid sudden jumps and skips.

frame interpolation (check out SVP if you wanna test it) analyses the difference between two frames and "guesses" the step inbetween (often taking multiple frames before and after into account). While this works remarkably well for stills and simple panning shots, anything fast moving (action scenes, etc.) and transforming will cause major artifacts unfortunately. Also, its pretty tough to calculate, so dont expect this to work properly on a laptop or something!

most smart TVs (imo smart means dumb, because they suck big time!) have such a feature built in with some ways to tune it (be it on/off or degree of interpolation).
Ceremony is offline   Reply With Quote
Old 27th April 2014, 23:07   #26258  |  Link
e-t172
Registered User
 
Join Date: Jan 2008
Posts: 568
There is also a fundamental difference in that the goal of Smooth Motion is to get closer to the way the content is supposed to look (native frame rate). It's much harder however to advocate motion interpolation because it tends to impose a "soap opera" look on everything, which arguably violates the artist's intent. Therefore it is a highly controversial feature and not at all an obvious "next step" or improvement. The real solution is to fix the source (i.e. HFR, high frame rate material) so that the artist can do whatever he wants, but that's still a long ways out it seems.
e-t172 is offline   Reply With Quote
Old 27th April 2014, 23:18   #26259  |  Link
Ver Greeneyes
Registered User
 
Join Date: May 2012
Posts: 445
Quote:
Originally Posted by e-t172 View Post
it tends to impose a "soap opera" look on everything,
Could you describe what this means? I don't tend to watch soap operas so I can't really picture it.
Ver Greeneyes is offline   Reply With Quote
Old 27th April 2014, 23:28   #26260  |  Link
StinDaWg
Registered User
 
Join Date: Jan 2014
Posts: 216
Quote:
Originally Posted by Ver Greeneyes View Post
Could you describe what this means? I don't tend to watch soap operas so I can't really picture it.
Everything looks unnaturally smooth and fake.
StinDaWg is offline   Reply With Quote
Reply

Tags
direct compute, dithering, error diffusion, madvr, ngu, nnedi3, quality, renderer, scaling, uhd upscaling, upsampling

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 09:10.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2019, vBulletin Solutions Inc.