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 June 2015, 03:47   #31341  |  Link
huhn
Registered User
 
Join Date: Oct 2012
Posts: 6,062
Quote:
Originally Posted by aufkrawall View Post
I suppose the bug that the tap menu für Jinc disappears is known?
I can't select the number of taps anymore for image upscaling with Jinc, it's grayed out. Tried resetting settings, didn't help.
4 and 8 tabs are removed. they are bad scalers there is no need for them.
huhn is offline   Reply With Quote
Old 26th June 2015, 03:49   #31342  |  Link
luk008
Registered User
 
Join Date: Aug 2011
Posts: 38
Quote:
Originally Posted by aufkrawall View Post
I suppose the bug that the tap menu für Jinc disappears is known?
I can't select the number of taps anymore for image upscaling with Jinc, it's grayed out. Tried resetting settings, didn't help.
It's not a bug, madshi has removed Jinc 4 from the options because there's any significative advantage over Jinc 3 since both sharpness and aliasing increase in the same proportions.
luk008 is offline   Reply With Quote
Old 26th June 2015, 04:25   #31343  |  Link
RyuzakiL
Registered User
 
Join Date: May 2015
Posts: 36
Quote:
Originally Posted by James Freeman View Post
Most TV's are built around Limited range and their CMS and processing works best in limited range if not exclusively, like yours TV.
My Panasonic ST60 Plasma behaves best (no banding) only in Limited range when 4:4:4 is enabled.

I would recommend sending Limited Range to your TV.
GPU Limited, madVR Full, TV limited (Good all around but two [invisible] conversion stages).
Or GPU Full, madVR Limited, TV Limited (No stage conversions, best for Video but clipped PC range).

The question is whether your TV does 4:4:4 only in PC mode or in TV mode also (if at all)?
I would aim my priority for Smooth (no banding) 4:4:4 first, then on less range conversions.



Exactly!

I'm not advocating for limited range when using TV, but in my experience limited range gives more accurate picture with my TV and without unnecessary conversion stages if GPU+madVR set accordingly.
The Panasonic ST/VT 50/60 only like limited range with 4:4:4 like you already noticed.

If a TV behaves well in full range like a PC monitor, then by all means, use GPU Full, madVR Full, TV Full.
arrggh...stuck between two configs

I now understand that my Samsung 40inch LedTV can do both FULL and LIMITED by way of toggling "HDMI Black Level" to NORMAL (0-255) and LOW (16-235) and also my panel can do FULL and LIMITED range RGB 4:4:4

so now the question is what are the advantages of using FULL, FULL, FULL config compared to FULL, LIMITED, LIMITED config?

it seems my panel is a very flexible one and can do either both config, so i'm confused what config to use.

P.S. just want to put a nail on this topic and move on.
RyuzakiL is offline   Reply With Quote
Old 26th June 2015, 07:59   #31344  |  Link
RainyDog
Registered User
 
Join Date: May 2009
Posts: 176
Quote:
Originally Posted by RyuzakiL View Post
arrggh...stuck between two configs

I now understand that my Samsung 40inch LedTV can do both FULL and LIMITED by way of toggling "HDMI Black Level" to NORMAL (0-255) and LOW (16-235) and also my panel can do FULL and LIMITED range RGB 4:4:4

so now the question is what are the advantages of using FULL, FULL, FULL config compared to FULL, LIMITED, LIMITED config?

it seems my panel is a very flexible one and can do either both config, so i'm confused what config to use.

P.S. just want to put a nail on this topic and move on.
If you've confirmed that your panel can definitely do Full RGB 4:4:4 then use that.

What James says about his plasma doesn't apply to your LCD. Many LCD telly's are perfectly capable of replicating a PC monitor and will do full RGB 4:4:4 fine.

I calibrated my Sony telly running AVS patterns in madVR outputting full RGB 4:4:4 using an EyeOne Pro with Colour HCFR and the results were nigh on reference.
RainyDog is offline   Reply With Quote
Old 26th June 2015, 08:09   #31345  |  Link
RyuzakiL
Registered User
 
Join Date: May 2015
Posts: 36
Ok thanks, then my next mission is too calibrate this panel but it seems i'll be needing some high-end expensive equipment just for that. hm.. I think I'll just have to settle with default settings of my Samsung Telly (Default Movie Profile)
RyuzakiL is offline   Reply With Quote
Old 26th June 2015, 09:31   #31346  |  Link
James Freeman
Registered User
 
Join Date: Sep 2013
Posts: 919
To test your TV for 4:4:4 and Banding,

Banding Test: Smooth Grey Gradient
Banding Test 2Smooth Color Gradient
4:4:4 Test: Chroma 4:4:4 Image

You should see 4:4:4 and no banding at all.
If your TV/Projector/Monitor can display 4:4:4 Full range without banding, it would be ideal because madVR's dithering is top quality.
__________________
System: i7 3770K, GTX660, Win7 64bit, Panasonic ST60, Dell U2410.

Last edited by James Freeman; 26th June 2015 at 10:41.
James Freeman is offline   Reply With Quote
Old 26th June 2015, 12:14   #31347  |  Link
lanzorg
Registered User
 
Join Date: Aug 2011
Posts: 43
Quote:
Originally Posted by huhn View Post
Finesharp is the current topic.
FineSharp is really too bad, madvr needs implementation of LimitedSharpenFasterMC and UnsharpHQ from the avisynth world.

Avisynth script chain:

Code:
SetMTMode(3,0)
ffdshow_source()
SetMTmode(2,0)
LimitedSharpenFasterMC(1)
UnsharpHQ(THRESHOLD=80,SHARPSTR=0.1,SMOOTH=0.2)
GradFun2db()
SetMTMode(1)
GetMTMode(false) > 0 ? distributor() : last
LimitedSharpenFasterMC avisynth function:

Code:
function LimitedSharpenFasterMC(clip clp,int "strength")
{
ox=clp.width
oy=clp.height
ss_x=1.5
ss_y=1.5
wide=true
softdec=-1
soft=softdec!=-1 ? softdec : sqrt((((ss_x+ss_y)/2.0-1.0)*100.0))*10
soft=soft>100 ? 100 : soft 
ssx=round(float(ox*ss_x/8)*8)
ssy=round(float(oy*ss_y/8)*8)
clp.isYV12() ? clp : clp.converttoyv12()
ss_x !=1.0 || ss_y !=1.0 ? last.spline144resize(ssx,ssy):last
tmp=last
dark_limit1=tmp.mt_inpand()
bright_limit1=tmp.mt_expand()
dark_limit=(wide==false) ? dark_limit1 : dark_limit1.removegrain(20,-1).mt_inpand()
bright_limit=(wide==false) ? bright_limit1 : bright_limit1.removegrain(20,-1).mt_expand()
minmaxavg=mt_average(dark_limit1,bright_limit1)
Str=string(float(strength)/100.0)
normsharp=mt_lutxy(tmp,tmp.removegrain(11,-1),"x y == x x x y - abs 16 / 1 2 / ^ 16 * "+Str+
\ " * x y - 2 ^  x y - 2 ^ "+Str+" 100 * 25 / + / * x y - x y - abs / * + ?")
last=mt_clamp(normsharp,bright_limit,dark_limit,1,1)
clp.isYV12() ? tmp.mergeluma(last) : tmp.mergeluma(last.converttoyuy2())
return last
}
lanzorg is offline   Reply With Quote
Old 26th June 2015, 14:58   #31348  |  Link
aufkrawall
Registered User
 
Join Date: Dec 2011
Posts: 1,717
Quote:
Originally Posted by huhn View Post
4 and 8 tabs are removed. they are bad scalers there is no need for them.
Thanks. It always displayed Jinc3 for me until I changed settings, a little confusing.
aufkrawall is offline   Reply With Quote
Old 26th June 2015, 19:29   #31349  |  Link
XMonarchY
Registered User
 
Join Date: Jan 2014
Posts: 489
How can I compare madVR renderer image with all them HQ settings I use to the "original image", whatever that means. In fact, what IS the original image? Which madVR settings should I use to get that image? Set everything to bicubic?
__________________
8700K @ 5Ghz | ASUS Z370 Hero X | Corsair 16GB @ 3200Mhz | RTX 2080 Ti @ 2100Mhz | Samsung 970 NVMe 250GB | WD Black 2TB | Corsair AX 850W | LG 32GK850G-B @ 165Hz | Xonar DGX | Windows 10 LTSC 1809
XMonarchY is offline   Reply With Quote
Old 26th June 2015, 19:34   #31350  |  Link
Asmodian
Registered User
 
Join Date: Feb 2002
Location: San Jose, California
Posts: 3,773
Quote:
Originally Posted by XMonarchY View Post
How can I compare madVR renderer image with all them HQ settings I use to the "original image", whatever that means. In fact, what IS the original image? Which madVR settings should I use to get that image? Set everything to bicubic?
There is no "original" image anymore.

If you mean to the image you used to get with EVR or something you simply compare to the other renderer. Don't try to configure madVR to give an image similar to EVR.
__________________
madVR options explained
Asmodian is offline   Reply With Quote
Old 26th June 2015, 19:54   #31351  |  Link
XMonarchY
Registered User
 
Join Date: Jan 2014
Posts: 489
OK, gotcha.

I have another problem. I don't know which info I need to report, but at times when I watch a video without any issues at all and then pause & resume it, the video begins to stutter badly until I shut down MPC-HC completely and restart the video...
__________________
8700K @ 5Ghz | ASUS Z370 Hero X | Corsair 16GB @ 3200Mhz | RTX 2080 Ti @ 2100Mhz | Samsung 970 NVMe 250GB | WD Black 2TB | Corsair AX 850W | LG 32GK850G-B @ 165Hz | Xonar DGX | Windows 10 LTSC 1809
XMonarchY is offline   Reply With Quote
Old 26th June 2015, 20:56   #31352  |  Link
omarank
Registered User
 
Join Date: Nov 2011
Posts: 180
Quote:
Originally Posted by madshi View Post
Thanks for the feedback. I don't like NEDI for chroma upscaling, either. FWIW, I've found a little bug in super-xbr chroma upscaling. Please compare again. Maybe super-xbr could replace Jinc as the "best medium speed" chroma upscaler?
I don’t think super-xbr can replace Jinc as the best medium speed chroma upscaler, because Jinc is more natural looking. I prefer suber-xbr chroma (sharpness 100) over Jinc AR chroma in two instances, when the source is rather soft or the image is being downscaled by a large factor using CR AR LL.


For image doubling, super-xbr is good, but NNEDI3 is much superior. In the screenshots below, NNEDI3 clearly looks the best on this sample:
Jinc3 AR
NNEDI3 64 neurons
Super-xbr sharpness 100
Original

Last edited by omarank; 27th June 2015 at 03:19.
omarank is offline   Reply With Quote
Old 26th June 2015, 23:34   #31353  |  Link
Warner306
Registered User
 
Join Date: Dec 2014
Posts: 1,127
Is there a place for debanding when outputting at 10-bits with dithering? I find it harder to detect banding with the sources I use.
Warner306 is offline   Reply With Quote
Old 26th June 2015, 23:36   #31354  |  Link
Warner306
Registered User
 
Join Date: Dec 2014
Posts: 1,127
Also, can someone tell me if this chroma subsampling chain is correct?

HTPC Chroma Subsampling: (Source) Y'CbCr 4:2:0 -> (madVR) Y'CbCr 4:4:4 to RGB -> (TV) Y'CbCr 4:2:2 or Y'CbCr 4:4:4 to RGB
Warner306 is offline   Reply With Quote
Old 26th June 2015, 23:38   #31355  |  Link
huhn
Registered User
 
Join Date: Oct 2012
Posts: 6,062
Quote:
Originally Posted by Warner306 View Post
Is there a place for debanding when outputting at 10-bits with dithering? I find it harder to detect banding with the sources I use.
the madVR dithering isn't removing banding it simply doesn't add more that's it so the usage of bedanding doesn't change with 10 bit output it just depends on the source quality. and even 10 bit sources in the future will have banding.

Quote:
HTPC Chroma Subsampling: (Source) Y'CbCr 4:2:0 -> (madVR) Y'CbCr 4:4:4 to RGB -> (TV) Y'CbCr 4:2:2 or Y'CbCr 4:4:4 to RGB
not sure what you mean but.

there are a lot of possibilities.

and it is really hard to judge what is really happening in a TV. but you can assume the final signal given the pixel is based on full range RGB signal i guess...
a GPU can output 4:2:0, 4:2:2, 4:4:4, limited range RGB and full range RGB (and that with 8-12 bit to make things more complicated). the madVR output is RGB with all possible ranges thanks to the custom levels but madVR is kind of telling the GPU his signal is full range RGB always.

Last edited by huhn; 26th June 2015 at 23:45.
huhn is offline   Reply With Quote
Old 27th June 2015, 00:13   #31356  |  Link
Warner306
Registered User
 
Join Date: Dec 2014
Posts: 1,127
Quote:
Originally Posted by huhn View Post
the madVR dithering isn't removing banding it simply doesn't add more that's it so the usage of bedanding doesn't change with 10 bit output it just depends on the source quality. and even 10 bit sources in the future will have banding.



not sure what you mean but.

there are a lot of possibilities.

and it is really hard to judge what is really happening in a TV. but you can assume the final signal given the pixel is based on full range RGB signal i guess...
a GPU can output 4:2:0, 4:2:2, 4:4:4, limited range RGB and full range RGB (and that with 8-12 bit to make things more complicated). the madVR output is RGB with all possible ranges thanks to the custom levels but madVR is kind of telling the GPU his signal is full range RGB always.
Does the TV always convert back to YCbCr before conversion to RGB, or is it possible for the TV to display RGB from a PC without conversion?
Warner306 is offline   Reply With Quote
Old 27th June 2015, 00:27   #31357  |  Link
QBhd
QB the Slayer
 
QBhd's Avatar
 
Join Date: Feb 2011
Location: Toronto
Posts: 492
As stated, it's very hard to know exactly what the TV is doing internally... We can only guess

QB
__________________
QBhd is offline   Reply With Quote
Old 27th June 2015, 00:33   #31358  |  Link
Warner306
Registered User
 
Join Date: Dec 2014
Posts: 1,127
Quote:
Originally Posted by James Freeman View Post
To test your TV for 4:4:4 and Banding,

Banding Test: Smooth Grey Gradient
Banding Test 2Smooth Color Gradient
4:4:4 Test: Chroma 4:4:4 Image

You should see 4:4:4 and no banding at all.
If your TV/Projector/Monitor can display 4:4:4 Full range without banding, it would be ideal because madVR's dithering is top quality.
I see 4:4:4 in purple lettering with no blinking. Is this an assumed pass?
Warner306 is offline   Reply With Quote
Old 27th June 2015, 00:44   #31359  |  Link
huhn
Registered User
 
Join Date: Oct 2012
Posts: 6,062
Quote:
Originally Posted by Warner306 View Post
I see 4:4:4 in purple lettering with no blinking. Is this an assumed pass?
and if you didn't see 4:2:2 (sharping can show the 4:4:4) than you can at least say it isn't down sample.

i personally look at the pixel to judge 4.4:4.
huhn is offline   Reply With Quote
Old 27th June 2015, 01:28   #31360  |  Link
JarrettH
Registered User
 
Join Date: Aug 2004
Posts: 842
Once this bug in xbr is fixed, you believe it to be a better chroma upsampler than jinc, madshi?
JarrettH 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 04:11.


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