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 13th February 2017, 21:30   #42441  |  Link
pirlouy
_
 
Join Date: May 2008
Location: France
Posts: 671
You don't get it with your screenshots.
The problem is not what you send, but what the TV does AFTER you send signal. That's what huhn is trying to say.

From rtings review, 24Hz is not judder-free for your TV. But maybe its motion interpolation is able to give a good result.

Last edited by pirlouy; 13th February 2017 at 21:32.
pirlouy is offline   Reply With Quote
Old 13th February 2017, 21:52   #42442  |  Link
Damien147
Registered User
 
Join Date: Mar 2011
Posts: 313
The last two pictures state the final result and what the tv receives supposedly.
I don't have rtings highly in my opinion but I think I agree with the judder.
Damien147 is offline   Reply With Quote
Old 13th February 2017, 21:53   #42443  |  Link
tranfa
Registered User
 
Join Date: Oct 2006
Posts: 68
Quote:
Originally Posted by tranfa View Post
Yes, exactly: I'm curious in understanding the logic behind.

Will post a screen as soon as I can.

Thank you!
25hz (when I have in Display modes only 4096x2160p24, 4096x2160p25):




Last edited by tranfa; 13th February 2017 at 21:55.
tranfa is offline   Reply With Quote
Old 13th February 2017, 21:58   #42444  |  Link
tranfa
Registered User
 
Join Date: Oct 2006
Posts: 68
30hz (when in Display Modes I add 4096x2160p30 to the previous):


Last edited by tranfa; 13th February 2017 at 22:18.
tranfa is offline   Reply With Quote
Old 13th February 2017, 22:14   #42445  |  Link
Damien147
Registered User
 
Join Date: Mar 2011
Posts: 313
https://www.youtube.com/watch?v=p85xwZ_OLX0

4096x2160p29 and not 4096x2160p30?The same?
It can be done as it says.I guess you have to play with display mode settings.Something similar happened to me yesterday and I'm trying to reproduce it again.

Last edited by Damien147; 13th February 2017 at 22:18.
Damien147 is offline   Reply With Quote
Old 13th February 2017, 22:20   #42446  |  Link
tranfa
Registered User
 
Join Date: Oct 2006
Posts: 68
Well, here at least it's stable...

PS: situation unchanged with p29 and/or p30.

Last edited by tranfa; 13th February 2017 at 22:35.
tranfa is offline   Reply With Quote
Old 13th February 2017, 22:45   #42447  |  Link
Damien147
Registered User
 
Join Date: Mar 2011
Posts: 313
Found it,if I put:

4096x2160p23, 4096x2160p24, 4096x2160p25, 4096x2160p29, 4096x2160p30

24p fails and goes to 60hz

If I put:

4096x2160p24, 4096x2160p25, 4096x2160p29, 4096x2160p30

everything's fine.
Damien147 is offline   Reply With Quote
Old 14th February 2017, 08:02   #42448  |  Link
huhn
Registered User
 
Join Date: Oct 2012
Posts: 7,357
Quote:
Originally Posted by tranfa View Post
30hz (when in Display Modes I add 4096x2160p30 to the previous):

http://i.imgur.com/T0S6bff.jpg

doesn't make sense to me.
it doesn't even use p30 in the screen it is 29p...

and please use links/url to add images. this is my personal recommendation nothing else.
huhn is offline   Reply With Quote
Old 14th February 2017, 12:11   #42449  |  Link
tranfa
Registered User
 
Join Date: Oct 2006
Posts: 68
Quote:
Originally Posted by huhn View Post
doesn't make sense to me.
it doesn't even use p30 in the screen it is 29p...

and please use links/url to add images. this is my personal recommendation nothing else.
Still this is the case, both with p30 and p29 in the Display Modes... and it is what puzzles me.

Ok for the url, I'll make sure of using this next time.
tranfa is offline   Reply With Quote
Old 14th February 2017, 16:23   #42450  |  Link
clsid
*****
 
Join Date: Feb 2005
Posts: 5,462
Just use 24/50/59
clsid is offline   Reply With Quote
Old 14th February 2017, 17:30   #42451  |  Link
Damien147
Registered User
 
Join Date: Mar 2011
Posts: 313
Quote:
Originally Posted by tranfa View Post
Well, here at least it's stable...

PS: situation unchanged with p29 and/or p30.
Guess what...!!The same thing happened to me with a series that I have.30p with 25fps.Same settings as I stated earlier.But every other 25 fps material I have plays in 25hz.So to conclude for now there's something in the source that triggers it.Can I stupidly ask if you tried another 25fps fps movie or whatever(?)

@clsid gave a nice idea for the end of the day.

Last edited by Damien147; 14th February 2017 at 18:50.
Damien147 is offline   Reply With Quote
Old 14th February 2017, 19:18   #42452  |  Link
Damien147
Registered User
 
Join Date: Mar 2011
Posts: 313
UPDATE:

issue fixed,just added 4096x2160p50 and it plays the particular 25fps source in 50hz.

@tranfa
try :

4096x2160p24, 4096x2160p25, 4096x2160p29, 4096x2160p30, 4096x2160p50

Last edited by Damien147; 14th February 2017 at 19:25.
Damien147 is offline   Reply With Quote
Old 14th February 2017, 20:36   #42453  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by Bloodred View Post
Hello! I've been using madVR for a long time, though I never posted in this forum. I've taken a few comparison shots comparing NGU/NGU pixart/NNEDI3, since I actually had a look at this thread after noticing that NGU really wasn't doing too well with a certain 720x480 anime source compared to NNEDI3, upscaled to 3840x2160.
Thanks, this image is quite helpful. Do you have some more images like this, which show NGU stumbling? If so, that might help improving NGU (non-pixart). I only need the unscaled image.

Quote:
Originally Posted by The 8472 View Post
Did something change wrt. to taking screenshots with subtitles from xy-subfilter?
Not that I know of right now. But screenshot functionality is going to be revisited at some point in the future.

Quote:
Originally Posted by TheShadowRunner View Post
Yes, the same happens in graphstudionext.
After much testing I was able to find the breaking point:
build 0.90.4.0.
Any build prior to that works OK, any build after (and including) 0.90.4.0 triggers the freeze when seeking during DVD playback.
Hope it helps pinpoint what change broke it.

Edit: after checking the changelog, my bet would be on either of these 2 changes to be the culprit..
* fixed: OSD stopped working when no new frames came from decoder
* last video frame is now remembered for 2 seconds when stopping graph
Interesting. Does this happen with every DVD, or just with some? How can I reproduce this problem exactly (step-by-step guide)?

Quote:
Originally Posted by StinDaWg View Post
madshi, I am still getting constant black screen lockups when existing/closing a video, that require a reboot of the computer. I have tried every combination I can think of. D3D9, D3D11, FSE and windowed, DontRenderAfterStop folder with and without.
It even happens if you disable fullscreen exclusive mode!?!? That seems very weird. Which OS and GPU are you using? If it's NVidia or Intel, have you tried Overlay mode? Otherwise you could also try reducing the number of pre-presented frames ("present in advance"), or turning pre-present off completely.

Have you tried different GPU driver versions? This is most likely a GPU driver bug.

Quote:
Originally Posted by jkauff View Post
@madshi: I've watched a lot more SD movies using the test build, and I've found that NGUpixart High is the sweet spot for my setup. It's as good as NNEDI3 on all but the poorest quality sources. NGU Very High is too much, though, and gives me that "processed" look I mentioned before. In all cases, I'm superscaling, then downscaling to 1920x1080. This is using a GTX1060.

For the poor sources, though, NNEDI3 does a better job. Maybe a few more tweaks are in order, but I'm almost ready to kiss NNEDI3 goodbye.
Can you post some (unscaled) test images where you like NNEDI3 better than NGU pixart?

Quote:
Originally Posted by Ceremony View Post
I have a strange new issue since I installed the most recent windows updates (win10 x64):
Whenever I switch to fullscreen mode (be it exclusive or not), the player suddenly freezes and starts consuming loads and loads of memory (RAM, not VRAM), till either the OS freezes up due to OOM.

I am running Win10 Pro x64 with 8GB of RAM (2GB of it as dedicated VRAM) with the latest madVR version and driver for the AMD A10-7850K APU.
It seems to occur for some win10 AMD users, it's a driver bug. The usual "solution" is to reduce the number of pre-presented frames to a lower number. I think 6 frames usually worked, or maybe 4, I don't recall right now.

Quote:
Originally Posted by MS-DOS View Post
I'm against it. SXBR + SR3 still looks better than any NGU for me overall (more universal source wise) *and* is faster than anything above "Low".
Are you talking standard NGU or the test build with the new "NGU pixart" algorithm? In the latter case, have you compared SXBR+SR with NGUpixart+SR or to straight NGUpixart without SR? I don't see how SXBR+SR could be faster than NGU medium without SR.

Can you post a test frame (BMP or PNG) of the unscaled source, for which you prefer SXBR+SR over NGU (pixart)?

Quote:
Originally Posted by age View Post
Is it possibile to add bilinear scaling done by shaders ?not the same by gpu texture unit,this last one should work with 8 bit precision only
GPU texture unit supports any bitdepth, so doing bilinear *up*scaling via shaders has no purpose. Bilinear *down*scaling via shaders produces better results for scaling factors bigger than 2.0x, though. But I don't think I'm doing to add an option for that because I don't expect Bilinear downscaling via shaders to be noticeably faster than e.g. Mitchell (without AR) downscaling. And Mitchell should look better than Bilinear.

Quote:
Originally Posted by kitame View Post
speaking of pixart, when is the next official update for madVR? since it seems like pixart is already ready to go.
I don't know, still trying different things.

Quote:
Originally Posted by kitame View Post
also, can we have this part of the image scaler be manually defined like this?
and what would the result be if we mixed up different scaling algorithms in luma/chroma image scaling?
The whole scaling settings will be revisited soon, don't know the details yet.

Quote:
Originally Posted by tranfa View Post
In MadVR settings I try to input the Display modes fit for my videos (some at 24fps, some other at 25, a very few at 30).
As long as I input just "4096x2160p25, 4096x2160p24" I don't have problems: according to the video, it adapts the refresh rate of my tv (a Panasonic 4K led). Only thing, as one might expect, stuttering for the few videos at 30fps.
If I input as well the 30, with the following logic: "4096x2160p30, 4096x2160p25, 4096x2160p24", the referesh rate goes to 30 also for the 25fps videos....and I don't understand why.
The video card is a Radeon 7700, supporting through HDMI the 4k res at 24hz, at 25 and at 30 too.
Some suggestions:

1) Please use display modes in the physical resolution of your TV, if possible, to avoid extra processing steps (which might harm image quality) in your TV. If your TV is really 4096x2160, then that's the right mode to use. But if your TV is actually 3840x2160, then please configure your HTPC to output 3840x2160 instead.

2) Make extra sure that all the refresh rates you enter into the madVR modes list are actually available, otherwise strange things may happen.

3) If things still react weird, you can upload a debug log and I'll see what's going on.

Quote:
Originally Posted by ryrynz View Post
You might need another variation of NGU to deal with this sort of content, I think slight alterations might not be enough as NGU and NGU pix look similar to me, and I can't imagine tweaks will be enough..
Here's an image I'll put forward for your NGU pix variant, easy enough for me to find more which work well with NNEDI3 if you want anything.
Can you mark areas in this image where you prefer NNEDI3? With your Anime image, I could see the difference, with this image I'm struggling a bit.

In all your tests, are you comparing just NGU pixart "very high"? Or have you also tested the lower speed variants? They don't seem to produce the same misinterpretations as "very high" sometimes does.
madshi is offline   Reply With Quote
Old 14th February 2017, 21:18   #42454  |  Link
TheShadowRunner
Registered User
 
TheShadowRunner's Avatar
 
Join Date: Feb 2004
Posts: 399
Quote:
Originally Posted by madshi View Post
Interesting. Does this happen with every DVD, or just with some? How can I reproduce this problem exactly (step-by-step guide)?
Yes it happens with all DVDs I could try (10+).
To reproduce: just start any dvd playback, on the menu (if there's one) select the main movie or some extra. And then just seek during playback or jump to a random position : the player instantly becomes unresponsive (interestingly the bug doesn't happen when going from one chapter to another).
__________________
XP SP3 / Geforce 8500 / Zoom Player
TheShadowRunner is offline   Reply With Quote
Old 14th February 2017, 23:19   #42455  |  Link
Damien147
Registered User
 
Join Date: Mar 2011
Posts: 313
Just have to ask even if it's stupid.

My TV{(Samsung Ku6000 (or else ku6300)}is 60hz and says in the manual that it supports film mode only for 1080i in HDMI.
In windows display adapter properties supported screen refresh rates are:24,25,29,30,50,59,60 hz.
I played with display modes in Madvr and it seems that 24p,25p,30p,50p is activated and displayed in 4096x2160(Madvr and TV report it).
If ''bypasses'' is the proper word do you guys think that what I am doing might cause problem/damage to the display?
I mean I haven't found support for that officially.

----------

If everything's ok,24p(24hz) in 23.976 fps gives me 1 frame repeat every 42 seconds and no dropped frames in OSD.Does this look good?Is there room for improvement?23hz is not an option and no reclock because I'm 64bit.

Last edited by Damien147; 14th February 2017 at 23:22.
Damien147 is offline   Reply With Quote
Old 14th February 2017, 23:42   #42456  |  Link
Shiandow
Registered User
 
Join Date: Dec 2013
Posts: 753
Quote:
Originally Posted by madshi View Post
GPU texture unit supports any bitdepth, so doing bilinear *up*scaling via shaders has no purpose. Bilinear *down*scaling via shaders produces better results for scaling factors bigger than 2.0x, though. But I don't think I'm doing to add an option for that because I don't expect Bilinear downscaling via shaders to be noticeably faster than e.g. Mitchell (without AR) downscaling. And Mitchell should look better than Bilinear.
Not that it's particularly important, but proper bilinear downscaling (i.e. a box filter) gives better results than bilinear interpolation for pretty much any scaling factor other than 2 (when they're equivalent).
Shiandow is offline   Reply With Quote
Old 15th February 2017, 00:07   #42457  |  Link
ryrynz
Registered User
 
ryrynz's Avatar
 
Join Date: Mar 2009
Posts: 3,527
Quote:
Originally Posted by madshi View Post
Can you mark areas in this image where you prefer NNEDI3? With your Anime image, I could see the difference, with this image I'm struggling a bit.

In all your tests, are you comparing just NGU pixart "very high"? Or have you also tested the lower speed variants? They don't seem to produce the same misinterpretations as "very high" sometimes does.
I chose this image because the entire image is that much softer, there's no area in particular that NGU pix really fails here it just doesn't quite pop like NNEDI3 does. It's a good indicator of how sharp I believe this new variant would have to be.

There is one area where it has been a bit aggressive and I've marked it here. There's a soft smudge that NGU pix removes that NNEDI3 keeps and also the curve on the button gets pulled back to the left. I use NGU pix very high for comparison. High definitely looks better in the marked area, it's less aggressive so I prefer it for that reason but it's sharpness is quite low. Anyway you're right the video I linked you before is a much better test of low res animated content.. only NGU very high does okay with the lines (everything else loses to NNEDI3 32 neurons) but suffers from the line issues I've mentioned previously.

Last edited by ryrynz; 15th February 2017 at 00:34.
ryrynz is offline   Reply With Quote
Old 15th February 2017, 00:39   #42458  |  Link
Bloodred
Registered User
 
Join Date: Feb 2017
Posts: 11
Quote:
Originally Posted by madshi View Post
Thanks, this image is quite helpful. Do you have some more images like this, which show NGU stumbling? If so, that might help improving NGU (non-pixart). I only need the unscaled image.
I've tried to find a few more scenes, though I think the one I first posted is one of the best examples, as in the image itself isn't of very poor quality and also showcases the issue. I've zipped a few images, just the unscaled versions:
NGU problematic scenes
Out of those images, the Slayers frame is by far the worst, but it's also very low quality, so it's to be expected.

One thing I've noticed is that the smudged/wavy look is most apparent when doing NGU image quadrupling, it's not much of an issue otherwise, so this particular usecase I ran into (upscaling 480p > 2160p) is what makes it apparent. Some of those shots aren't all too bad, but if you switch back and forth between NNEDI3 and NGU quadrupling it's easily visible. It's also most apparent around areas with a lot of black outlines, but it does happen in colored areas too.

I've used NGU vhigh across the board when looking for these scenes, with Jinc AR to upscale all the way to 2160p and deband set to Low.

Let me know if you could use even more images, I don't have a lot of low-res sources, but I'll try to look for more scenes where the issues are visible if you need more stuff to test with.
Bloodred is offline   Reply With Quote
Old 15th February 2017, 01:35   #42459  |  Link
jkauff
Registered User
 
Join Date: Oct 2012
Location: Akron, OH
Posts: 486
Quote:
Originally Posted by madshi View Post
Can you post some (unscaled) test images where you like NNEDI3 better than NGU pixart?
I take it all back. I took several screenshots from the worst SD sources I could find, and NGU pixart High did a better job on every set (using NNEDI3 128 neurons and 256 neurons both).

I'm sold.

Last edited by jkauff; 15th February 2017 at 01:38.
jkauff is offline   Reply With Quote
Old 15th February 2017, 02:23   #42460  |  Link
MS-DOS
Registered User
 
Join Date: Sep 2012
Posts: 77
Quote:
Originally Posted by madshi View Post
Are you talking standard NGU or the test build with the new "NGU pixart" algorithm? In the latter case, have you compared SXBR+SR with NGUpixart+SR or to straight NGUpixart without SR? I don't see how SXBR+SR could be faster than NGU medium without SR.
Both. SXBR+SR1 (mistake, sorry) against straight NGU. I didn't test NGU with SR, because apart from you not recommending it it's also very slow. Here's what I just got:

SXBR+SR = 4.92ms
NGU Med = 5.41ms

Nevermind it though, I probably shouldn't mention it in the first place as that's an issue with the AMD driver + RX 480 and not with NGU.
Quote:
Originally Posted by madshi View Post
Can you post a test frame (BMP or PNG) of the unscaled source, for which you prefer SXBR+SR over NGU (pixart)?
I currently don't have a good video frame example to post, so here's a couple of images: 1 2. Double the first one and quadruple the second. With pixart, the pinned flag (or whatever it's called) on the first one looks aliased, while with the second one the whole image turns into an artificial mess with aliased edges.
With real life content I often need to scale bad quality sources with compression artifacts. Guess what NGU does? It enhances (sharpens) the quality of the artifacts while leaving the actual data visually almost untouched compared to SXBR+SR. Sometimes it produces aliasing, sometimes it just looks too artificial to me, and sometimes it looks better than SXBR+SR. Therefore it's not a universal upscaler that I would use for everything.
MS-DOS 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 00:00.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, vBulletin Solutions Inc.