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. Domains: forum.doom9.org / forum.doom9.net / forum.doom9.se |
|
|||||||
![]() |
|
|
Thread Tools | Search this Thread | Display Modes |
|
|
#2801 | Link |
|
Registered User
Join Date: Oct 2012
Posts: 8,623
|
this should stop all none security updates for now:
http://abload.de/img/update8mxgw.png and there is a file/option from microsoft to stop "third party" updates. and there is a software to block some updates here a german page with it: http://www.chip.de/downloads/Microso..._81466169.html I personally use 353.62 so I don't use any of this. and here a work around from the beta times: http://answers.microsoft.com/en-us/i...llReplies#tabs |
|
|
|
|
|
#2802 | Link |
|
Registered User
Join Date: Dec 2011
Posts: 1,812
|
Problem is that not the device is hidden from WU, but only the specific driver. That means if a new driver is released on WU, it will automatically be downloaded and installed if you don't run the tool from MS to hide drivers timely.
The setting in the advanced system options to not download driver updates from WU doesn't really have an effect, it downloaded drivers with it disabled for many users (me included). The only safe way is to manually disable Windows Update service and manually start it again and directly after this start the update hide tool from MS before searching for updates. But this can't be reommended to most users, since they will forget to enable the service again and miss out security updates. |
|
|
|
|
|
#2808 | Link |
|
MPDN Extensions Dev
Join Date: Apr 2015
Posts: 52
|
I went for this tool first : http://www.guru3d.com/files-details/...-download.html
To remove all old drivers and everything that could be problematic. It also disable for you the automatic update of drivers. Then when I restarted and I installed the version of the driver I wanted. I've done the the same for the Intel drivers (I was having some crash with it). |
|
|
|
|
|
#2810 | Link |
|
Registered User
Join Date: Dec 2011
Posts: 1,812
|
This time, MPDN crashed with this error while being in the render script settings:
Well, however playback so far is totally fine when doing nothing else. There's some stutter during playback when changing volume via mouse scroll, but that's not a huge issue to me. Btw: Jinc2D with few taps can be well used together with SuperRes for scaling below factor 1.5x, image isn't doubled and so a lot of GPU resources are saved, compared to the other available algorithms. Scaling BD content to WQHD looks great this way. I lowered SuperChromaRes strength a bit (to 1.7), since it brightens colors (mostly red) up a bit (or more than a bit). MPDN Jinc with 4 taps w.o. AR looks good and isn't too expensive together with SuperChromaRes. However, Jinc3 AR of madVR is still better for this purpose, imho. Last edited by aufkrawall; 31st July 2015 at 18:03. |
|
|
|
|
|
#2811 | Link | |
|
Suptitle, MediaPlayer.NET
Join Date: Nov 2001
Posts: 1,721
|
Quote:
Startup time with extensions had also been improved and on my system it is now less than half a second. We will continue to look into startup speed improvements (the bulk of the time is spent in creating and initialising the playlist extension). |
|
|
|
|
|
|
#2812 | Link |
|
Suptitle, MediaPlayer.NET
Join Date: Nov 2001
Posts: 1,721
|
Quote:
MPDN's Jinc2D is similar to AviSynth's - in fact, you'll find the code to reference AviSynth's Jinc implementation too. I'm not sure what madVR's Jinc is but I think it is similar (i.e. both EWA) to MPDN's Jinc2D. Perhaps madVR uses different weights, but only madshi can answer that (MPDN's Jinc2D is open source, madVR's Jinc is closed source). |
|
|
|
|
|
#2813 | Link |
|
Registered User
Join Date: Dec 2011
Posts: 1,812
|
MPDN's Jinc2D would probably be fine for chroma, but yet it can only be used for luma (and only for upscaling).
That other Jinc of MPDN (that comes with the player itself, not with the render scripts) is muuuch softer. My config isn't anything special, I think crashes are rather related to specific actions in a row than to a specific config. Maybe enable some kind of debug mode? Then I can provide logs. Last edited by aufkrawall; 1st August 2015 at 16:27. |
|
|
|
|
|
#2814 | Link |
|
Registered User
Join Date: Dec 2011
Posts: 1,812
|
2.39 hasn't yet crashed even once while changing render script settings.
Deband seems noticeably improved: In the example I posted earlier, higher power now reduces the banding better. And even if set to 1, it doesn't seem to be a huge detail killer. However, I really dislike the grain. It is easy to perceive but doesn't seem to have a real positive effect. Thankfully it can be turned off. In some areas, Shiandow's algorithm is better than madshi's and vice versa. For SuperRes, smoothness is still in the GUI (but doesn't seem to have an effect). I found it very useless anyway, introducing lots of aliasing (shouldn't it do the exact opposite?). As for chroma, I'm now with custom chroma scaler (B: 1; C: 0.2) + SuperChromaRes (1 pass, strength 1.3). Hard to notice a difference to madVR Jinc3 AR without a comparison with some cartoon examples. Jinc2D with configurable AR filter, like already existing for luma, would probably be better though. |
|
|
|
|
|
#2815 | Link |
|
Registered User
Join Date: Dec 2011
Posts: 1,812
|
Ok, here's an example (filmed skies) where madVR deband still looks clearly better:
http://www41.zippyshare.com/v/hqA15LOW/file.html |
|
|
|
|
|
#2818 | Link |
|
Registered User
Join Date: Jun 2005
Posts: 513
|
Just noticed that MPDN's FSE mode actually performs worse than desktop mode, due to spikes of high rendering time which causes in-fluidity during playback of videos. Desktop mode plays well though.
Windows 8.1 64-bit, nVidia 880M Optimus system, with latest version of MPDN, render script SuperRes used. |
|
|
|
|
|
#2819 | Link | |
|
Registered User
Join Date: Mar 2009
Posts: 3,697
|
Quote:
|
|
|
|
|
|
|
#2820 | Link |
|
Registered User
Join Date: Jun 2015
Posts: 10
|
Thanks for such an awesome player. I love it!
I came from Potplayer where I could jump forward and backward instantaneously (but inaccurately). ![]() I had a look to see if I could modify one of your extensions, Navigation.cs. Since it is trying to calculate where the next frame is, it's quite slow. StepFrame() is the only quick one but it pauses the playback and it's only forwarding by 1 single frame). Going back 1 frame on the other hand, Jumpframe (-1), is as slow as the Jump(x) method. The next or previous keyframe isn't exposed in the methods of the PlayerControl class, unfortunately. I was wondering if you would consider adding a feature like that (it's also implemented in all the major media players, such as VLC, kmplayer, MPC-HC, etc) that allows users to instantly jump to keyframes so that we don't need to wait. The intervals between keyframes are usually 5-10 seconds long, really nice to have when precision isn't required (we can still have the ability to jump single frames), which is most of the time. Last edited by kingpage; 2nd August 2015 at 08:33. |
|
|
|
![]() |
| Tags |
| direct3d, mpdn, nnedi3, opencl, reclock |
| Thread Tools | Search this Thread |
| Display Modes | |
|
|