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. |
20th May 2017, 17:24 | #43762 | Link |
Registered User
Join Date: Oct 2012
Posts: 7,923
|
NGU very high
in FSE ~32 AVG and ~35 ms max. but i'm not a huge friend of rendertimes i get rendertiems of 39 ms with super xbr because the cards goes for a nap. with advanced rendertimes NGU takes ~29 ms. it is hard to read because it is called finalsteps there right before smoothmotion. GPU usages is abut 90% and the AVG boost clock is 1700 i use other filter at the same time. in short NGU very high 32 ms NGU takes about 29 ms GPU load 90 % boost clock 1700 NGU high 16-17 ms NGU take about 11 ms (a ~3rd like it should be) GPU load is 45 % boost clock is 1900 (don't think about this) |
20th May 2017, 20:14 | #43764 | Link |
Registered User
Join Date: Feb 2002
Location: San Jose, California
Posts: 4,407
|
Interesting, it does seem like NGU has improved on Polaris.
My Pascal Titan X doesn't do that much better at NGU Sharp very high 1080p -> 2160p. These are the rendering times for the NGU step by itself (ShowRenderSteps). If you disable subtitles and smooth motion you can see it reported by itself, the times weren't quite the same when it was mixed into another line. I used chroma doubling set to normal (Bicubic 60). madVR v0.91.9 of course, DX11 fullscreen windowed, Win10 w/ creators update, Nvidia driver version 382.05. I added image enhancements and changed chroma upscaling to keep the GPU clocks up. The GPU was at 1860 MHz unless otherwise noted. NGU Sharp (low): 1.7 ms NGU Sharp (med): 2.6 ms NGU Sharp (high): 5.2 ms NGU Sharp (very high): 16.8 ms edit: I remember NNEDI3 differences being interesting too: NNEDI3 32: 4.18 ms NNEDI3 64: 7.54 ms NNEDI3 128: 18.83 ms NNEDI3 256: 32.93 ms (1835 MHz)
__________________
madVR options explained Last edited by Asmodian; 20th May 2017 at 20:51. Reason: added NNEDI3 and NGU med, low |
20th May 2017, 20:34 | #43765 | Link | ||||||||||||||
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
Quote:
Quote:
Quote:
I can understand if you want to save power, and have your GPU fan be silent. So I can understand that you want low chip/memory frequency *during normal video playback*. However, frequencies being temporarily higher after changing the window size or after seeking is totally irrelevant, IMHO. I don't see why that would matter in any way. Quote:
Quote:
Quote:
Quote:
Quote:
http://www.avsforum.com/forum/139-di...argyllcms.html Quote:
Quote:
Quote:
Quote:
Are you using fullscreen exclusive mode? Quote:
It shouldn't. It certainly doesn't for me. Quote:
|
||||||||||||||
20th May 2017, 21:11 | #43766 | Link |
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
madVR v0.91.10 released
http://madshi.net/madVR.zip Code:
* improved Windows 10 HDR metadata support * added new option "hack Direct3D to make 24.000Hz and 60.000Hz work" * performance improvement with "trust DXVA" + "use D3D11 for presentation" * fixed: Intel GPU driver flickering bug with "use D3D11 for presentation" * fixed: madVR(64).ax files were larger than necessary * Intel: 10bit DXVA surfaces are now converted to 10bit RGB (instead of 8bit) 2) There was a long standing bug with Intel GPUs where activating the "use D3D11 for presentation" option could result in video frames jumping back and forth. I was finally able to reproduce the bug on my new Kaby Lake test PC. It turned out to be a GPU driver bug (as usual). I've found a workaround, so the problem should hopefully be fixed now. 3) I've had a hack in madVR which hooked deep into D3D9 and D3D11 to make sure that 24.000Hz and 60.000Hz display modes were correctly used in Windows 8.1 and Windows 10. However, I've found that this hack can sometimes lead to presentation queues not filling properly. So this hack is now optional (and disabled by default). Without the hack, you will probably always get 23.976Hz and 59.940Hz, even if you want 24.000Hz and 60.000Hz. With the hack, you will probably get proper refresh rates, but may run into issues with the presentation queue not filling. Pick your poison... |
20th May 2017, 23:33 | #43768 | Link | |
Registered User
Join Date: Jul 2016
Posts: 171
|
Quote:
Does this mean we dont have to manually switch the windows hdr slider anymore ? if so, it didnt work on my system. Last edited by imhh11; 20th May 2017 at 23:37. |
|
20th May 2017, 23:41 | #43769 | Link | |
Registered User
Join Date: Jul 2011
Posts: 33
|
Quote:
I think, it could be due to whiter-than-white values or something like that in the original video. But all that leads to unpleasant artifacts in the video. Is it possible to still convert these values for them to not be out of sRGB gamut? Last edited by igvk; 21st May 2017 at 11:57. |
|
20th May 2017, 23:45 | #43770 | Link | |
Registered User
Join Date: Dec 2007
Posts: 128
|
Quote:
|
|
21st May 2017, 00:07 | #43771 | Link |
Registered User
Join Date: Feb 2002
Location: San Jose, California
Posts: 4,407
|
If you use the rendering -> general settings -> "enable windowed overlay (Windows 7 and newer)" option, and don't have exclusive mode or DX11 enabled, WtW values are clipped to in gamut.
__________________
madVR options explained |
21st May 2017, 00:34 | #43772 | Link | |
Registered User
Join Date: Jul 2011
Posts: 33
|
Quote:
Although I would prefer using exclusive mode with D3D11. Is it really a difficult problem to solve ? It seems to me that many computer-generated videos contain such colors. |
|
21st May 2017, 01:39 | #43773 | Link |
Registered User
Join Date: Feb 2002
Location: San Jose, California
Posts: 4,407
|
Are you using a 3DLUT with an absolute or relative intent? Some of the intent options clip colors that are outside the gamut which can cause artifacts like that.
__________________
madVR options explained |
21st May 2017, 02:52 | #43774 | Link |
Registered User
Join Date: Aug 2004
Location: Canada
Posts: 860
|
Pretty major bug since 0.91.9 or 0.91.10 (I'm on 0.91.8): I can't exit out of full screen (d3d9, automatic full screen exclusive) by double clicking nor does alt+enter work. Both produce a black screen - no desktop - audio continues. I have to alt+F4 the program to close it. ctrl+alt+del doesn't overlay for me to access task manager. In d3d11 exclusive (automatic full screen exclusive), switching between full screen and windowed now takes 4-5 seconds.
550 Ti Windows 10 Home (not on Creators Update yet) Latest stable MPC-HC 378.78 drivers Dell U2417H over HDMI Last edited by JarrettH; 21st May 2017 at 14:36. |
21st May 2017, 02:58 | #43775 | Link |
Registered User
Join Date: Feb 2002
Location: San Jose, California
Posts: 4,407
|
It works fine here, both D3D9 and D3D11 in and out of full screen exclusive almost instantly.
Titan X (Pascal) Windows 10 Pro Zoom Player 382.05 drivers
__________________
madVR options explained |
21st May 2017, 04:11 | #43777 | Link |
Registered User
Join Date: Jan 2005
Posts: 171
|
With madVR v.0.91.10 I now get the following HDR behaviors on my GTX 1060 with Nvidia driver version 382.19
In windowed mode, with HDR on in Windows settings, and using the Windows 10 HDR metadata transmission API I get HDR output that seems correct. In FSE mode, I get the same incorrect output I was getting prior to version 0.91.10. If I turn automatic FSE off and try to go fullscreen, I almost immediately get a black screen followed within about a minute by a reboot. I can consistently reproduce this and it happens with the windowed overlay on or off. |
21st May 2017, 04:17 | #43778 | Link | |
Registered User
Join Date: Oct 2012
Posts: 7,923
|
Quote:
just to give you an idea what the real problem is a r9 270 is not much slower when NGU is used compared to a rx 480. and there is no reason for that! but it looks like NGU got a lot faster on polaris so maybe it get's fixed someday. edit: only the nnedi3 numbers: 32: 10 ms 64: 19 ms 128: 37ms 256: 72 ms Last edited by huhn; 21st May 2017 at 06:11. |
|
21st May 2017, 07:17 | #43779 | Link |
Registered User
Join Date: Mar 2009
Posts: 3,650
|
Benchmarking has come up a few times, I was thinking a new thread specifically testing some hand picked samples using a preconfigured config might be an idea.
Perhaps madshi could help here adding a simple benchmark mode specifically to display and or maybe log the minimum, average and maximum render times, dropped frames etc of whatever video is played showing an average at the end on a blank screen after playback until the player is closed, since this basically exists already in the OSD and logging functions I'm guessing this in basic form would be pretty easy to add. I think though this was more a 1.0 feature perhaps fleshed out a bit more, but some basic logging/display and some specifically chosen samples would provide some apple to apple comparisons going forward. |
21st May 2017, 07:23 | #43780 | Link |
Registered User
Join Date: Feb 2014
Posts: 139
|
The intel fix has also resolved intermittent frame flickering with low queues (such as when starting playback/seeking without waiting for queues to fill) that I experienced with a laptop running dual intel/AMD gpu setup. Well done!
|
Tags |
direct compute, dithering, error diffusion, madvr, ngu, nnedi3, quality, renderer, scaling, uhd upscaling, upsampling |
Thread Tools | Search this Thread |
Display Modes | |
|
|