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. |
2nd January 2015, 20:59 | #27941 | Link |
Guest
Posts: n/a
|
I am very sorry if this has been re-posted a dozen times, but I have been out of 3DLUT-making scene for a while now. I just made a madVR 3DLUT with dispcalGUI 2.6.0.0, but HCFR does not detect madVR. I uninstalled madVR using the included .bat in normal mode and As Administrator. Then I restarted. Then I placed madVR on root directory of drive D to prevent any long directory names with spaces and ran the install .bat file in normal mode and As Administrator. However, none of that worked. When I select madVR as a pattern generator in the latest HCFR 3.1.6, I get an error saying that madVR DLL was not found... I have the latest madVR x.x.11. Maybe all these new connection options need to be used, but I do not see any options for them. I do see my local IP in madTPG - 192.168.1.96. When my Windows 8.1 firewall asked me about allowing madTPG access, I clicked "Allow" for both public and private networks. I also removed and re-installed HCFR Calibration several times in normal and As Administrator, but that didn't help... It keeps telling me DLL was not found and asking me whether madVR was installed...
Please don't yell at me for asking, but how do I fix this problem? I figured the problem must be with HCFR because dispcalGUI 2.6.0.0 detected madTPG just fine when Profiling (reading patterns) for 3DLUT creation. I applied 3DLUT and I can definitely see image differences, which means that 3DLUT is working, but I need to verify its accuracy with HCFR (or is there some other way?). I can't post on AVS forums because they banned me (yes SHAME on me, its embarrassing and I regret it and still feel guilty, please don't rub it n... ) and I tried 10x times to write them e-mails to get back, but it didn't work and they catch IP's if you try to create another account, etc. Please help! Last edited by XMonarchY; 2nd January 2015 at 21:12. |
4th January 2015, 22:22 | #27946 | Link | ||||||||||||||
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
Quote:
Quote:
Quote:
Not at this time. Quote:
Quote:
Quote:
madVR/madTPG is supposed to be backwards compatible to older calibration software versions. So it's my fault. Which is good, because I can fix it easily. Quote:
Quote:
Quote:
Quote:
Quote:
Quote:
Quote:
Quote:
To get an answer to the real question, read my comment here: http://www.avsforum.com/forum/24-dig...l#post28854034 |
||||||||||||||
4th January 2015, 22:27 | #27947 | Link |
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
madVR v0.87.12 released
http://madshi.net/madVR.zip Code:
* fixed: v0.87.11 broke madTPG API compatability * madTPG: added dropdown box to allow manual gamut/3dlut selection * madTPG: loading a 3dlut now takes immediate effect in madTPG * madTPG: added API "GetVersion" * madTPG: added API "Get/SetSelected3dlut" * madTPG: number of pre-presented frames is now limited to max 3 |
5th January 2015, 10:57 | #27949 | Link |
Registered User
Join Date: Jan 2014
Posts: 216
|
I've had an issue for awhile now where I will play a video in FSE mode and the render time will settle down to say, 30ms for 720p->1080p NNEDI3 upscaling. Then I'll back out to windowed mode and back into FSE and the render time will drop down to a steady 25ms. I'm not sure what's happening but I can reproduce this consistently. I have my gpu overclocked and it's almost like madVR isn't accepting the overclock in FSE even though Afterburner says it's using 1050 core the entire time. No change with v0.87.12. I'm using an AMD 7850 and it happens on all firmware versions.
Does this happen to anyone else? |
5th January 2015, 21:27 | #27954 | Link |
Registered User
Join Date: Jul 2009
Posts: 51
|
Hi Madshi, the problem seemed to start with 0.87.11 - it was not present in 0.87.10.
The problem is when I am using a second display (e.g. my Samsung UE40H6400 TV, via HDMI, and an AMD HD7790 GPU). I can force film mode on the primary display, but not when the MPC-HC window is on the secondary display. When I toggle through modes, it just goes through auto and video, and skips film. |
6th January 2015, 03:49 | #27956 | Link |
Registered User
Join Date: Oct 2012
Posts: 7,925
|
DXVA and CUVID decoding (it's the same any way) doesn't have an effect on this.
these settings effect madVR when DXVA scaling is used. next thing is deinterlacing. madVR uses DXVA deinterlacing with a interlaced input stream. and again there is no difference between CUVID or DXVA deinterlacing (CUVID HQ is deinterlacing with inverse telecine checked in the driver). you can change the type of deinterlacing at least with AMD but the sharpening settings and so on doesn't have an effect on this (just tested with amd). there is "nothing" to gain from CUVID any more it only forces your GPU in a high powerstate. and with active deinterlacing you can't use madVR IVTC for telecine streams any more. which can be easily trigger with a hot key. |
6th January 2015, 15:51 | #27957 | Link |
Registered User
Join Date: Jan 2007
Posts: 729
|
http://www.pcper.com/news/Displays/C...esync-Monitors
FreeSync monitors finally out (almost). It appears that at least some will hit the shelves this month? |
6th January 2015, 21:56 | #27958 | Link |
Registered User
Join Date: Feb 2002
Location: San Jose, California
Posts: 4,407
|
That is very good news but not relevant for madVR as it is currently. It would be interesting to have madVR be G-sync (or FreeSync) aware, it seems like it would be possible to have the audio and video sync be very exact without smooth motion, basically like Reclock but adjusting the video frame timing instead of re-sampling the audio. This would also be great for bit-streaming without smooth motion.
I have been playing with G-sync (PG278Q) and madVR. If madVR goes into full screen exclusive mode G-sync activates but the image flickers quite often which does not happen in games. Also, as expected, it does not refresh at the move frame rate but at 143.5 Hz (I have "use fastest mode available" enabled). If I don't have g-sync enabled or don't use FSE madVR reports a refresh rate of 144 Hz. I believe madVR would need to not present frames in advance but instead only present new a frame when it should be displayed. As it is now I cannot use full screen exclusive with G-sync enabled but this is not a problem. |
7th January 2015, 00:10 | #27959 | Link |
Registered User
Join Date: Mar 2007
Posts: 934
|
Without looking it up, I'm assuming G-Sync and Freesync have fluid refresh rates where you can update the screen at any time (with a minimum time between updates of course, which'd translate to a maximum refresh rate). If so, that is pretty damned cool and I'm surprised it's taken so long to happen. I came up with the idea about a decade ago.
In terms of video playback, it's probably most useful for mixed telecined/interlaced content in 60 Hz regions.
__________________
TV Setup: LG OLED55B7V; Onkyo TX-NR515; ODroid N2+; CoreElec 9.2.7 |
7th January 2015, 01:38 | #27960 | Link | |
Registered User
Join Date: Jan 2008
Posts: 589
|
This article, which unfortunately is in French, contains the following information, which I'm translating:
Quote:
|
|
Tags |
direct compute, dithering, error diffusion, madvr, ngu, nnedi3, quality, renderer, scaling, uhd upscaling, upsampling |
Thread Tools | Search this Thread |
Display Modes | |
|
|