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 7th May 2019, 19:57   #56181  |  Link
Asmodian
Registered User
 
Join Date: Feb 2002
Location: San Jose, California
Posts: 3,782
Quote:
Originally Posted by el Filou View Post
I don't get what the Kodi wiki means by 'scaling' and at what step.
That is a misunderstanding of how limited range YCbCr gets converted to full range RGB. They think one YCbCr -> RGB step gets turned into two steps, the conversion of YCbCr to RGB and a conversion of limited range to full range. In reality madVR simply uses a different matrix for YCbCr -> RGB to get full range RGB so the number of steps does not change and there is no extra scaling process.

Quote:
Originally Posted by Klaus1189 View Post
And of course that 50 fps is not worse in terms of smoothness just because there are 10 fps less.
Of course! At best you only get the smoothness of the source when using smooth motion. And no one is arguing that matching the refresh rate is not preferred, only that smooth motion is smooth even with 50p at 60 Hz!

Yes, matching refresh rates can be worth it. You do not need to say smooth motion is not smooth to say this, you can stick to explaining the real limitations of smooth motion.
__________________
madVR options explained

Last edited by Asmodian; 7th May 2019 at 20:01.
Asmodian is online now   Reply With Quote
Old 7th May 2019, 20:24   #56182  |  Link
nussman
Registered User
 
Join Date: Nov 2010
Posts: 215
50i/p with smooth motion @60hz doesnt work "perfect" for all kind of sources (i.e. live sport @50hz). Its still way better for "60hz only PC monitors" than without smoothmotion, but I prefer a matching refresh rate too if possible.

For 23/24p sources (film content) @60hz with smooth motion works very well for me.
nussman is offline   Reply With Quote
Old 7th May 2019, 20:27   #56183  |  Link
Klaus1189
Registered User
 
Join Date: Feb 2015
Location: Bavaria
Posts: 831
Quote:
Originally Posted by nussman View Post
50i/p with smooth motion @60hz doesnt work "perfect" for all kind of sources (i.e. live sport @50hz). Its still way better for "60hz only PC monitors" than without smoothmotion, but I prefer a matching refresh rate too if possible.
I fully agree. But there was, probably still is a feature in madvr to add other refresh rates for 60hz only monitors. I don‘t need it now so I am not sure.

Last edited by Klaus1189; 7th May 2019 at 20:29.
Klaus1189 is offline   Reply With Quote
Old 7th May 2019, 21:59   #56184  |  Link
Siso
Registered User
 
Siso's Avatar
 
Join Date: Sep 2013
Location: Bulgaria
Posts: 447
Speaking of refresh rates, I have an optimised 71.928 hz for 23.976 fps movies, do I need to make another one for example 72 hz for 24 fps movies?
Siso is offline   Reply With Quote
Old 7th May 2019, 22:06   #56185  |  Link
Klaus1189
Registered User
 
Join Date: Feb 2015
Location: Bavaria
Posts: 831
Yes, just google intel 24p bug, there were 23.976 files played at 24.000 so every 42 secs a repeated frame was shown. You need two different refresh rates for these two similar but different framerates.
Klaus1189 is offline   Reply With Quote
Old 7th May 2019, 22:20   #56186  |  Link
toki
Registered User
 
Join Date: Apr 2019
Posts: 60
so the chain equation has always confused me a bit. I have always set madVR to Limited, GPU full when calibrating. I guess this is wrong, even though it doesn't look natural and too dark?
toki is offline   Reply With Quote
Old 7th May 2019, 22:36   #56187  |  Link
Asmodian
Registered User
 
Join Date: Feb 2002
Location: San Jose, California
Posts: 3,782
That is not wrong if your TV is set to limited as well. Everything but madVR will be wrong but madVR will be correct.

This is actually the optimal configuration if your display needs limited range for some reason. It also matches the old standards all this stuff was based on so you still have the WtW and BtB information at the display, but given the calibration should clip it anyway this isn't actually beneficial.

Edit: WtW and BtB does make manual calibration easier, so it can be a benefit that way.
__________________
madVR options explained

Last edited by Asmodian; 7th May 2019 at 23:58.
Asmodian is online now   Reply With Quote
Old 8th May 2019, 09:37   #56188  |  Link
chros
Registered User
 
chros's Avatar
 
Join Date: Mar 2002
Posts: 1,567
Quote:
Originally Posted by toki View Post
so the chain equation has always confused me a bit. I have always set madVR to Limited, GPU full when calibrating. I guess this is wrong, even though it doesn't look natural and too dark?
Just to close this conversation
https://forum.doom9.org/showthread.p...16#post1271416
__________________
Ryzen 5 2600,Asus Prime b450-Plus,16GB,MSI GTX 1060 Gaming X 6GB(v385.28),Win10 LTSB 1607,MPC-BEx64+LAV+MadVR,Yamaha RX-A870,LG OLED65B8(2160p@23/24/25/29/30/50/59/60Hz)
chros is offline   Reply With Quote
Old 8th May 2019, 13:48   #56189  |  Link
chrisssj2
Registered User
 
Join Date: Apr 2019
Posts: 18
QuestioN I was playing Aquaman 4k HDR. And i pressed CTRL+J for the OSD.
And it said 1506 nits..
I have madvr@using pixel shaders target peaks 1000 nits.
Do I need to set it to 1506 or higher nits since that is being reported in Madvr OSD?
Also, do you recommend the option "present a frame for every Vsync" in madvr to be checked or not?
And lastly how can you know for sure if Vsync is being used in kodi dsplayer/madvr etc?

I had created a profile for kodi and MPC HC/BE with adaptive performance and Vsync On.
But i just want to check wether it is really on.

And Vsync does that mean a performance hit? Or generally really advised to use?

Also on the RGB vs YCBCR debate

https://www.avsforum.com/forum/465-h...-10-bpc-2.html

This guy says:
"YCbCr 422 10 bit/12 bit is the best you can go for HDR video sources at the moment I'm afraid."
Rather then RGB 12 bcp FULL range @23 hz..?
Still not sure what is best.

Someone on reddit nvidia"For movies you want 420, 12bit(they are all made in 420), for non HDR games rgb full, for HDR games 420 12bit or 422 10bit. 444 is for monitors/primary displays."

Is what I found on this reddit from an user. Is this true though? What settings is best for HDR Movies, which best for SDR movies and which best for SDR games and HDR games?

Option 1: 4K@60Hz 4:2:2 YCBR 12bpc Limited Range
Option 2: 4K@60Hz 4:4:4 YCBR 8bcp Limited Range
Option 3: 4K@60Hz 4:4:4 RGB 8bpc Full range
Option 4: 4k@23Hz 4:4:4 RGB 12bcp Full range

Option 1 probably best for HDR gaming according to reddit user
Option 2 probably best for SDR movie playback according to reddit user
Option 3 probably best for SDR gaming according to reddit user
Option 4 probably is best for HDR movie playback then according to this topic? Yet AVSforum seems to claim otherwise.

Jesus what a mess is this... One setting for each end use? Could someone clarify the situation?

Last edited by chrisssj2; 8th May 2019 at 14:18.
chrisssj2 is offline   Reply With Quote
Old 8th May 2019, 14:46   #56190  |  Link
huhn
Registered User
 
Join Date: Oct 2012
Posts: 6,160
Quote:
QuestioN I was playing Aquaman 4k HDR. And i pressed CTRL+J for the OSD.
And it said 1506 nits..
I have madvr@using pixel shaders target peaks 1000 nits.
Do I need to set it to 1506 or higher nits since that is being reported in Madvr OSD?
peak nits should be choicen on your display performance not the movie.
so no.

Quote:
I had created a profile for kodi and MPC HC/BE with adaptive performance and Vsync On.
But i just want to check wether it is really on.
do you have tearing no? then Vsync is working. this doesn't need checking as long as it works.

Quote:
And Vsync does that mean a performance hit? Or generally really advised to use?
yes it means a performance hit not sure if it is measurable.
yes Vsync should be who even gave you the idea that this may not be the case?

Quote:
Also, do you recommend the option "present a frame for every Vsync" in madvr to be checked or not?
if you have problems you can try this option and it was needed in the past things changed.
Quote:
Also on the RGB vs YCBCR debate
RGB end of story if your screen can do it.
and this is the case with games too as long as the game is still rendered in 10 bit that's different from madVR where rendering in 8 bit is not rarely superior to 10 bit.

i don't care about some random reddit user that make a list same for AVS.
huhn is offline   Reply With Quote
Old 8th May 2019, 15:28   #56191  |  Link
toki
Registered User
 
Join Date: Apr 2019
Posts: 60
the first thing is to turn off Aquaman. that movie is so terrible!
toki is offline   Reply With Quote
Old 8th May 2019, 17:15   #56192  |  Link
chrisssj2
Registered User
 
Join Date: Apr 2019
Posts: 18
Quote:
Originally Posted by huhn View Post
peak nits should be choicen on your display performance not the movie.
so no.


do you have tearing no? then Vsync is working. this doesn't need checking as long as it works.


yes it means a performance hit not sure if it is measurable.
yes Vsync should be who even gave you the idea that this may not be the case?


if you have problems you can try this option and it was needed in the past things changed.

RGB end of story if your screen can do it.
and this is the case with games too as long as the game is still rendered in 10 bit that's different from madVR where rendering in 8 bit is not rarely superior to 10 bit.

i don't care about some random reddit user that make a list same for AVS.
Ok so you saying so use RGB FULL 8 BIT@60 hz for SDR games and HDR games and SDR movies?

And then the RGB 12 BIT FULL @23 hz for HDR/MADVR movie playback?

Isn't it weird to use only 8 bits for HDR games?
chrisssj2 is offline   Reply With Quote
Old 8th May 2019, 20:07   #56193  |  Link
Asmodian
Registered User
 
Join Date: Feb 2002
Location: San Jose, California
Posts: 3,782
Quote:
Originally Posted by chrisssj2 View Post
Isn't it weird to use only 8 bits for HDR games?
No, they dither too. 8 bit is much better than losing 1/3 of the video data (YCbCr 422) or running 30 Hz. Can you tell if the game is running in 8 or 10 bit? Of all the things you can sacrifice the bit depth has the least impact.

Have a third party randomly change it and then try to determin what the bit depth is, do you do better than 50:50?
__________________
madVR options explained
Asmodian is online now   Reply With Quote
Old 8th May 2019, 23:48   #56194  |  Link
ryrynz
Registered User
 
ryrynz's Avatar
 
Join Date: Mar 2009
Posts: 3,277
@chrisssj2 you don't need to quote an entire reply consisting of answers to your questions when you post. If there are many branches of conversation it can be useful to quote but there's no real need when the topic is focussed, it just wastes space on the forum, cheers.
ryrynz is offline   Reply With Quote
Old 9th May 2019, 01:36   #56195  |  Link
tp4tissue
Registered User
 
tp4tissue's Avatar
 
Join Date: May 2013
Posts: 446
Quote:
Originally Posted by toki View Post
the first thing is to turn off Aquaman. that movie is so terrible!
It was a great movie, it had amber hurd.
__________________
Ghetto | 2500k 5Ghz
tp4tissue is offline   Reply With Quote
Old 9th May 2019, 15:29   #56196  |  Link
toki
Registered User
 
Join Date: Apr 2019
Posts: 60
Quote:
Originally Posted by tp4tissue View Post
It was a great movie, it had amber hurd.
I couldn't even finish it, it was that bad
toki is offline   Reply With Quote
Old 10th May 2019, 17:30   #56197  |  Link
SirMaster
Registered User
 
Join Date: Feb 2019
Posts: 51
Does anyone have much experience with the anamorphic lens option in madVR?

I am helping set up a new HTPC for a friend and we put in a RTX 2070. Anyways, with the settings we are using the render time is about 28-30ms which seems decent.

However, when I enable the anamorphic lens option, the render times go way up to 130-150ms!

Also, the anamorphic mode on his actual projector will stretch the content off the top and bottom edges of the screen (which is intentional). Like if it's a 16:9 native video and he wants to view it on his 2.40 screen, the 16:9 needs to be vertically stretched where the top and bottom get stretched off the screen. How can you do this with madVR anamorphic mode? I tried playing with the stretch numbers, but I can't get it to stretch the video vertically off the top and bottom of the screen. Maybe I have to use the "define visible screen area by cropping masked borders" feature?

I just assumed madVR would be really flexible with this sort of video manipulation, but I just can't seem to figure it out.

Anyone have any insights on either of these issues (the terrible anamorphic stretch performance) and (getting 16:9 content to vertically stretch)?

Last edited by SirMaster; 10th May 2019 at 17:35.
SirMaster is offline   Reply With Quote
Old 10th May 2019, 17:59   #56198  |  Link
huhn
Registered User
 
Join Date: Oct 2012
Posts: 6,160
a screen of the OSD should help but using a lens can increase render times heavily you have to do at least an extra scaling.

if you mean with stretching of the screen cropping you have to tell the player to do that that would be touch window from outside in MPC-HC no sure why someone would do something like that and removes big parts of movies but that'S how it is done.
huhn is offline   Reply With Quote
Old 10th May 2019, 20:36   #56199  |  Link
el Filou
Registered User
 
el Filou's Avatar
 
Join Date: Oct 2016
Posts: 585
I think leeperry just raised this issue a few posts earlier: when using 'touch window from outside', image gets scaled to bigger than screen size and then displayed with areas outside the screen, which wastes GPU, instead of being pre-cropped then scaled to full screen.

Also, when using NGU in combination with anamorphic scaling you can possibly end up using two different scaling steps depending on how you configured NGU settings.
For example I get this with broadcast SD when scaling to FHD:
luma x: 720 > [NGU] 1440 > [selected upscaling algorithm] 1920
luma y: 576 > [NGU] 1152 < [selected downscaling algorithm] 1080
Even if I choose the exact same algo for both upscaling and downscaling in NGU config, e.g. Lanczos 3 AR, the OSD still shows two different scaling steps for x and y.
__________________
HTPC: Windows 10 1909, MediaPortal 1, LAV Filters, ReClock, madVR. DVB-C TV, Panasonic GT60, 6.0 speakers Denon 2310, Core 2 Duo E7400, GeForce 1050 Ti
el Filou is offline   Reply With Quote
Old 10th May 2019, 20:42   #56200  |  Link
SirMaster
Registered User
 
Join Date: Feb 2019
Posts: 51
Quote:
Originally Posted by el Filou View Post
I think leeperry just raised this issue a few posts earlier: when using 'touch window from outside', image gets scaled to bigger than screen size and then displayed with areas outside the screen, which wastes GPU, instead of being pre-cropped then scaled to full screen.

Also, when using NGU in combination with anamorphic scaling you can possibly end up using two different scaling steps depending on how you configured NGU settings.
For example I get this with broadcast SD when scaling to FHD:
luma x: 720 > [NGU] 1440 > [selected upscaling algorithm] 1920
luma y: 576 > [NGU] 1152 < [selected downscaling algorithm] 1080
Even if I choose the exact same algo for both upscaling and downscaling in NGU config, e.g. Lanczos 3 AR, the OSD still shows two different scaling steps for x and y.
Yeah, I am getting 2 NGU steps, but I can't see how to avoid this.

Honestly I just want a simple scale for the stretch heh.

I can post my madVR settings and my OSD later tonight.
SirMaster 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 20:50.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, vBulletin Solutions Inc.