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 17th May 2010, 12:31   #2681  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by starla View Post
This would cause the motion vectors to get "corrupted" and the movement would be still juddering. Some frames would contain more of the linear movement than others and the frame duration is still limited with the HW v-sync.
I think I didn't properly explain how it works. Basically the renderer would always do a weighted blend of two adjacent frames. The weights are defined by the distance of the hardware vsync to the "ideal vsync". I'm sure that this solution would be perfectly smooth. However, due to the blending it might be slightly less sharp. Not sure about that...

Actually this frame blending is exactly what many studios do when converting natively interlaced NTSC TV series to PAL (60i -> 50i). Motion is perfectly smooth this way.

Quote:
Originally Posted by starla View Post
This would work only depending on a movie type. Just imagine how hard such sweet spot could be to archive on a Planet Earth blu-rays for example (they contain pretty much constant panning / fly-overs scenes).

Again I think it would be only a sub par solution.
I think you are underestimating how well it would work. Ok, there will be a couple of movies where it might not work, e.g. handy cam movies like Cloverfield. Such movies make me motion sick, anyway. But I think for 99% of all movies, solution 2) should work pretty well. The key is that if you need to drop only e.g. 3 frames per movie, there would only have to be one single motion-less scene every half hour of the movie for 2) to work. Just a simple fadeout to black (scene change) does the trick!

Quote:
Originally Posted by starla View Post
About the drop few frames during whole movie. I cannot tolerate that.
I can't tolerate *noticeable* frame drops during a movie, either. I have no problem with frame drops, though, if they're not visible to the naked eye.

Last edited by madshi; 17th May 2010 at 12:34.
madshi is offline   Reply With Quote
Old 17th May 2010, 12:32   #2682  |  Link
tetsuo55
MPC-HC Project Manager
 
Join Date: Mar 2007
Posts: 2,317
Quote:
Originally Posted by starla View Post
There are few options. If you know any other (something thaI have no plans to implement a perfect audio renderer from audio point of view since it will disable possibility to have perfectly judder free video playback. I'm a guy who falls into the category who tolerates minor audio quality loss (not including pops, clicks or any noises!) to have perfectly fluid video playback.
Having the audio stream as perfect as possible still minimised glitches introduced when the audio is modified to fit the video timing.
I dont think your goal contradicts perfect audio processing in any way.
__________________
MPC-HC, an open source project everyone can improve. Want to help? Test Nightly Builds, submit patches or bugs and chat on IRC
tetsuo55 is offline   Reply With Quote
Old 17th May 2010, 12:45   #2683  |  Link
starla
Registered User
 
Join Date: Jan 2004
Posts: 56
Quote:
Originally Posted by tetsuo55 View Post
Having the audio stream as perfect as possible still minimised glitches introduced when the audio is modified to fit the video timing.
I dont think your goal contradicts perfect audio processing in any way.
When resampling audio (even for few brief moments during the whole movie) is not going to give perfect audio playback and thus I'm sure it wont be good enough for the people who have golden ears (people who freeze their CD disks before listening... or say that they are able to hear some minimal jitter introduced in the audio chain fall into this hole ).

I'm not after a bit perfect audio playback, instead I'm fine with a bit molested audio stream. So, currently I'm not planning to make the audio renderer perfect in the way that was proposed (I just don't have enough resources and maybe even the skills needed).

I think a separate thread would be good since we are drifting away from the madVR into a more common direction.
starla is offline   Reply With Quote
Old 17th May 2010, 14:10   #2684  |  Link
Mark_A_W
3 eyed CRT supporter
 
Join Date: Jan 2008
Location: Or-strayl-ya
Posts: 563
I think the only practical solution, other than resampling the audio (a la Reclock), is to fiddle the video timings. Stretch and compress the actual video timings on the card.

Ar-jar did this with gothplayer. He moved away from this method, I never really understood why, as it is one of the two possible solutions.


That said, Reclock sounds fine to me. It's all whizzbangy and upsampling now.


And Powerstrip is going the way of the dodo. The latest ATi cards don't support it, and I don't think they ever got the last couple of generations of Nvidia cards working properly, both due to insufficient info from the manufacturers.
Mark_A_W is offline   Reply With Quote
Old 17th May 2010, 15:09   #2685  |  Link
tetsuo55
MPC-HC Project Manager
 
Join Date: Mar 2007
Posts: 2,317
Yeah a new thread is possibly best, and as starla suggest, we need to think of scientific ways of measuring audio quality before we can do any coding in that area.

I suggest building a scientifically accurate test-suite for the video renderer too.
__________________
MPC-HC, an open source project everyone can improve. Want to help? Test Nightly Builds, submit patches or bugs and chat on IRC
tetsuo55 is offline   Reply With Quote
Old 17th May 2010, 15:54   #2686  |  Link
pie1394
Registered User
 
Join Date: May 2009
Posts: 212
Quote:
Originally Posted by madshi View Post
I think I didn't properly explain how it works. Basically the renderer would always do a weighted blend of two adjacent frames. The weights are defined by the distance of the hardware vsync to the "ideal vsync". I'm sure that this solution would be perfectly smooth. However, due to the blending it might be slightly less sharp. Not sure about that...

Actually this frame blending is exactly what many studios do when converting natively interlaced NTSC TV series to PAL (60i -> 50i). Motion is perfectly smooth this way.
This reminds me that MPEG encoded video often gets sharper image on P-frame, while it is often foggy (+ even blocky on MPEG2) on 2 consequent B-frames. This is often how it is encoded even with high video bit rate DVD / BD, as well as good BS-HI / ATSC HDTV contents. Yet most human eyes cannot observe B-frame's unsharpness or blocky 16x16 macroblock during normal playback speed.

If the P-frame is encoded with insufficient bit rate, however, the picture loses its sharpness at all.
pie1394 is offline   Reply With Quote
Old 17th May 2010, 18:01   #2687  |  Link
FoLLgoTT
And so it begins...
 
FoLLgoTT's Avatar
 
Join Date: Nov 2005
Location: Hannover, Germany
Posts: 64
Quote:
Originally Posted by madshi View Post
2) drop/repeat frames only in scenes where there's (almost) no motion

A similar solution to 2) could also be done in the audio renderer: Don't resample audio at all. Instead drop/repeat "silence".
ReClock does dropping/repeating with S/PDIF throughput. AC3/DTS samples get dropped/repeated periodically when video refresh rate does not match exactly. This can be audible when occuring in the wrong scene. Dropping/reapeating only in silent scenes would be a good enhancement.

But if refresh rate is tweaked correctly there is no need to drop/repeat for many many hours.

Last edited by FoLLgoTT; 17th May 2010 at 18:09.
FoLLgoTT is offline   Reply With Quote
Old 17th May 2010, 19:30   #2688  |  Link
iSeries
Registered User
 
Join Date: Jan 2009
Posts: 625
About 'tweaking' refresh rates slightly during playback to ensure smooth playback - with my ATI 4550 I get absolutely no audio if anything other than the standard refresh rates from the ATI panel are forced.
iSeries is offline   Reply With Quote
Old 17th May 2010, 21:03   #2689  |  Link
6233638
Registered User
 
Join Date: Apr 2009
Posts: 1,019
Quote:
Originally Posted by FoLLgoTT View Post
But if refresh rate is tweaked correctly there is no need to drop/repeat for many many hours.
Any way to do this with nvidia?

I'm not sure how you would do without reclock though, bluray releases are a combination of 23.976, 24.000 and 25.000 fps, with the original source being 24.000. Reclock plays all three back at 24.000 and upsamples audio to 32/192 to minimise quality loss.
6233638 is offline   Reply With Quote
Old 17th May 2010, 21:11   #2690  |  Link
Razoola
Registered User
 
Join Date: May 2007
Posts: 454
I'm one of those people like starla who would rather have perfect video over audio. I don't like the sound of frame interpoliation no matter how its presented. LCD panel makers have been trying it for years and still havent got it right without artifacts or some form. Its one of the reasons I have gone off of reclock somewhat (though I still use it), the tool used to put video smoothness (judder) over audio. That was the entire point of the tool back in the day of gfx cards with TV out, and I'm not talking about DVI or HDMI. Since it got taken over from Ogg the program has become more turned to audio rather than video, not that its a bad thing but it seems these days the current devs are not intrested in perfect video as much as perfect audio. I just wish more effort was put into bitstreaming with reclock. I don't use reclock resamplers at all, for me the tool is best when bitstreaming, even the odd dropped or repeated audio packet every couple of minutes is no problem against the miniscule CPU usage it requires compared to resampling. Anyway, maybe i'm talking a bit much in the wrong forum about that.

For me the video rendering should be a priority over audio when it comes to watching content. The video renderer should not start adjusting the image to allow audio to be perfect. I feel that could be a slippery slope, espicially when audio is already getting resampled left right and centre.

Btw, Im a person who likes the sound from vinyl more than CD
Razoola is offline   Reply With Quote
Old 17th May 2010, 21:22   #2691  |  Link
Razoola
Registered User
 
Join Date: May 2007
Posts: 454
Quote:
Originally Posted by FoLLgoTT View Post
ReClock does dropping/repeating with S/PDIF throughput. AC3/DTS samples get dropped/repeated periodically when video refresh rate does not match exactly. This can be audible when occuring in the wrong scene. Dropping/reapeating only in silent scenes would be a good enhancement.

But if refresh rate is tweaked correctly there is no need to drop/repeat for many many hours.
This is what I want from reclock or any new audio render in its place. Dropping/Repeating audio packets during silence via bitstreaming to cancel lip sync. I personally feel thats the perfect solution, and it also requires the least cpu cycles to boot (compared to audio resampling or video frame creation). How can you beat that?

Remember we are talking packet sizes of only 32ms for dolby and 11ms for DTS.

Last edited by Razoola; 17th May 2010 at 21:35.
Razoola is offline   Reply With Quote
Old 17th May 2010, 21:44   #2692  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by Razoola View Post
For me the video rendering should be a priority over audio when it comes to watching content. The video renderer should not start adjusting the image to allow audio to be perfect. I feel that could be a slippery slope, espicially when audio is already getting resampled left right and centre.
Every video renderer has to make sure that video presentation doesn't run out of sync. That means that sometimes frames have to be dropped or repeated - unless video and audio clocks are perfectly matched. Don't worry about madVR doing violence to the image. I won't do that, or if I do, it will be an option which can be turned off. I do plan to "hide" frame drops/repeats by doing them during scenes where it will be less noticeable. But if you tweak your audio/video clocks so well that no frames need to be dropped/repeated in the first place, madVR will be happy and not drop/repeat anything, of course.
madshi is offline   Reply With Quote
Old 17th May 2010, 21:59   #2693  |  Link
Razoola
Registered User
 
Join Date: May 2007
Posts: 454
Quote:
Originally Posted by madshi View Post
Every video renderer has to make sure that video presentation doesn't run out of sync. That means that sometimes frames have to be dropped or repeated - unless video and audio clocks are perfectly matched. Don't worry about madVR doing violence to the image. I won't do that, or if I do, it will be an option which can be turned off. I do plan to "hide" frame drops/repeats by doing them during scenes where it will be less noticeable. But if you tweak your audio/video clocks so well that no frames need to be dropped/repeated in the first place, madVR will be happy and not drop/repeat anything, of course.
Yes I understand. Being from PAL land maybe I sometimes forget these things (being used to content speeded up to match the refresh rate).
Razoola is offline   Reply With Quote
Old 17th May 2010, 22:47   #2694  |  Link
djsolidsnake86
Registered User
 
Join Date: Mar 2010
Posts: 139
with wmv files madvr don't show the video
djsolidsnake86 is offline   Reply With Quote
Old 17th May 2010, 23:06   #2695  |  Link
leeperry
Kid for Today
 
Join Date: Aug 2004
Posts: 3,477
Quote:
Originally Posted by djsolidsnake86 View Post
with wmv files madvr don't show the video
sample and more infos needed, or it didn't happen...WMV play fine here. maybe a DRM'ed file?
Quote:
Originally Posted by Razoola View Post
Yes I understand. Being from PAL land maybe I sometimes forget these things (being used to content speeded up to match the refresh rate).
yeah, 25fps needs to die...w/o Reclock doing its job bye bye 24fps.
Quote:
Originally Posted by Razoola View Post
For me the video rendering should be a priority over audio when it comes to watching content.
exactly, video glitches completely ruin the experience for me...I project on a very large screen, and dropped frames are just unbearable.

many ppl who complain about Reclock resampling actually play their movies in DirectSound...which does the exact same thing, as DirectSound cannot be made bit-perfect on Vista/W7(unlike XP when all its sliders are maxed out ): http://www.hydrogenaudio.org/forums/...dpost&p=522247

I'm not trolling or anything, but teh "evil resampling" will most likely occur anyway(except for the bistream/pass-through ppl...but no 25@24fps for them then ), and quite frankly Reclock in its highest libsamplerate SINC mode does a hell of a job...conduct some proper DBT, tell us how it went

and I'm quite sure that many ppl who are so proud of their bit-perfect audio don't EQ their room resonances down...
Quote:
Originally Posted by Razoola View Post
Im a person who likes the sound from vinyl more than CD
terribly OT, but I think this sort of statement boils down to:
-unbearable "loudness war" CD mastering...4dB of dynamics, great! http://www.dynamicrange.de/
-cheapo opamps in the CD players they tried , as a 4580/5532 will sound artifacty/agressive/edgy/narrow when two LT1028/AD4627-1/LT1363/OPA827/AD797 will sound as "analog" as can get...and far more detailed than vinyl will ever be able to(50dB crosstalk on vinyl I think?)

some high quality mastered CD on a CD player running from a linear regulated PSU w/ high end opamps will sound awfully good...w/o the nasty vinyl crackles and pops at that

Last edited by leeperry; 18th May 2010 at 02:41.
leeperry is offline   Reply With Quote
Old 17th May 2010, 23:46   #2696  |  Link
Thunderbolt8
Registered User
 
Join Date: Sep 2006
Posts: 2,197
this seems to be a newb question, but have all those chroma and luma up-/downsampling settings any effect when the displayed size of the video and resolution of the output screen is 1:1 ? or is this something entirely different and has nothing to do with that?
Thunderbolt8 is offline   Reply With Quote
Old 18th May 2010, 00:17   #2697  |  Link
leeperry
Kid for Today
 
Join Date: Aug 2004
Posts: 3,477
Quote:
Originally Posted by Thunderbolt8 View Post
have all those chroma and luma up-/downsampling settings any effect when the displayed size of the video and resolution of the output screen is 1:1 ?
luma nope, but chroma yes as it's stored at half resolution in the YV12 stream so it needs to be upscaled...and the softer the better IMHO.

you can always use my "Bronzés" sample(the one shown in the OP) and run tests: http://www.mediafire.com/file/zirwxawggmy/Bronz%E9s.mkv

Last edited by leeperry; 18th May 2010 at 02:30.
leeperry is offline   Reply With Quote
Old 18th May 2010, 02:17   #2698  |  Link
ajp_anton
Registered User
 
ajp_anton's Avatar
 
Join Date: Aug 2006
Location: Stockholm/Helsinki
Posts: 805
I'm having problems with some wmv files. Found a small one that won't play:
http://www.multiupload.com/PFEET9HBGS
Using MPC-HC with internal decoder.
ajp_anton is offline   Reply With Quote
Old 18th May 2010, 04:46   #2699  |  Link
slvbutn
Registered User
 
Join Date: Sep 2009
Posts: 1
Is it a bug?

I have intel G45 video card, which support non power of 2 texture, But madVr report: "graphics card only supports power of 2 textures".
slvbutn is offline   Reply With Quote
Old 18th May 2010, 08:43   #2700  |  Link
djsolidsnake86
Registered User
 
Join Date: Mar 2010
Posts: 139
General
Complete name : C:\Users\Andy\Desktop\Nuova cartella\doll.wmv
Format : Windows Media
File size : 10.3 MiB
Duration : 2mn 21s
Overall bit rate : 612 Kbps
Maximum Overall bit rate : 1 024 Kbps
Encoded date : UTC 1970-01-01 00:00:00.000

Video
ID : 1
Format : WMV2
Codec ID : WMV2
Codec ID/Info : Windows Media Video 8
Description of the codec : wmv2
Duration : 2mn 21s
Bit rate : 354 Kbps
Width : 638 pixels
Height : 274 pixels
Display aspect ratio : 2.35:1
Frame rate : 23.976 fps
Bit depth : 8 bits
Bits/(Pixel*Frame) : 0.085
Stream size : 5.96 MiB (58%)

Audio
ID : 2
Format : WMA
Format version : Version 2
Codec ID : 161
Codec ID/Info : Windows Media Audio
Description of the codec : Windows Media Audio V8
Duration : 2mn 21s
Bit rate : 224 Kbps
Channel(s) : 2 channels
Sampling rate : 44.1 KHz
Bit depth : 16 bits
Stream size : 3.77 MiB (37%)

this is the mediainfo of the wmv file that don't work with madvr (with evrsync yes)
you are right, other wmv files work correctly!
djsolidsnake86 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 05:54.


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