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 5th June 2013, 15:26   #19021  |  Link
callannn
Registered User
 
Join Date: Apr 2013
Posts: 6
Quote:
Originally Posted by Soukyuu View Post
pens, it could be that the decoder queue doesn't fill up fast enough to not drop frames vs. having enough time if you watch the whole episode. I solved nearly all my buffering problems with LAV by just increasing the cpu queue to 64. I'm saying "problems with LAV" because on my test files I get no frame drops with haali, but I do with LAV. So it might be the same for you, I already notified nevcariel about it.
Yes I use Lav, not haali. Would you recommend I increase my CPU queue to a higher number than default? Same with the GPU queue?

Quote:
Originally Posted by madshi View Post
The OSD screenshot only helps if it's directly from the situation when the frame drops occur. The additional info we need is for example which queues are (near) empty when those frame drops occur.
Here is a screenshot taken the exact moment of a delayed frame. I don't get any dropped frames during playback at all (the dropped frames in this screenshot are from initally booting the video up). As you can see none of the queues seem to drop when this occurs, so i'm stumped.
callannn is offline   Reply With Quote
Old 5th June 2013, 15:29   #19022  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Delayed frames can occur during windowed playback. Of course ideally they shouldn't, but they sometimes do. If you want to avoid this, try using fullscreen exclusive mode.
madshi is offline   Reply With Quote
Old 5th June 2013, 15:38   #19023  |  Link
callannn
Registered User
 
Join Date: Apr 2013
Posts: 6
Quote:
Originally Posted by madshi View Post
Delayed frames can occur during windowed playback. Of course ideally they shouldn't, but they sometimes do. If you want to avoid this, try using fullscreen exclusive mode.
Ah so it's not necessarily a problem with playback, my GPU or anything, just something that happens? Is there any way to get rid of them at all?

I posted a problem I seem to have with FEM not so long ago but it seemed to have been overlooked, in that when i switch to exclusive mode my average rendering time seems to go through the roof. From my understanding when in exclusive mode the video works directly from the GPU, so you should receive much better playback, so I have no idea why it seems worse when i switch to it. I have left all exclusive mode settings as default in the madVR settings, could this be why?

I apologise for all the questions as well!
callannn is offline   Reply With Quote
Old 5th June 2013, 16:51   #19024  |  Link
kopija
Registered User
 
Join Date: May 2012
Posts: 49
Hi guys,
a pair of questions:
I am sending video to my 720p plasma and am wondering whether I should set its resolution to 50 or 60hz when using smooth motion FRC?
What are advantages of using "separate device for presentation" on Intel HD3000?ž
Thanks for replies, and godspeed our video emperor Madshi!
kopija is offline   Reply With Quote
Old 5th June 2013, 17:17   #19025  |  Link
leeperry
Kid for Today
 
Join Date: Aug 2004
Posts: 3,477
Quote:
Originally Posted by kopija View Post
I am sending video to my 720p plasma and am wondering whether I should set its resolution to 50 or 60hz when using smooth motion FRC?
madshi said that the higher the better as FRC simulates an infinite refresh rate. So if I were you I'd try to reach as high as you can get, some displays can reach up to 72Hz(and 120Hz if 3D capable) FWIW...even a bit over 60Hz wouldn't hurt

Last edited by leeperry; 5th June 2013 at 17:30.
leeperry is offline   Reply With Quote
Old 5th June 2013, 19:47   #19026  |  Link
n3w813
Registered User
 
Join Date: Jan 2006
Posts: 80
Quote:
Originally Posted by cyberbeing View Post
First, a head-up once again that the name of new subtitle interface filter is XySubFilter.

XySubFilter = New Subtitle Interface (XySubFilter.dll)
xy-VSFilter = Legacy Filter (VSFilter.dll)
* fixed: #46: xy-vsfilter: 3DLUT was not applied to frames with subtitles

So does the above issue only affect XySubFilter.dll and NOT VSFilter.dll? (If I'm not updating to 0.86.2)
n3w813 is offline   Reply With Quote
Old 5th June 2013, 20:25   #19027  |  Link
raul31
Registered User
 
Join Date: Dec 2012
Posts: 5
madvr

Quote:
Originally Posted by Asmodian View Post
I'll give some answers to keep madshi free to keep up the good work. These have been covered many times in this thread and I don't think madshi needs to keep answering.

(hardware acceleration):
It sounds like you want "none" to me. This is what I use as well, all the other options use the GPU. The "hardware" doing the acceleration is on the GPU.

RGB Output Levels:
Untouched, leave converting color spaces to MadVR.

Dithering Mode:
Random unless you plan to re-encode the output. Random is visually higher quality but it does not encode as well.

You can turn on Yadif in LAV for software deinterlacing, I use video mode for real interlaced video (30i in my case).

All you need to do in LAV to keep the GPU free is to use "none" for hardware acceleration and don't use any shader scripts in MPC-HC.
thanks a lot asmodian!
though i would have loved to hear madshi's thoughts as well as i do see him replying to new users of madvr... no matter how simple the question. a bit disappointed.

i understand what u r saying regarding LAV filters. thank u.
btw, what is "shader scripts in MPC-HC?" & how do i avoid them?

now that u have explained to me how i can get the LAV filters powered by the CPU only, how do i get madVR working with the Nvidia Gpu exclusively?
because i mean like how do i know that madvr is instead being powered by the intel integrated graphics?

thanks a lot!
raul31 is offline   Reply With Quote
Old 5th June 2013, 21:21   #19028  |  Link
mindbomb
Registered User
 
Join Date: Aug 2010
Posts: 576
question about the second smooth motion option- "or if the display refresh rate is an exact multiple of the movie refresh rate"
what is the difference between this and the always on option?
cause doesn't the second option cover every possibility?
mindbomb is offline   Reply With Quote
Old 5th June 2013, 21:22   #19029  |  Link
kasper93
MPC-HC Developer
 
Join Date: May 2010
Location: Poland
Posts: 586
Quote:
Originally Posted by n3w813 View Post
So does the above issue only affect XySubFilter.dll and NOT VSFilter.dll? (If I'm not updating to 0.86.2)
VSFilter.dll draw subtitles onto video frame before it reaches madVR. madVR doesn't even know that subtitles were drown. And sure 3DLUT is applied to whole frame (with subs). With XySubFilter.dll it works quite differently, so yes only XySubFilter.dll was affected.

I was testing XySubFilter.dll and I really like it especially for PGS subtitles. http://screenshotcomparison.com/comparison/27166/ (#3 are HD subtitles that's why quality is similar, but we can see that ISR render at wrong levels it seems)

I have issue with blinking video frame when subtitle appear and DXVA deinterlacing is enabled. Reported it already on trac. I wonder why no one spots this bug before.
kasper93 is offline   Reply With Quote
Old 5th June 2013, 21:31   #19030  |  Link
Focalom
Registered User
 
Join Date: Aug 2012
Posts: 2
I got a problem after updating these today:

-Ati drivers
-MPC-HC
-Lav filters
-madVR

I guess all of the versions were from 6 months to 12 months old.

The problem is that mpc-hc won't close properly, the process i still running, preventing running new video files. I tried to search, and some people points at madVR. When I use an another renderer (EVR custom pres), the problem dissappears. Any idea?

Last edited by Focalom; 5th June 2013 at 21:34.
Focalom is offline   Reply With Quote
Old 5th June 2013, 21:43   #19031  |  Link
Danat
Registered User
 
Join Date: May 2013
Posts: 10
Quote:
Originally Posted by mindbomb View Post
question about the second smooth motion option- "or if the display refresh rate is an exact multiple of the movie refresh rate"
what is the difference between this and the always on option?
cause doesn't the second option cover every possibility?
EDIT: my answer was wrong. sorry for confusion. madshi replied to your question here

Last edited by Danat; 6th June 2013 at 12:21.
Danat is offline   Reply With Quote
Old 5th June 2013, 22:13   #19032  |  Link
kiccolsd
Registered User
 
Join Date: Mar 2010
Posts: 14
Quote:
Originally Posted by madshi View Post
I'm not sure I understand what you mean. Do you think that your playback chain somehow adds ringing artifacts to the movie which aren't in the original movie source? Can you show me a screenshot of the exact artifact you mean?
At first thx for your help and you madvr.
I tried to take a screenshot but if I stop the movie seem that artifacts disappear. It's something related to motion in the frames.

I'll try to explain, imagine a panning quite fast scene, around the edge\border of a character I can see artifacts, something that similar to ringing artifacts.

Another typical scene for this issue is a panning frame with a wire mesh, the border of the mesh shows artifacts.

By the way I'm a real noob (with HQ love) and maybe is not a madvr fault but if you have some tips i'll be thankfull!!!
kiccolsd is offline   Reply With Quote
Old 5th June 2013, 22:28   #19033  |  Link
truexfan81
Registered User
 
truexfan81's Avatar
 
Join Date: Nov 2012
Posts: 138
Quote:
Originally Posted by Focalom View Post
I got a problem after updating these today:

-Ati drivers
-MPC-HC
-Lav filters
-madVR

I guess all of the versions were from 6 months to 12 months old.

The problem is that mpc-hc won't close properly, the process i still running, preventing running new video files. I tried to search, and some people points at madVR. When I use an another renderer (EVR custom pres), the problem dissappears. Any idea?
i used to have that issue it is a mpc-hc bug, it has been fixed in the nightlies

http://nightly.mpc-hc.org/
truexfan81 is offline   Reply With Quote
Old 5th June 2013, 23:07   #19034  |  Link
cyberbeing
Broadband Junkie
 
Join Date: Oct 2005
Posts: 1,859
Quote:
Originally Posted by n3w813 View Post
So does the above issue only affect XySubFilter.dll and NOT VSFilter.dll? (If I'm not updating to 0.86.2)
Yes, this was a bug in prior version of madVR with XySubFilter.dll (not yet released). This is why I asked madshi to update the changelog, since these bugs which were fixed have nothing to do with xy-VSFilter (VSFilter.dll). It looks as though the changelog in the first post of the thread wasn't updated yet.

Quote:
Originally Posted by kasper93 View Post
I wonder why no one spots this bug before.
We are aware of this issue with black boxes and/or flashing between lines with bitmap subtitles (PGS/DVB/VOBSUB). I can occasionally reproduce it with progressive content as well. At this point in time we believe this is caused by a performance issue in XySubFilter which results in sending a bogus bitmap to madVR, and something on our to-do list before we move on to user testing and a beta release. Only bugs which we believe were caused by madVR were added to madshi's bug tracker, all bugs we believe are caused by XySubFilter specifically we are tracking privately internally.

Quote:
Originally Posted by madshi View Post
So I'm asking the yuvMatrix again for every rendered frame. That's a very simple "tell me your yuvMatrix, please" question which the madVR rendering thread asks XySubFilter. And this question seems to get stuck inside of the XySubFilter code for a very long time.
Maybe were missing a bit of logging around this area, since I not seeing anything that madVR is requesting yuvMatrix for every frame. All I see is that XySubFilter updates the yuvMatrix once immediately after the resolution rectangle change, and immediately after start sending new bitmaps and positions to madVR. I do agree that XySubFilter shouldn't be getting stuck on a yuvMatrix request if that's really what's occurring, I also believe that requesting yuvMatrix for every frame is a bit overkill especially if you wanted to mimic xy-VSFilter which at this point in time doesn't support changing yuvMatrix mid-stream at all, except video user override in the settings GUI. Maybe this should be changed so its the responsibility of the subtitle provider to inform the subtitle consumer of the yuvMatrix only when something may have changed. It seems possible considering our logging, that this is what XySubFilter is already doing, but I don't really know.


Quote:
Originally Posted by madshi View Post
Actually, there is an indication (but no proof) that madVR is not refusing anything, but that the render queue is getting stuck because XySubFilter gets stuck when madVR asks it about yuvMatrix.
Could you clarify this once more.

Theoretically if a subtitle provider only updates yuvMatrix once at the beginning of the stream and ignores all further yuvMatrix requests from the subtitle consumer, but continues to respond to RequestFrame & GetBitmap requests nominally, what is the intended behavior by madVR? Assuming our logs are accurate, XySubFilter does not seem to be stuck internally, since immediately after the resolution change and updating yuvMatrix, we start responding to all RequestFrame & GetBitmap requests.

What requirements need to be met in madVR internally before it renders subtitles from the subtitle queue?


Quote:
Originally Posted by madshi View Post
We don't need to discuss it, but if my guess is right, fixing this in XySubFilter should be rather easy.
Quote:
Originally Posted by madshi View Post
P.S: @cyberbeing, I've double checked the rendering code. It's probably either ISubRenderProvider::GetString("yuvMatrix") or ISubRenderFrame::GetOutputRect() which stalls inside of XySubFilter.
Okay. After I confirm with the developer what's going on I'll email you. He has not yet tested madVR 0.86.2 with this behavior change yet, so there could be an oversight somewhere. Overall, GetOutputRect sounds a bit more likely as a cause of stalling rather than yuvMatrix.

Last edited by cyberbeing; 6th June 2013 at 01:24.
cyberbeing is offline   Reply With Quote
Old 5th June 2013, 23:37   #19035  |  Link
kasper93
MPC-HC Developer
 
Join Date: May 2010
Location: Poland
Posts: 586
Quote:
Originally Posted by cyberbeing View Post
Only bugs which we believe were caused by madVR were added to madshi's bug tracker, all XySubFilter bugs we are tracking privately internally.
Oh, I see. So I should not post it on bug tracker. But I did it already so if you like you can comment and clarify that this is NOT madVR bug so madshi can close the ticket.

By the way this public repo is the most recent?

Last edited by kasper93; 5th June 2013 at 23:42.
kasper93 is offline   Reply With Quote
Old 6th June 2013, 00:47   #19036  |  Link
cyberbeing
Broadband Junkie
 
Join Date: Oct 2005
Posts: 1,859
Quote:
Originally Posted by kasper93 View Post
By the way this public repo is the most recent?
No, we've recently only been occasionally merging changes there. The most up-to-date development commits usually end up on one of the xy_sub_filter branches of my GitHub repository. Just be aware that at this point in time we give no guarantees that any given commit will actually build successfully or not contain a new critical bug. Our process usually goes, commit a bunch of experimental changes and/or bug-fixes, test, commit fixes via force push which wipes out and replaces a bunch of older commits. Before you build anything, you should always do a pull and hard reset to ensure you have the latest versions of commits.

Overall neither our XySubFilter or xy-VSFilter development branches, including the one on repo.or.cz, do we consider suitable for user testing, since at any given time they could be in an unstable state of flux. As a rule of thumb, we release new binaries on GoogleCode whenever our codebase is stable without any critical issues. If we haven't released anything, it doesn't mean we are holding back changes, but rather there is a critical bug, regression, or unfinished feature which we don't want to expose end-users to. This applies to both xy-VSFilter and XySubFilter, so just keep this in mind whenever testing unreleased code.
cyberbeing is offline   Reply With Quote
Old 6th June 2013, 00:49   #19037  |  Link
Delerue
Registered User
 
Join Date: Jun 2005
Posts: 365
I'm using XP x64 with MPC-HC x86 last build from xhmikosr + Haali splitter + madVR last build + 9600GT @301.42 driver and getting this:



Any idea?

Last edited by Delerue; 6th June 2013 at 00:52.
Delerue is offline   Reply With Quote
Old 6th June 2013, 00:58   #19038  |  Link
kasper93
MPC-HC Developer
 
Join Date: May 2010
Location: Poland
Posts: 586
@cyberbeing: Thanks, I understand what you're saying
kasper93 is offline   Reply With Quote
Old 6th June 2013, 02:29   #19039  |  Link
Stephen R. Savage
Registered User
 
Stephen R. Savage's Avatar
 
Join Date: Nov 2009
Posts: 327
The changelog mentions a "matrix=2020" and "primaries=2020" tag -- how do you set these?
Stephen R. Savage is offline   Reply With Quote
Old 6th June 2013, 02:42   #19040  |  Link
leeperry
Kid for Today
 
Join Date: Aug 2004
Posts: 3,477
^ in the file or folders name.
leeperry 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 06:07.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.