L-SMASH Source

[StainlessS]

I have a modest proposal: how about you upgrade to an OS from the last decade

No !

I dont expect that Zathor was aware that he was breaking MeGUI for XP when downloading The_Weirdo's offering. I'm perfectly OK with The_Weirdo not supporting XP, what I do have a problem with is breaking stuff due to not stating requirement.

EDIT:
Yes, I shouldn't need to post this in another thread, and yet I am :).
And I should not have to post this.

[burfadel]

And I should not have to post this.

It's 2015, not 2007. In 2007, there was a reason to stick to XP - Windows Vista. There is really no valid reason to to stick with XP any more. System can't run Windows 7/8.1/10? Well, it probably shouldn't be used for encoding either. Stressing old components and all.

Not to mention Windows XP is no longer supposed security patch wise, unless on that paid plan or whatever (which would be stupid to be part of), which they're ending shortly anyway. It's hardly a point worth saying. It's old, defunct, full of security holes, and regardless of any complaints I might get because of this statement, they shouldn't be encouraged to continue using it. If they're too stubborn to change from XP why on earth are they wanting to latest encoding tools?

[Wilbert]

@burfadel, stop trolling. It's not up to you to say which os people should use.

[burfadel]

It wasn't my intention to troll. Actually Microsoft have summed up what I said (and keeping in mind this page isn't new):
http://windows.microsoft.com/en-AU/windows/end-support-help

With that in mind, and people still using XP for banking and other financial transfers, I just fully believe it's continued use shouldn't be encouraged.

[hello_hello]

It's 2015, not 2007. In 2007, there was a reason to stick to XP - Windows Vista. There is really no valid reason to to stick with XP any more.

Granted, software/driver support for XP isn't universal any more, but even in 2015 the best Windows upgrade scenario I can expect still seems to be the software I use will continue to work as it always has.

If they're too stubborn to change from XP why on earth are they wanting to latest encoding tools?

It's a tangible benefit thing and a not an upgrading just for the sake of it thing.
The next PC I build (which is well overdue) won't run XP, but this one does (and very possibly will until it dies) so it'd be nice to know when updated software no longer supports XP. If that's okay with you?

[jones1913]

@xp users try this: https://www.mediafire.com/folder/17q7t31dib18t/LSMASH-Works

Some time ago I build the dll for personal use, but then I stopped when the_weirdo offers his builds. It was easy to reactivate my build environment now...
Build with VS2010, tell me if it works and how speed compares. :)

I mostly agree that it is a bad idea to use xp on production/online systems these days, but it is not up to us to force anybody to upgrade.
Everybody must self decide and weight up the risks.

I personally use a NT 5.2 x64 virtual machine on my Linux PC for Windows specific tasks, because a Win7 VM would need 3 times the disk space.
And this system got all updates via offline method and never saw an online connection.

[hello_hello]

@xp users try this: https://www.mediafire.com/folder/17q7t31dib18t/LSMASH-Works

Some time ago I build the dll for personal use, but then I stopped when the_weirdo offers his builds. It was easy to reactivate my build environment now...
Build with VS2010, tell me if it works and how speed compares. :)

I haven't checked for any speed difference yet but it definitely works on XP. Thank you!

I mostly agree that it is a bad idea to use xp on production/online systems these days, but it is not up to us to force anybody to upgrade.
Everybody must self decide and weight up the risks.

I honestly don't know how people get infected. It must take a bit of effort, even when running XP. It's not like we're all still using IE6 and in danger of an ActiveX drive-by infection wherever we go.
No antivirus program here. No anti-malware program. No software firewall. Just me, XP and Firefox (behind a router). I haven't even run Windows update in over a year. Around once per year or so I do install anti-virus software and run a full scan before restoring the previous Windows image. Just to see what it'll find. Then I restore the image, update everything and create a new one. I've been doing that for years and so far the nastiest thing a scan has found is a naughty cookie or two.

[the_weirdo]

Well at least please let people know of the limitations everywhere that you post your links.
Thank you.

You're right. It's my fault. Sorry about that. When I made the change, I forgot to announce it here. Then I'd completely forgotten about it until hello_hello reported his problem.

I'm not forcing anyone to upgrade. They're free to use whichever OS that suits them. That said, I've my own decision and I'm not trying to please everyone.

[StainlessS]

I'm not forcing anyone to upgrade.

Yes I know, sorry if I sounded curt, was not my intent and is of course your decision.
Anyways, Peace bro :)

PS, Thank you Jonesy (I hate being called that), I'll give it a try :thanks:

[Groucho2004]

I honestly don't know how people get infected.
It's a mystery to me as well. I think it's Pebcac (http://www.urbandictionary.com/define.php?term=Pebcac).

No antivirus program here. No anti-malware program. No software firewall. Just me, XP and Firefox (behind a router). I haven't even run Windows update in over a year. Around once per year or so I do install anti-virus software and run a full scan before restoring the previous Windows image. Just to see what it'll find. Then I restore the image, update everything and create a new one. I've been doing that for years and so far the nastiest thing a scan has found is a naughty cookie or two.
Pretty much the same here. I do however use a software firewall, mainly to monitor any odd activities. Call me paranoid. :)

The only time I got a virus/worm was in the early 2000's (Blaster). At that time I used Win2000, ISDN dialup without any firewall. I just restored my Ghost image and the problem was solved in 5 minutes.

[Selur]

Only the LWLibavVideoSource builds with libav should be able to handle 10bit x264 content, right?

[the_weirdo]

Only the LWLibavVideoSource builds with libav should be able to handle 10bit x264 content, right?

No. Both builds with FFmpeg and Libav should be able to do that.

[Selur]

Strange, for me both don't.
L-SMASH-Works-r729-20141117-32bit-Libav
L-SMASH-Works-r784-20150301-32bit
both only produce garbled output and my script looks like this:
LoadPlugin("G:\MkvCutter\LSMASHSource.dll")
LWLibavVideoSource("H:\Output\HIGH10~1.MKV", cache=false)
http://s17.postimg.org/sjy0vk9zj/test.png
Here's also the mediainfo data of the source:
General
Count : 303
Count of stream of this kind : 1
Kind of stream : General
Kind of stream : General
Stream identifier : 0
Unique ID : 173785410236381370390882345944681015379
Unique ID : 173785410236381370390882345944681015379 (0x82BDDA40884D25E1AF571D15FC4FE853)
Count of video streams : 1
Video_Format_List : AVC
Video_Format_WithHint_List : AVC
Codecs Video : AVC
Video_Language_List : English
Complete name : h:\Output\high10_10bit.mkv
Folder name : h:\Output
File name : high10_10bit
File extension : mkv
Format : Matroska
Format : Matroska
Format/Url : http://packs.matroska.org/
Format/Extensions usually used : mkv mk3d mka mks
Commercial name : Matroska
Format version : Version 4 / Version 2
Codec : Matroska
Codec : Matroska
Codec/Url : http://packs.matroska.org/
Codec/Extensions usually used : mkv mk3d mka mks
File size : 7234313
File size : 6.90 MiB
File size : 7 MiB
File size : 6.9 MiB
File size : 6.90 MiB
File size : 6.899 MiB
Duration : 17160
Duration : 17s 160ms
Duration : 17s 160ms
Duration : 17s 160ms
Duration : 00:00:17.160
Duration : 00:00:17:04
Duration : 00:00:17.160 (00:00:17:04)
Overall bit rate : 3372640
Overall bit rate : 3 373 Kbps
Frame rate : 25.000
Frame rate : 25.000 fps
Frame count : 429
Stream size : 143963
Stream size : 141 KiB (2%)
Stream size : 141 KiB
Stream size : 141 KiB
Stream size : 141 KiB
Stream size : 140.6 KiB
Stream size : 141 KiB (2%)
Proportion of this stream : 0.01990
Encoded date : UTC 2015-03-17 20:03:51
File creation date : UTC 2015-03-17 20:03:44.514
File creation date (local) : 2015-03-17 21:03:44.514
File last modification date : UTC 2015-03-17 20:03:51.913
File last modification date (local) : 2015-03-17 21:03:51.913
Writing application : mkvmerge v7.5.0 ('Glass Culture') 64bit built on Jan 26 2015 13:43:13
Writing library : libebml v1.3.1 + libmatroska v1.4.2
Writing library : libebml v1.3.1 + libmatroska v1.4.2
Encoding Gui : Hybrid 2015.03.16.1

Video
Count : 297
Count of stream of this kind : 1
Kind of stream : Video
Kind of stream : Video
Stream identifier : 0
StreamOrder : 0
ID : 1
ID : 1
Unique ID : 6920789142181302940
Format : AVC
Format/Info : Advanced Video Codec
Format/Url : http://developers.videolan.org/x264.h
tml
Commercial name : AVC
Format profile : High 10@L3.0
Format settings : CABAC / 4 Ref Frames
Format settings, CABAC : Yes
Format settings, CABAC : Yes
Format settings, ReFrames : 4
Format settings, ReFrames : 4 frames
Internet media type : video/H264
Codec ID : V_MPEG4/ISO/AVC
Codec ID/Url : http://ffdshow-tryout.sourceforge.net
/
Codec : V_MPEG4/ISO/AVC
Codec : AVC
Codec/Family : AVC
Codec/Info : Advanced Video Codec
Codec/Url : http://ffdshow-tryout.sourceforge.net
/
Codec profile : High 10@L3.0
Codec settings : CABAC / 4 Ref Frames
Codec settings, CABAC : Yes
Codec_Settings_RefFrames : 4
Duration : 17160
Duration : 17s 160ms
Duration : 17s 160ms
Duration : 17s 160ms
Duration : 00:00:17.160
Duration : 00:00:17:04
Duration : 00:00:17.160 (00:00:17:04)
Bit rate : 3305524
Bit rate : 3 306 Kbps
Width : 640
Width : 640 pixels
Height : 352
Height : 352 pixels
Pixel aspect ratio : 1.000
Display aspect ratio : 1.818
Display aspect ratio : 16:9
Frame rate mode : CFR
Frame rate mode : Constant
Frame rate : 25.000
Frame rate : 25.000 fps
Frame count : 429
Resolution : 10
Resolution : 10 bits
Colorimetry : 4:2:0
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 10
Bit depth : 10 bits
Scan type : Progressive
Scan type : Progressive
Interlacement : PPF
Interlacement : Progressive
Bits/(Pixel*Frame) : 0.587
Delay : 0
Delay : 00:00:00.000
Delay, origin : Container
Delay, origin : Container
Stream size : 7090350
Stream size : 6.76 MiB (98%)
Stream size : 7 MiB
Stream size : 6.8 MiB
Stream size : 6.76 MiB
Stream size : 6.762 MiB
Stream size : 6.76 MiB (98%)
Proportion of this stream : 0.98010
Writing library : x264 - core 146 r2538 121396c
Writing library : x264 core 146 r2538 121396c
Writing library/Name : x264
Writing library/Version : core 146 r2538 121396c
Encoding settings : cabac=1 / ref=3 / deblock=1:0:0 / analyse=0x3:0x111 / me=hex / subme=5 / psy=1 / psy_rd=1.00:0.00 / mixed_ref=1 / me_range=16 / chroma_me=1 / trellis=0 / 8x8dct=1 / cqm=0 / deadzone=21,11 / fast_pskip=0 / chroma_qp_offset=0 / threads=12 / lookahead_threads=2 / sliced_threads=0 / nr=0 / decimate=1 / interlaced=0 / bluray_compat=0 / constrained_intra=0 / bframes=3 / b_pyramid=2 / b_adapt=1 / b_bias=0 / direct=3 / weightb=1 / open_gop=0 / weightp=1 / keyint=250 / keyint_min=25 / scenecut=40 / intra_refresh=0 / rc=cqp / mbtree=0 / qp=18 / ip_ratio=1.40 / pb_ratio=1.30 / aq=0
Default : Yes
Default : Yes
Forced : Yes
Forced : Yes
colour_description_present : Yes
Matrix coefficients : BT.601
Color range : Limited

[qyot27]

Unless something has changed, it's LSMASHSource itself that doesn't support >8 bit in AviSynth. It's probably just doing a straight input->output treatment of the pixfmt, and since AviSynth doesn't have >8 bit pixfmts, that's what you get. In VapourSynth you'd get the proper image.

[sneaker_ger]

Yes, you have to force a compatible output format:
format="YUV420P8" or format="YUV420P16" (stacked)

[MeteorRain]

both only produce garbled output and my script looks like this:

This is exactly the correct output of a high bit depth video.

You'll need a tool that can display high bit depth preview, or, you'll have to force a 8 bit output, which is YUV420P8 as said above.

[stax76]

What is the runtime situation currently? Doesn't it work without installing a runtime? A few StaxRip users have problems running the plugin, for me it runs but I have installed Visual Studio 2013 Community Edition of course.

[sneaker_ger]

If this is still true they need VC++2013: http://avisynth.nl/index.php/LSMASHSource

[stax76]

How much bigger would the file size be if it was statically linked and provided as 7z download? Could anybody help with a building and uploading this?

[kuchikirukia]

So, is L-SMASH broken for decoding interlaced content out of MKV? I get jumps no mater which field order I use when using QTGMC.

[LigH]

A broken interlacing is more probable; or Telecine instead of Interlacing. Did you provide a sample of your material?

[kuchikirukia]

It deinterlaces with DGDecode. It deinterlaces with L-SMASH out of .h264 (though it sets as 119/2 fps).
Out of MKV with QTGMC it thinks it's 120fps and it jerks back and forth if you set L-SMASH to 30000/1001 (though not quite as bad as if the field order is wrong).

Out of MP4 with LSMASHVideoSource and no filters I got several instances of "failed to make a frame."
And that ain't right: E: never mind, that's in the mp4. First time I used this drive was for this MP4, so there might be an issue with it.
http://s14.postimg.org/t0wxc5cvh/untitled.jpg

[Music Fan]

The problem can come from your mkv : if the original container of your video was ts, and if you opened it directly in MKVmerge, the framerate can be wrong, even if you specified the correct framerate in MKVmerge.
I noticed this problem with several interlaced H264 videos. After remux in MKV, MediaInfo detects VFR while it's CFR.
To avoid this, the ts has to be demuxed in elementary streams before remux in mkv.
But it's not too late, you can still demux mkv in elementary streams and remux it.

[kuchikirukia]

It was extracted with HD streams extractor, and it played at 30fps interlaced / 60fps LAV filter YADIF bobbed.

E: Ok, now it's not doing it. We're getting 60000/1001 and no jumping.

[Music Fan]

E: Ok, now it's not doing it. We're getting 60000/1001 and no jumping.
What did you do to make it work now ? What I explained ?

[Overdrive80]

Can not LSmash indexing avi files encoded with magicyuv???

http://forum.doom9.org/showthread.php?p=1722304#post1722304

[Reel.Deel]

Can not LSmash indexing avi files encoded with magicyuv???

http://forum.doom9.org/showthread.php?p=1722304#post1722304

FFmpeg/Libav does not include (http://forum.doom9.org/showthread.php?p=1708190&highlight=ffmpeg#post1708190) MagicYUV. I just use AviSource.

[LigH]

As long as libavcodec doesn't contain a decoder for MagicYUV, L-SMASH Source may not support this video format. Even though I believe it would be possible to index an AVI regardless of the used codec; but why index a clip if the video codec is not supported in libavcodec, thus neither in ffmpeg nor in related projects like FFMS2 or L-SMASH Source.

[Overdrive80]

Ok, thanks for both. I´ll use avisource.

[stax76]

I asked DG how he sends progress from DGIndex and DGIndexNV, his answer:

void OutputProgress(int progr)
{
static int lastprogress = -1;

if (progr != lastprogress)
{
char percent[20];
DWORD written;

sprintf(percent, "%d\r", progr);
WriteFile(GetStdHandle(STD_OUTPUT_HANDLE), percent, (DWORD) strlen(percent), &written, NULL);
lastprogress = progr;
}
}

It would be very useful for GUI users to see progress.

[videoh]

This is important too for an implementer:

"Every 30 frames, this function is called to write to stdout"

and

"progr is an int 0-100 indicating the percent completed. The caller determines the percent complete from the ratio of the current file position and the size of the file. That way you don't need to know the number of frames in the file."

[Kurtnoise]

why are you posting this ? we are waiting for your patch instead...:rolleyes:

[stax76]

why are you posting this ? we are waiting for your patch instead...:rolleyes:

If it's still to difficult I consider making a patch. :rolleyes:

[foxyshadis]

I think the VapourSynth readme should really point out that the prefix is lsmas. Without it, there's no way to use the filter!

[LouieChuckyMerry]

First, many thanks for L-SMASH works :) . Second, would someone please help me understand why I can't index this (https://www.mediafire.com/?fuvbl9eb9h4sgdd) with LSW? I've tried demuxing-remuxing it as a .ts, .m2ts, .mp4, and a fresh .mkv, all of which result in flashbacks when viewing the preview window in MeGUI ;) . FFMS2 and DGAvcIndex index the file without any issues. Thanks for your time.

[Selur]

adding ,format="YUV420P8" might help with the preview, alternatively you could also add
f3kdb_dither(stacked=false, input_depth=10)

[LouieChuckyMerry]

Thanks for your help (and patience), Selur, and for not simply shouting "10 BIT!", not that I would've known what to do even if I'd noticed ;) . Both methods did the job (although the first was three-times faster), so now I've now a new trick for my bag :thanks: . In your opinion, is there a three-times improvement in quality? I can't see any difference, but my "monitor" is a 14" laptop screen.

[Selur]

Sorry, not sure now FFVideoSource does the dithering from 10bit to 8bit compared to f3kdb_dither.
I normally only add f3kdb_dither for preview purposes, for encoding I stick with 10bit. :)

Cu Selur

[foxyshadis]

Sorry, not sure now FFVideoSource does the dithering from 10bit to 8bit compared to f3kdb_dither.
I normally only add f3kdb_dither for preview purposes, for encoding I stick with 10bit. :)

Cu Selur

ffms2 & lsmash leave it default, where SWScale uses a very fast, basic fixed 8x8 pattern grid for >8-bit to 8-bit YUV. For processing & encoding that's good enough. It uses a nicer Floyd-Steinburg noise diffusion for all yuv->rgb.

[LigH]

I guess the fast pattern is "Bayer" dither?

[foxyshadis]

Exactly. They really should at least use DitherPost's method of creating some patterns at run-time, instead of a single compiled-in one, but ffmpeg does things their way.

[8-BaLL]

I was getting Access Violation trying to open a MKV file using LWLibavVideoSource. Looks like LSMASH cant handle MKV containers, right?

[LigH]

Not in general. If the content is not corrupt, and the container spec compliant, there is no reason to crash.

Try to re-multiplex this file with mkvmerge. Any error messages?

[8-BaLL]

Thanks for the info. In fact Ive just checked again and now it works :) Looks like I had some weird behaviour in AVSPmod. It would say "Access Violation" on MKV files no matter which one I selected. After selecting a .ts file, it seems to handle all MKVs again.

Again thanks for clarifying ;) LSMASH is the only source I use for HD stuff, pretty stable and precise. Not saying its not recommended for SD stuff, I just think DGdecode is more suited for mpeg2 stuff (DVDs).

[videoh]

I just think DGdecode is more suited for mpeg2 stuff (DVDs). To avoid anyone misinterpreting this, I point out that DGDecNV handles AVC, MPEG2, and VC1 in the same frame-accurate manner; there is no special sauce for MPEG2 streams. It handles HD perfectly fine. And finally, it is not buggy for MKVs as you said you experienced with LSMASH. In no way is it specialized for MPEG2 or DVDs.

I would be very surprised if LSMASH didn't also work perfectly fine for MPEG2 and DVDs. Did you experience any issues?

[LigH]

L-SMASH Works will probably not support a set of VOB segments as easily as DGMPGDec did. It will prefer one contiguous source file instead.

But as long as the source is not corrupt, it should handle all supported container and content formats really well.

[stax76]

Is there a technical reason why MP4 files can be opened directly without indexing but not MKV files?

I just think DGdecode is more suited for mpeg2 stuff (DVDs).

DG tools are solid but I think FFVideoSource will work just as well, with LWLibavVideoSource I had problems last time with a NTSC source, I'll try it again. StaxRip is x64 only so DGDecode isn't even a option.

[LigH]

L-SMASH Works contains two demultiplexers: libavformat and L-SMASH.

The L-SMASH demultiplexer supports only "ISO Media" containers and relies on media file internal index chunks. So it does not need to create an own, but fails if one is missing in the media file or is corrupt.

The libavformat demultiplexer does not rely on any media file internal index, therefore it supports also containers which do not even have an internal key frame index; but that means that it must create its own.

To support MKV containers while relying on their internal key frame index, it would have to support a third demultiplexer, probably.

[AMED]

I seem to be getting a weird grey color flash at the beginning of one of my encodes when using L-Smash and MVTools2. It doesn't seem to happen when using FFMS2.

# Set DAR in encoder to 16 : 9. The following line is for automatic signalling
global MeGUI_darx = 16
global MeGUI_dary = 9
LoadPlugin("D:\MEGUI\tools\lsmash\LSMASHSource.dll")
LWLibavVideoSource("H:\Above The Law (1988) (1)-001.mkv")
#crop
super = MSuper(pel=2, sharp=1)
backward_vec2 = MAnalyse(super, isb = true, delta = 2, overlap=4)
backward_vec1 = MAnalyse(super, isb = true, delta = 1, overlap=4)
forward_vec1 = MAnalyse(super, isb = false, delta = 1, overlap=4)
forward_vec2 = MAnalyse(super, isb = false, delta = 2, overlap=4)
MDegrain2(super, backward_vec1,forward_vec1,backward_vec2,forward_vec2,thSAD=400)

It only seems to occur when using MDegrain 2 or 3.

I've tested with Avisynth (2.6final), Avisynth+ (r1825), MVTool2 (v2.6.0.5) and L-Smash works (v785).

SAMPLE (http://www.mediafire.com/download/wsukekelrw1ag4r/Above_The_Law.zip)

[hello_hello]

As long as libavcodec doesn't contain a decoder for MagicYUV, L-SMASH Source may not support this video format. Even though I believe it would be possible to index an AVI regardless of the used codec; but why index a clip if the video codec is not supported in libavcodec, thus neither in ffmpeg nor in related projects like FFMS2 or L-SMASH Source.

I was looking for some info and found LigH's post above, but I was wondering.....

Why is it I can successfully index an MKV containing MagicYUV video with MeGUI and ffms2, but when I save the script and try to open it in MPC-HC, it displays an error about not being able to find a video stream. It displays normally in MeGUI's preview. Is it a splitter thing?

L-Smash kind of works the same way, except when indexing it creates the index file, then complains about not being able to find the video stream, but if I open MeGUI's Script Creator and load the index file as the input video it works just as ffms2 does. If I save the script and try to load it for encoding though, it's back to the "can't find the video" error.

A pic of MeGUI's Script Creator opening an MKV containing MagicYUV video via LWLibavVideoSource. In the background is the Script Creator preview happily displaying the video. If I save the script (keeping the script creator and preview open for the screenshot below) and try to load it for encoding, the result is the error message being displayed in the foreground.
I'm curious as to how that'd work.......

http://s8.postimg.org/kipmybx0x/magicyuv.jpg (http://postimg.org/image/kipmybx0x/)