View Full Version : LAV Filters - DirectShow Media Splitter and Decoders
mrg155
7th April 2012, 23:37
Hi,
The playback of main titles from BDMV folders using the index.bdmv file is working perfectly for me in WMP/WMC (Win7) - many thanks for this feature.
The only issue I can't figure out is how to get Windows 7 Media Center to recognise the index.bdmv file so that it can start playback from the Movie Library.
If I double click on an index.bdmv file from explorer it will start fine in WMC. Usually WMC will recognise a BDMV folder and throw an error when I try to play saying I need a 3rd party blu-ray player. Changing the name of the BDMV folder doesn't work; doing this together with registering the .bdmv extension as a video file type does allow WMC to 'see' the index.bdmv file but playback fails.
Any ideas? Getting ripped blu-ray folder playback working to avoid the need for remuxing to mkv and without the need for Arcsoft or PowerDVD would be great!
Cheers
Mikey2
8th April 2012, 18:12
I was looking at Process Explorer and I noticed that LAVSplitter by far uses the highest amount of CPU/Cycles Delta of anything in the entire media-playing (MPC-HC) tree. Is this normal? (See screenshot below: I am using some pretty intense avisynth scripts and even they do not peg my processor as much as the splitter.)
If this is normal, out of curiosity, why does the splitter take so many CPU cycles? I would have thought that the decoding and/or rendering process would take the most resources (that or my avisynth script that does some advanced sharpening...)
Thanks much in advance for any replies!
MikeY
http://i155.photobucket.com/albums/s297/mikeyurkus/LAVSplitter.png
nevcairiel
9th April 2012, 13:19
Thats a common misinterpreation of those values.
The way DirectShow works, the source filter creates the threads that push the data through the filters.
That means, LAV Splitter creates a thread that sends the video frames to the decoder, and most commonly the decoder then uses that same thread to decode the data, attributing all the CPU load to the splitters thread.
So, what you are seeing there is the CPU load of the video decoder and video post-processors, not the splitter. Its just running on the splitters thread.
My guess is that the first LAVSplitter thread is video, the second audio, and a hidden third down in the list somewhere actually doing the "splitting".
fastplayer
9th April 2012, 13:52
I thought the msvcrt.dll!_endthreadex...-threads are the actual decoding threads.
nevcairiel
9th April 2012, 13:53
I thought the msvcrt.dll!_endthreadex...-threads are the actual decoding threads.
For a multi-threaded decoder probably are.
My guess is that the highest LAV thread is ffdshows post-processing and (parts of) AviSynth.
fastplayer
9th April 2012, 14:02
For a multi-threaded decoder probably are.
Since there are 6 of those in Mikey2's case, my guess was that it's all happening on a quad-core CPU.
My guess is that the highest LAV thread is ffdshows post-processing and (parts of) AviSynth.
Probably. I have yet to see LAVSplitter.ax as a top CPU consumer. It's usually madVR.ax or - most of the time - the msvcrt.dll thingies. :)
aufkrawall
9th April 2012, 16:08
There's a little glitch with LAV splitter.
With it madVR it reports the fps for the following file as 59.94 but in fact they should be 60.
MPC HC splitter reports 60 too.
Here's the sample:
http://www.mediafire.com/?j78uquuq32eyip5
nevcairiel
9th April 2012, 17:56
There's a little glitch with LAV splitter.
With it madVR it reports the fps for the following file as 59.94 but in fact they should be 60.
MPC HC splitter reports 60 too.
Here's the sample:
http://www.mediafire.com/?j78uquuq32eyip5
Fixed, for some reason 60 fps was not included the fps guessing code.
aufkrawall
9th April 2012, 17:59
Thanks, will test it with the next build. :)
edwrap
9th April 2012, 22:38
Adding a pure source filter to the external filter list has no real advantages, because MPC-HC doesn't know which formats to use it for. Adding it does no harm of course, and there may be 1-2 cases where it actually helps, but only if you have more then one source filter for a particular format installed.
If you want to use the Source Filter, use the Installer. It should set everything up for all the common formats. If there is more common formats that the installer doesn't take care of, just let me know.
I've already manually added it via the registry, but is it possible to add .m4v if mp4/mov is selected?
(I haven't checked in awhile if this was changed, but Quicktime/ios have fewer issues with embedded subtitles with the .m4v extension [source (http://blog.jharding.org/2008/11/subtitles-on-appletv-and-iphone.html)], or else I wouldn't bother using it.)
skanskan
10th April 2012, 01:03
Hello
aac audio inside some files, such as flv or mp4, doesn't work for me.
I can't get any sound from them using Potplayer, VLC or other players.
Anyway they sound in a different computer and mediainfo identifies aac audio streams inside the video.
How can I solve my problem?
I don't have any extra codec pack installed.
Windows 7 x64
regards
for example this video
http://www.youtube.com/watch?v=SK9z0z_MSJw
meetajhu
10th April 2012, 01:55
I love LAV filters and Nvidia Cuda Video! Thanks nevcairiel
LDSoft
10th April 2012, 05:04
Why so slowly reading stream from udp?
wanezhiling
10th April 2012, 06:27
nev, what's the principle of HQP? difference between enable it or not? and does HQP only affect interlaced video? Thx~:)
nevcairiel
10th April 2012, 06:55
nev, what's the principle of HQP? difference between enable it or not? and does HQP only affect interlaced video? Thx~:)
The HQP option instructs the CUVID decoder to use D3D/DXVA2 to access the hardware instead of direct hardware calls. This is required for high quality deinterlacing.
It also has some influence on progressive, for example if you set the post-processing options in the NVIDIA control panel (edge enhancement, etc), those will be applied as well if HQP is active.
wanezhiling
10th April 2012, 07:09
I'll copy your authoritative words to my those friends in question.
Many :thanks:~~
wanezhiling
10th April 2012, 07:35
New AC3Filter 2.0a released (http://ac3filter.net/wiki/Download_AC3Filter), I connected lav to it for e-ac3, but lav audio jumped. http://i.imgur.com/oK4IE.png
Does it mean AC3Filter failed to accept what lav splitter created?
Yup I think ac3filter is a little bit outdated.:)
hubblec4
10th April 2012, 12:34
New AC3Filter 2.0a released (http://ac3filter.net/wiki/Download_AC3Filter), I
Yup I think ac3filter is a little bit outdated.:)
thanks for this info.
AC3-Filter was the best and with this update i think i will go back.
nevcairiel
10th April 2012, 12:42
Good luck with that (you'll need it), but you're off-topic for this thread, so go somewhere else to discuss other DirectShow filters. :)
aufkrawall
11th April 2012, 00:24
Thanks, will test it with the next build. :)
Tested with the latest build from xhmikosr.1f0.de and I can confirm it's fixed.
Everything else seems to be working too.
Joniii
11th April 2012, 08:04
Am I missing something? I installed LAV filters to clean Windows 7 SP1, disabled media foundation then used the Win7DSFilterTweaker_5_2 to make LAV video default. Windows still uses it's own decoders for all blu-ray video formats, LAV audio filter and splitter works fine.
btcdtc
11th April 2012, 10:37
bug report:yadif not working when play 10bit video
nevcairiel
11th April 2012, 10:39
bug report:yadif not working when play 10bit video
The real bug report is that someone made a 10-bit interlaced video.
8-bit progressive is most likely much better quality then 10-bit interlaced will ever be.
Defenestration
12th April 2012, 02:48
Will LAV Audio Decoder eventually offer post-processing like Normalisation and Gain, or is it recommended to use ffdshow or AC3Filter for post-processing, while using LAV Splitter/Decoder for media splitting and audio/video decoding ?
nevcairiel
12th April 2012, 06:37
LAV Audio will most likely never support Normalisation or things of that sort.
ryrynz
12th April 2012, 07:05
"Feel free to submit a patch" ™ -Nevcairiel :D
nevcairiel
12th April 2012, 07:06
Actually, in this case i would not accept it.
I don't want such a feature.
kypec
12th April 2012, 07:37
Actually, in this case i would not accept it.
I don't want such a feature.
Yeah, I fully support your stand, nev. Proper normalization can't be done in one-pass scenario so it's not feasible for real-time playback use at all.:sly:
Anyone who wants normalized volume should re-encode their material beforehand, period. Otherwise use volume knobs and live with possible clipping, that's all you can get.
nevcairiel
12th April 2012, 09:32
I found something interesting today.
http://us.download.nvidia.com/Windows/301.24/301.24-win7-winvista-desktop-release-notes.pdf
Quote from the fixed issues in 301.24
[GeForce 500/400 Series]: Media Player Classic–there is no picture when using LAV
DirectShow filters MPEG4 video playback. [952108]
LAV was mentioned in the NVIDIA release notes. :D
I should install 301.24 later, see how things work (on 301.10 right now)
Octo-puss
12th April 2012, 09:36
I am stunned these companies do give a sh*t about non-commercial software.
Andy o
12th April 2012, 09:54
I found something interesting today.
http://us.download.nvidia.com/Windows/301.24/301.24-win7-winvista-desktop-release-notes.pdfI presume this is unrelated to the madVR problems?
nevcairiel
12th April 2012, 10:11
Who knows, go try it.
I suspect madVR does something thats not really officially documented to work like that in D3D though.
Its really not a big issue though, just turn off "use separate device for presentation" and everything is fine again.
Doesn't seem to have any bad side effects.
dukey
12th April 2012, 10:42
I am stunned these companies do give a sh*t about non-commercial software.
Nvidia have always been pretty good about driver support. AMD on the other hand. On your graphics card is 4-5 years old ? Sorry we won't fix those critical bugs.
madshi
12th April 2012, 10:55
madVR uses an official D3D feature for the option "use separate device for presentation", namely sharing surfaces between different D3D devices. That feature is rarely used by other software, though, so that's probably why nobody else is affected by what appears to me to be an NVidia driver bug.
aufkrawall
12th April 2012, 12:05
madVR uses an official D3D feature for the option "use separate device for presentation", namely sharing surfaces between different D3D devices. That feature is rarely used by other software, though, so that's probably why nobody else is affected by what appears to me to be an NVidia driver bug.
There's a chance that we will see a similar changelog entry with one of the next drivers for madVR just like for LAV. :)
madshi
12th April 2012, 12:18
Is NVidia even aware of the issue? I'm not sure...
aufkrawall
12th April 2012, 12:34
The issue has been reported, just like previously the LAV one.
I'm confident that we just have to wait (could take some months though, but we'll see).
madshi
12th April 2012, 13:42
Where has it been reported? And the LAV one? Do you have a link? Just wondering where to report other issues myself. Thx.
aufkrawall
12th April 2012, 13:56
Where has it been reported?
I don't have any more details, I'm sorry. :(
I don't want to leak private messages information.
Do you have a link? Just wondering where to report other issues myself. Thx.
You can either report it at the Nvidia forums or send me a PM how to reproduce if it's not too difficult.
Then I can report it to someone else who has better access.
The way via the forums may not be the better one. ;)
madshi
12th April 2012, 14:41
Great, thx!
Skibicki
12th April 2012, 14:49
nevcairel
Does Lav override GPU driver deinterlacing settings? I have a problem with motion adaptive deinterlacing being switched off randomly in Catalyst.
wanezhiling
12th April 2012, 14:51
http://www.station-drivers.com/page/nvidia%20forceware.htm
;)
kasper93
12th April 2012, 14:52
Where has it been reported? And the LAV one? Do you have a link? Just wondering where to report other issues myself. Thx.
I don't know really, but maybe there http://www.nvidia.com/object/driverqualityassurance.html and for AMD http://emailcustomercare.amd.com/
PeQuE
12th April 2012, 15:32
I found something interesting today.
http://us.download.nvidia.com/Windows/301.24/301.24-win7-winvista-desktop-release-notes.pdf
Quote from the fixed issues in 301.24
LAV was mentioned in the NVIDIA release notes. :D
I should install 301.24 later, see how things work (on 301.10 right now)
Congrats... or not :D
mindbomb
12th April 2012, 17:45
nevcairel
Does Lav override GPU driver deinterlacing settings? I have a problem with motion adaptive deinterlacing being switched off randomly in Catalyst.
for amd, no, unless you use yadif.
mark0077
12th April 2012, 18:30
I found something interesting today.
http://us.download.nvidia.com/Windows/301.24/301.24-win7-winvista-desktop-release-notes.pdf
Quote from the fixed issues in 301.24
LAV was mentioned in the NVIDIA release notes. :D
I should install 301.24 later, see how things work (on 301.10 right now)
Nice one! Do you think they would listen if you were to mention to them that cuda vc-1 decoding produces blocky artifacts on VP2 hardware, eg GTX295
They certainly don't listen to me when I tell them on their forums etc..... :(
nevcairiel
12th April 2012, 18:33
Do you think they would listen if you were to mention to them that cuda vc-1 decoding produces blocky artifacts on VP4 hardware, eg GTX295
Thats VP2 hardware, not VP4. Its only a problem because those ancient cards only support partial acceleration.
It works fine on VP4, where there is full acceleration for VC-1.
mark0077
12th April 2012, 18:40
Sorry VP2 yes. I guess thats a no that you don't want to mention it to them ;) did you inform them via a post or PM? Just want to know whats the best way to get a message through to them..
aufkrawall
12th April 2012, 19:44
Did it ever work with any driver?
If it's rather a hardware issue they hardly will be able to fix it via driver.
And VP2 isn't that new anymore.
My advice: Don't wait for a fix. ;)
mark0077
12th April 2012, 19:49
Yeah it worked, I even posted the list of drivers where it worked / didn't. Yeah its a pity they don't seem to care about older hardware. The drivers get progressively better and better in some areas, like to support new games, and get worse and worse over time in other areas / or for older games and hardware... Oh well. A little bit more regression testing wouldn't go astray :)
vBulletin® v3.8.11, Copyright ©2000-2025, vBulletin Solutions Inc.