Weird judder in encoded mp4?

[The.King]

Copy dgdecode.dll into the Avisynth plugins folder. It's in your Program Files under Avisynth (or Avisynth 2.5, or Avisynth 2.6, or however Avisynth was installed), in the "plugins" subfolder.

DGMPGDec was already there, that's where I had it installed.

You can load a vob into VirtualDub and do the same thing with the deinterlace and bob filters.

I just tried that, but without success? If I load the original .vob in VirtualDub, I get the error: MPEG Import filter: invalid pack at position 3, marker bit not set; possibly MPEG-2 stream?

Nobody shoots movies at 25fps except in PAL land.

Agreed, but PAL regions cover most of the habitable planet https://en.wikipedia.org/wiki/PAL_region which sucks.

(Donald Sutherland with a squeaky voice? No, please.)

That made me laugh!

Unwarranted assumption, not to mention incorrect. The mkv didn't play "fine" on my LED or my plasma, and the re-encode was stuttery.

Finally - someone who could see the judder/stutter like I could. The question is why?

I've followed further instructions here (good, bad or misleading - I don't know) and I've finally been able to write my very first short script [blow me over] and also sort out the .d2v file via DGIndex to it run through VirtualDub. (see, it may be a very slow & tedious process, but I'm learning!) What I get as I pan through it is; aa bb cc dd ee (as per the guide found here about source video http://web.archive.org/web/20140703141744/http://neuron2.net/faq.html) so it's obviously progressive. Yet when I try to encode it and do not de-interlace it (treat it as progressive), the resulting encode still has that judder when the camera pans? Is it just a bad source? Pity if it is, because it's nigh on impossible to find a good copy of that old movie.

I'm sorry to have alluded to quality issues in the earlier post. I realize the Q word freaks out people in this forum, so I'll try to contain myself. I usually stay away from these "I want my perfectly good DVD to fit on a 1.4MB floppy drive" threads.

That's not what this thread is about at all. I have 48TB of space rigged on 2 x tower raids, so space is not an issue for me and I will always prefer quality over space regardless. In saying that, I want to be efficient and don't need an encoded x264 file from a DVD source taking up 5GB like can be found online. I'd prefer to keep 5GB, 10GB, 15GB for high encodes of full 1080p with DTS soundtracks, etc.

I recall running into info about mod-8,mod-16 and encoders and decoders, etc., from one of the clique members here (jagabo, poisondeathray, or whoever). But you guys don't even follow your own rules, so why should I rummage for those posts?

All I can say about your earlier post (yesterday) on my mod-4 usage in Staxrip is - I was following instruction from other forum members from here a few years ago when I was told to lower to mod-4 (from mod-16) if I had a PAR issue following some minor cropping. I'm not sure what it even is or does, I was just told that's what people did and it cleared the problem with Width x Height.

@LMotlow - Can I ask your opinion on two things please?

(1)
After I have determined the source content as progressive, interlaced, etc. (by using the original .vob files to create a .d2v in this case) – when I’m loading the source into my encoding tool (a GUI like Staxrip, MeGUI or Hybrid – because I don’t know enough to encode via command lines), do you recommend inputting the original source vob’s (can be 3-8 files), or do you recommend creating a single .mkv file with all of the .vob’s via Makemkv to load as the source instead? I appreciate that you wouldn't touch a GUI and possibly do all of your own encoding via complex commands, but considering you have far more knowledge & experience than I do, it would be good to get your opinion.

i.e. Many forums have suggested loading an .mkv as the source in this case – instead of multiple .vob’s which can (supposedly) be unstable? Even one of the three GUI’s I mentioned, auto-suggest that I create an .mkv file from multiple .vob files as it was more stable? This was after I was trying to input 4 x .vob’s as the source. So when it comes to encoding old DVD’s and once I have confirmed the source content first in VirtualDub, which will give me the better / more stable result?

(2) Regarding the link above from Neuron on source video and the 'a b c d e' or 'aa bb cc dd ee', etc. to determine if a source is interlaced, progressive - I have another source but it's doing none of the things mentioned? I get 3 x frames which all slowly move forward and then 1 x frame which is stationary - then it repeats - 3 x frames forward and 1 x stationary? So it's like 'a b cc'?? It doesn't seem to be solely interlaced & it's not progressive, so is it Telecined or IVTC'd, Blended (whatever they are)?

This happens with the script:
LoadPlugin("C:\Program Files (x86)\AviSynth\plugins\dgdecode.dll")
MPEG2Source("C:\Desktop\New folder\movie.d2v")
AssumeTFF()
SeparateFields()

I was almost uploading the .d2v file for convenience - but then I thought 'what good would that be to anyone'? So I've extracted a short clip of a panning scene with a car to hopefully show what I mean. I did this with mkvtoolnix like yesterday, but if you would prefer me to extract the clip from the .vob file and still give you a smaller .vob or MPEG2 stream clip, then you'll need to tell me what to use and how to do it?

Thanks for your time & help, I appreciate it.

[The.King]

@blud7

Thanks for that, but most of it went straight over my head. I'm trying to learn about DGIndex, VirtualDub, writing .avs scripts, using Mkvtoolnix, etc. so I don't need to wreck my head with more stuff right now.

For the record, Mediainfo was reading the MPEG info straight from the .vob source as well as reading it from the .mkv later on. Furthermore, DGIndex also showed the content as 'interlaced' after it finished reading/indexing the original .vob file?

[LMotlow]

I was almost uploading the .d2v file for convenience - but then I thought 'what good would that be to anyone'. So I've extracted a short clip a panning scene with a car to hopefully show what I mean. I did this with mkvtoolnix like yesterday, but if you prefer me to extract a clip from the .vob file, you'll need to tell me what to use and how to do it?

The sample mkv has really bad stutter, was telecined video. Don't deinterlace. The original appears to have some form of 2:2 pulldown, which was pretty well borked by deinterlacing. No sense getting into all the other questions, it's more useful to see something unprocessed from the original and stay on track.

We don't need a big vob to see the problems. We can't use a d2v, which is just an index to the original video. It's a text file, as you'll see if you open it with Notepad. A while back an acquaintance and former member here posted a tutorial on how to make a sample cut with DGIndex: https://forum.videohelp.com/threads/359295-Record-without-interlacing?p=2272359&viewful...=1#post2272359. After you set up your edit points, save the sample with "Save project and demux video". As you saw earlier, you get a video-only stream (m2v) and AC3 audio-only file. We really need just the m2v, but if you want audio with it you can post the AC3 and the m2v both. They're easy enough to rejoin in Avisynth or VirtualDub.

Yep, people often post mkv or mp4 as demos to save space with lower-bitrate h264. If there are specific questions about interlace, telecine, etc. to MPG, sometimes an mpg is posted as sample work. It depends on the situation.

[jagabo]

Yet when I try to encode it and do not de-interlace it (treat it as progressive), the resulting encode still has that judder when the camera pans? Is it just a bad source?

There's nothing wrong with the source and nothing wrong with your encoding. The problem is in your playback device. You just need to figure out what it's limitations are and work within them. I'd start by trying different containers or a remux with MkvTooknix. If none of that help try encoding with x264 at the veryfast preset. That disables a lot of the advanced features that might cause problems. If that plays smoothly work your way up the presets until you find one that chokes you player.

One thing I know about the WDTV Live is that its read-ahead buffer isn't big enough. It's not very tolerant of network delays when playing off shared folders. It starts having problems with Windows shares at around 40 Mb/s, far below the 100 Mb/s ethernet connection. It does much better playing off a local USB drive.

[poisondeathray]

The source video is fine. If anyone is having problems in various programs like avisynth, just demux it to m2v from the mkv container with mkvextract, or ffmpeg.

There is no stutter anywhere - it might be the new version of MKV container that is giving older hardware problems. There are revised MKV specs. For example if your TV is older (if you're playing on the TV directly), or your WDTV is older (maybe older firmware), it might have problems playing it back . Maybe try MP4 container instead, or using an older mkvmerge version

Furthermore, DGIndex also showed the content as 'interlaced' after it finished reading/indexing the original .vob file?

DGIndex cannot determine "content" ; only your eyes can reliably determine actual content - hence the exercise to examine fields

[The.King]

A while back an acquaintance and former member here posted a tutorial on how to make a sample cut with DGIndex: https://forum.videohelp.com/threads/359295-Record-without-interlacing?p=2272359&viewful...=1#post2272359.

Perfect - thanks a lot for that link - it was the easiest instruction to follow that I've seen in a long time.

Please find attached the sample from the original .vob file - now in m2v.

Thanks for your help.

[The.King]

There's nothing wrong with the source and nothing wrong with your encoding.

I'm certainly not qualified in that field to agree or disagree. Some folk here say it's fine, others are saying it's messed up. All I know is what I can see when I play it and it's not great with the jerky movement every couple of frames.

The problem is in your playback device. One thing I know about the WDTV Live is that its read-ahead buffer isn't big enough. It's not very tolerant of network delays when playing off shared folders. It does much better playing off a local USB drive.

Thanks, but there's a lot of assumption in what you say. For a start, I've had my WDTV for about 3yrs now and I've played all manner of things through it - some of which my TivX couldn't cope with and a Popcorn Hour load of crap failed to even load. The WDTV has performed flawlessly and it has the latest firmware installed. Secondly, it is directly connected to my NAS which I can plug in/out of my network when I like depending on if I am transferring content across to the 6TB drive within it, so nothing is reliant on any network. The WDTV is then connected directly to my 75" LED. As I mentioned previously, it plays everything flawlessly - even very high encodes with massive bit rates and there is never any playback issue, delays, or problems with buffer.

Aside from this, I also stated in an earlier post that I copied the same problematic file to my laptop, connected the laptop directly to my LED TV via the HDMI out and played it using a variety of players on the laptop (vlc, MPC-HD, WMP, etc.). No matter what - the encode still look a little judder-ish - as it did on my WDTV, so this also lends to eliminating the WDTV as the issue. I got my second laptop from work (which is a monster piece of kit) and did the same thing. Again, more judder.

So I'll have to disagree with you on the WDTV being the problem because quite simply - it's not. There's either something up with the original source, or something weird happened during the encode - this is what I'm trying to get to the bottom of. I'm currently running the same source through Hybrid to see if it gives a different result than Staxrip. I know they all fundamentally use the same x264 codecs, but I figured it was worth a short. It's a good few hours away yet.

[jagabo]

That video is a field blended NTSC->PAL conversion. You need to do something like:

Mpeg2Source("D:\Downloads\Original .vob sample.demuxed.d2v", Info=3) 
Yadif(mode=1) # or QTGMC() or some other double frame rate smart bob
SRestore()

You'll need to download and install the Yadif and SRestore packages. The result will be a 23.976 fps progressive video.

[LMotlow]

OOOOOPS! Gotta edit this post already??? Jagabo beat me to it. Fast on the draw, that jagabo.
I used the same script, with QTGMC at "very fast"). Got the same result as jagabo. Attached is 23.976 original film speed.

[The.King]

The source video is fine. If anyone is having problems in various programs like avisynth, just demux it to m2v from the mkv container with mkvextract, or ffmpeg.

I made a sample cut with DGIndex of the original .vob file as per the link LMotlow provided. I'm just waiting for someone to check out the sample and (a) confirm whether it's telecined, blended, etc. and (b) how they figured it out (so I know how to do it myself if it happens again with another source.)

My TV is not even 2yrs old, the WDTV is only about 3yrs old and both laptops are between 2yrs-4yrs old. All have an issue playing the encode? When I say issue, I mean in a panning scene - I'm getting 3-4 secs of normal motion, then a split sec of judder, then 3-4 secs of normal motion, then a split sec of judder, and so on.

It's not the TV or the WDTV or the Laptops, so there's something we're all missing. I'd be dumbfounded if it all came down to the container on this occasion, but the encode that's currently running at the moment is an .mp4, so we will see.

DGIndex cannot determine "content" ; only your eyes can reliably determine actual content - hence the exercise to examine fields

Fair enough, but then it shouldn't really have a 'Frame type' listed in the analysis because it's just misleading. I know how to assess the source code for a .vob now, so I will be using my eyes from now on. But if someone could reply to post #28 with regard to knowing how to load other sources, that would also be helpful.

[LMotlow]

I made a sample cut with DGIndex of the original .vob file as per the link LMotlow provided. I'm just waiting for someone to check out the sample and (a) confirm whether it's telecined, blended, etc. and (b) how they figured it out (so I know how to do it myself if it happens again with another source.)

See jagabo's post and mine, above.

People come up with these creative pulldown and blend effects all the time. We've seen worse, many of them can't be fixed. You'll need Avisynth for this. You can use yadif instead of QTGMC, but the results won't be as sharp or clean. Yadif is not "bad", it's often used as a quickie smart bobber to test for effect. QTGMC can be mighty slow depending on settings.

[poisondeathray]

My TV is not even 2yrs old, the WDTV is only about 3yrs old and both laptops are between 2yrs-4yrs old. All have an issue playing the encode? When I say issue, I mean in a panning scene - I'm getting 3-4 secs of normal motion, then a split sec of judder, then 3-4 secs of normal motion, then a split sec of judder, and so on.

It's not the TV or the WDTV or the Laptops, so there's something we're all missing. I'd be dumbfounded if it all came down to the container on this occasion, but the encode that's currently running at the moment is an .mp4, so we will see.

It doesn't occur on jagabo's WDTV, or any of my other hardware or software. I don't know why, but it's definitely on your end.

You can go frame by frame and see that there is nothing wrong, encoding wise. And it plays fine in all software players

Initially you said it played fine on the PC:

The resulting encode looks fine on my PC. The resulting encode when played on my TV via a WDTV Live is also fine EXCEPT, every couple of seconds - about 3-4 secs, the picture jumps a little.

To me, that indicates either configuration or HW problems.




What about other sources, other videos ?

DGIndex cannot determine "content" ; only your eyes can reliably determine actual content - hence the exercise to examine fields

Fair enough, but then it shouldn't really have a 'Frame type' listed in the analysis because it's just misleading. I know how to assess the source code for a .vob now, so I will be using my eyes from now on. But if someone could reply to post #28 with regard to knowing how to load other sources, that would also be helpful.

The frame type listed is the type of encoding. You can have progressive content encoded interlaced as you did there, or vice versa. (or the other combinations, progressive encoded progressive, interlaced encoded interlaced. Again, those are different things. Content vs. type of encoding.

[The.King]

[QUOTE=LMotlow;2409009]See jagabo's post and mine, above.

My sincere thanks to you both and for uploading the short clip of the restored copy. I transferred it played it on my WDTV - played perfectly.

People come up with these creative pulldown and blend effects all the time. We've seen worse, many of them can't be fixed.

Dare I even ask how you guys determined that this was a weird NTSC->PAL blend, or were you saving my head from exploding by not saying? I'm just curious in case I come across it again. It happened to me years ago with another old DVD and I never knew how to determine that blended source. Now the same thing has happened, so I'd love to know how to determine when a source is blended like this?

You'll need Avisynth for this. You can use yadif instead of QTGMC, but the results won't be as sharp or clean. Yadif is not "bad", it's often used as a quickie smart bobber to test for effect. QTGMC can be mighty slow depending on settings.

Yeah I've always found QTGMC to be quite slow, but it does give a better result. Now I've just got to figure out if any of the GUI's can do this, considering I'm useless at command lines. I recall running a blended source through Hybrid before and I distinctly remember something about Restore(), so hopefully I can figure that out.

[LMotlow]

I saw a little stutter in the m2v, bust mostly a lot of line twitter before QTGMC et el. Not really satisfied with that sRestore, because adding pulldown back into it gives some hard stutter on playback at 25fps. I suspect some odd blended/pulldown pattern in there somewhere, but jagabo and poisondeathray and manono are better at those odd patterns than I am.,

I know how to assess the source code for a .vob now, so I will be using my eyes from now on. But if someone could reply to post #28 with regard to knowing how to load other sources, that would also be helpful.

Are you talking about "loading" in Avisynth to examine frames? DGDecode is almost universal for MPEG. There are other openers for MTS, etc., in this article: http://avisynth.nl/index.php/FAQ_loading_clips. SOme folks use AviDemux to cut samples from h264, I often use VideoReDo or TMPGEnc Smart Renderer, neither of which are free.

[The.King]

It doesn't occur on jagabo's WDTV, or any of my other hardware or software. I don't know why, but it's definitely on your end.

But it was unstable on LMotlow LED & Plasma?

Initially you said it played fine on the PC:

I did. But very shortly afterwards I said that on closer inspection, I could see the judder on my PC too. It was also evident on all laptops. While I'm very much a novice when it comes to command lines of encoding and the more technical stuff, I'm well adept when it comes to most things AV. One PC monitor with lower Hz rates and a lack of technology will never display artefacts like a 2000Hz UHD LED can.

What about other sources, other videos ?

I have encoded several other old DVD's lately (original .vob files -> .mkv file -> GUI -> adjusted settings -> encoded with either Staxrip or Hybrid) without any issue. They all played fine on my WDTV & LED. It's obviously just these weird 'Blended NTSC -> PAL' conversions that messed things up for me and I've lost two days to it. Hopefully someone can explain how to determine when a source is blended in this way so I'll know how to confirm it and resolve it in the future?

Thanks.

[LMotlow]

PC media players, external players, and TVs don't operate the same way. I saw some small amount of stutter in all the videos on my PC, but not much. Mounted on a USB drive and played in my BluRays, the stutter was worse. But that was a matter of the way the samples were deinterlaced and whatnot, not to mention the sloppy way they seem to be created -- which ain't unusual nowadays.

I'm not all that great at figuring out some of the more weird patterns, but playing that m2v original it seemed there was no pattern, period. Looked like originally telecined, then frame blended with telecine intact to slow it down to 25fps. That led to a semi-educated guess that the video used blended frames. I usually follow jagabo's lead with that kind of oddball work, he's better at it than most of us around here. He beat me to the post this time (no surprise) but he was right as usual.

[poisondeathray]

To see the field blends, you use separatefields() as you have been doing. You will notice a double image on some fields, instead of a clean distinct image. That is, by definition, "field blending"

For his source and encoded mkv - you should see some stutter in North America 60Hz displays. 25 isn't evenly divisible into 60. 25 is evenly divisible into 50, so he shouldn't see any

[The.King]

I know how to assess the source code for a .vob now, so I will be using my eyes from now on. But if someone could reply to post #28 with regard to knowing how to load other sources, that would also be helpful.

Are you talking about "loading" in Avisynth to examine frames?

I think so? I was kinda referring to the first & second lines in a new .avs script. Assuming that whatever xxxxxx.dll I need to tell it to load will depend on the content/container of my source file?

i.e. Like for .vob/MPEG-2 sources, I can use DGIndex to create a .d2v file, then write the necessary script beginning with...
LoadPlugin("C:\Program Files\DGMPGDec\dgdecode.dll")
MPEG2Source("C:\Desktop\(file).d2v")
and then just load it in VirtualDub to see the frames, etc. and determine the content.

So if I had a different source (pick any file type/extension), I assume it wouldn't be compatible with DGIndex, so; (1) what would I use to open it, (2) am I still trying to create a .d2v file regardless of the source type, (3) That will determine what the first & second lines of a new .avs script would be to load the right plugin and set the right source and (4) do I still use VirtualDub to view the script?

That's what I meant.

Thanks.

[jagabo]

Dare I even ask how you guys determined that this was a weird NTSC->PAL blend

Start with:

Mpeg2Source()
Bob() # or SeparateFields()

Step through frame by frame. You'll see blended frames about every few frames.

By the way, I would have uploaded a sample earlier but I can't upload images or videos to this site any more. I've tried from three different computers. My account must have uploads locked out.

[The.King]

By the way, I would have uploaded a sample earlier but I can't upload images or videos to this site any more. I've tried from three different computers. My account must have uploads locked out.

If you and all the other kind folk who have helped here could just upload your brain & ability to write commands and understand the inner workings of encoding video - that would be great!

The command lines you guys write look like Putonghua dialect to me. I can get as far as being able to see a .d2v file in VirtualDub, but that's about it. As for integrating your command lines into a GUI - I know there are options, but I don't think I can just copy/paste and leave everything else out?

Hybrid is pretty thorough, so I'll give that a shot, otherwise the original .vobs are going in the bin because I'm not losing another day to trying to successful encode a rare DVD.

Thanks.

[LMotlow]

Thyanks to jagabo for excellent analysis methods. I would never have guess the tricks shown with TIVTC. That's a keeper.

Almost every Avisynth filter or plugin has several parameters, all of it explained in the docs that come with the filters. Even simple ole yadif has a few values like field order and output mode. Many people just use the defaults, for example the default QTGMC statement is "QTGMC()" . That runs at default settings, explained in the html and the avsi script itself (the default settings for QTGMC()" are the same as saying "QTGMC(preset="slow"), and it is slow!). Mainly you learn this stuff by watching others. Nobody comes here knowing this jargon from birth. In a short time you'll see that you use pretty much the same batch of filters for almost everything, so you don't need to get into 200 different plugins. If you run into a video type that you can't open or if it's confusing, ask here.

The link to the Avisynth FAQ for opening different video types explains functions and filters for different video and audio types, like FFMS2, DirectShowSource, LSMASH, etc. Except for DGDecode, these openers are decoded directly in Avisynth without the need for anything like a d2v project file.

Most people here have some sort of text file or file(s) with pre-typed command lines that they just cut and paste into a script. We also save bunches of sample scripts and methods instead of trying to reinvent the wheel with every project.

Automated converters are a crap shoot. Most of them use limited functions, some of which are not right for the job, and they offer too little control over particulars. And most of them make it tougher to learn more about video innards.

[manono]

Thanks, but that's the info on DGMPGDec. I was referring to an overall definitive guide for writing commands within an .avs file - i.e. if my source was .mp4, was .mkv, was M2TS, was AVI, etc.

Your immediate problem was getting your DVD to open in VDub using MPEG2Source and that's what those three docs explain clearly and completely.

Many forums have suggested loading an .mkv as the source in this case – instead of multiple .vob’s which can (supposedly) be unstable?

Repackaging the VOBs as MKV is just inviting trouble, as I think you've discovered.

[LMotlow]

as manono says (beat me to the post again. I need more coffee), the docs that come with these plugins explain much. The link at http://avisynth.nl/index.php/FAQ_loading_clips#How_do_I_load_MP4.2FMKV.2FM2TS.2FEVO_into_AviSynth.3F is a general guide to several openers. I see that the guide still mentions DGAVCindex for certain formats, but LSMASH has superceded it. I still use DGAVCindex myself sometimes. I haven't quite come up to speed with LSMASH yet. Some of it looks like that Putonghua dialect you mentioned.

[QUOTE=The.King;2409033]

i.e. Like for .vob/MPEG-2 sources, I can use DGIndex to create a .d2v file, then write the necessary script beginning with...
LoadPlugin("C:\Program Files\DGMPGDec\dgdecode.dll")
MPEG2Source("C:\Desktop\(file).d2v")
and then just load it in VirtualDub to see the frames, etc. and determine the content.

If you copy DGDecode.dll to your Avisynth plugins folder, you won't need that LoadPlugin line. Dll filters in the plugin folder load automatically.

You mentioned opening "AVI" files. Avi is a container, not a codec, it can contain many different encoded formats: lossless decodes with huffyuv or Lagarith or others, DV-AVI, XVid, Panasonic DV, etc. Most of the time you can use "AviSource("path to the video\video.avi"), assuming you have the codec\decoder installed in your system.

[blud7]

The judder occurs on my PC with the graphics card options off as well as an older TV. Turning the options on make sit go away. I see you ignored my post. I have had problems with MakeMKV doing this before.
I encoded your sample with ffmpeg, no deinterlacing and it played fine.
Sorry if my suggestion isn't as technical as the rest but it works.

[The.King]

Almost every Avisynth filter or plugin has several parameters, all of it explained in the docs that come with the filters.

Therein lies part of the problem. Reading takes way too long (generally speaking) - especially when there is so much to read about technical stuff and so little time in the day. (It's not like I'm chillin' on vacation with a good novel.) To use a slant on a well-known saying; I've always been the 'show me how to fish and I'll fish forever' kind of guy. Not the 'throw a book about fishing at me and I might learn how to' kind of guy. I've always been baffled that there are no reliable videos on YouTube about this sort of stuff when so many people are trying to learn? Someone could make a lot of money...

Nobody comes here knowing this jargon from birth.

A flaw in our evolution & DNA, me thinks.

The link to the Avisynth FAQ for opening different video types explains functions and filters for different video and audio types, like FFMS2, DirectShowSource, LSMASH, etc. Except for DGDecode, these openers are decoded directly in Avisynth without the need for anything like a d2v project file.

Excellent, thanks for that.

Automated converters are a crap shoot. Most of them use limited functions, some of which are not right for the job, and they offer too little control over particulars.

So true. But without reading a bible (on fishing) and with no comprehensive guides out there to say (for example); "ok, here's an interlaced .vob that I'm going to encode perfectly with x264 into an .mkv container. Want to know how to do it? Let's get started: 1, Using Google (or your preferred search engine) download the following programs[list them all]. 2, Ensure all programs are installed in this XXXXX directory, for example 'C:\Program Files\xxxxxxxxxxxx’. 3, Open the program DGIndex which can be found in the folder xxxxxx. 4, Open or load the .vob file........................... 5, .......

You get the point I'm sure and you might even agree that doing something like this as a 'one off' would end up being less time consuming that responding to a million questions from people about all manner of things? No, the guide won't work for every file source or for every encode - but it would be a damn sight better than anything else available out there and it would give thousands of folk like me the opportunity to see the process from start to finish. Anyone with a bit of savvy after that, should be able to learn & adapt, or else ask a Q here as you suggested. That's how you teach people how to (fish), not by throwing the book at them. And that's why (a) there are tonnes of GUI's out there and (b) there are even more people using them and staying away from (what looks like) highly complicated command line encoding. If someone did a screengrab video of the one-off tutorial process and put it on the likes of YouTube, I bet anything that it would have a tonne of hits.

[The.King]

Repackaging the VOBs as MKV is just inviting trouble, as I think you've discovered.

The main issue wasn't repackaging the .vob to an .mkv, it was doing so and assuming that both were interlace TFF as many programs suggested, when it turned out - it was a weird NTSC -> PAL blend. Having learned how to use DGIndex & VirtualDub, I know how to check for myself now so that I'm not de-interlacing something which isn't interlaced.

I'd also guess that all you guys use command line encoding (considering you're all well-versed on encoding) rather than GUIs. From that perspective, using the original .vobs might well be the only way to go, but for us lesser folk - loading a number of .vob files into a GUI can cause problems - least of which is the encoding time.

Example A: Movie loaded into GUI using 5 x original .vob's, settings applied and encoded - took 12hr:27mins.
Example B: Movie loaded into GUI using 1 x .mkv (made from original .vob's using Makemkv), settings applied and encoded - took 2hr:11mins.

The results of both were identical - the .vob encode file had a little bit more grain which (although minimal) actually made it look worse.

[The.King]

Some of it looks like that Putonghua dialect you mentioned.

Some?

If you copy DGDecode.dll to your Avisynth plugins folder, you won't need that LoadPlugin line. Dll filters in the plugin folder load automatically.

Thanks for that tip.

You mentioned opening "AVI" files. Avi is a container, not a codec, it can contain many different encoded formats: lossless decodes with huffyuv or Lagarith or others, DV-AVI, XVid, Panasonic DV, etc. Most of the time you can use "AviSource("path to the video\video.avi"), assuming you have the codec\decoder installed in your system.

Thanks again. Yeah I know my containers from my codecs - my head was fried by the time I posted that yesterday.

[The.King]

I see you ignored my post. I encoded your sample with ffmpeg, no deinterlacing and it played fine.
Sorry if my suggestion isn't as technical as the rest but it works.

For a start I didn't ignore your post. I responded way back in post #32 or something, you're just a little late to the party or didn't bother to check?

Second; while I've seen ffmpeg (and probably used it in an automated way within a GUI) a billion times, I have no idea what it is or how to use it. From clicking on your link, I see it's a command line tool to convert one video file format to another. I'm not sure what part of my posts where I've repeatedly said that "I don't know enough about encoding to get my head around the technicalities of command lines" (which must be a million times now) that people don't understand? I was trying to learn about DGIndex, Mkvtoolnix, VirtualDub, creating a .d2v and writing my very first script at the time, which might take a few minutes for you guys to do now - but once upon a time it also took you a day or two to figure out. I lost two days in the process between reading a million tutorials & trying to solve a problem with the help of the good folk here, so I wasn't particularly excited to try your route with another "command line tool".

[jagabo]

The main issue wasn't repackaging the .vob to an .mkv, it was doing so and assuming that both were interlace TFF as many programs suggested, when it turned out - it was a weird NTSC -> PAL blend.

That wasn't the case for the two videos in post #6.

[The.King]

That wasn't the case for the two videos in post #6.

Yeah, but it was as the post evolved and the more recent sample was uploaded which is what I was referring to before he responded.