Weird judder in encoded mp4?

[The.King]

Just to be sure, you should be running as administrator when you install avisynth.

Yeah, I normally install stuff when 'running as admin'. That's what I did both times when installing & re-installing it?

Look in C:\Windows\SysWow64. Do you see avisynth.dll there? Right click on it, select Properties, go to the Details tab. Does the information look right?

Everything looks fine - see photo.

I'm lost - have to say?

[poisondeathray]

In case your MPCHC wasn't compiled properly, try opening version.avs in virtualdub x86 as suggested earlier. Make sure the virtualdub plugins (or plugins32) folder for vdub is clean, or select ".avs" from files of type drop down box when opening the file (the point is to make sure no other input drivers are opening the avs)

[The.King]

Ok, here's one for the books - which (a) I never had to do before when installing Avisynth and (b) never expected it may be required?

I just uninstalled Avisynth again (2nd time), ran C-Cleaner again, cleared new reg again & rebooted (even though it wasn't required).

Here's the thing I never had to do before, because it never caused a problem. I gave this a shot out of desperation because I had tried everything else I know.

I temporarily disabled AVG, then ran Avisynth installer again (as Admin again) and then - re-activated AVG. Ran the very same script as before and 'bingo', everything suddenly worked?

What the hell is that all about? (I've always had AVG running every time I've installed Avisynth, updated it, or even tried installing the 64-bit Avisynth (successfully) a few weeks ago. The two never conflicted before?)

Anyway - it's working now, but seems very strange that AVG somehow prevented it or conflicted with it - without even giving me a splash-screen or warning that something was detected or wrong? The Avisynth installer 2.6 was downloaded from the Videohelp website, so hardly a bad copy and it's working now anyway, so was obviously ok?

Very odd.



Now I've run the usual script

MPEG2Source("C:\New Folder\Movie\VTS_01_2.d2v")
AssumeTFF()
SeparateFields()

to check a .vob source for structure and it appears to be what I think you guys call 3:2 pulldown? i.e. According to Neuron's wiki, my 23.976 NTSC .vob is going... aaa bb ccc dd eee ff. I also tried switching to AssumeBFF() but part of it looked worse, like a frame was going backwards?

All I need to do now, is figure out how to correct this without screwing it up. I know sRestore() is for Blended NTSC -> PAL. So is this where the IVTC thing comes in, or is there something about decimate()? I read a tonne of stuff (mostly in what looks like a foreign language), so now I'm as confused as ever? Any words of wisdom?

Thanks.

[jagabo]

Now I've run the usual script

MPEG2Source("C:\New Folder\Movie\VTS_01_2.d2v")
AssumeTFF()
SeparateFields()

to check a .vob source for structure and it appears to be what I think you guys call 3:2 pulldown? i.e. According to Neuron's wiki, my 23.976 NTSC .vob is going... aaa bb ccc dd eee ff.

Yes.

I also tried switching to AssumeBFF() but part of it looked worse, like a frame was going backwards?

Yes, that's what happens when you have the wrong field order.

All I need to do now, is figure out how to correct this without screwing it up. I know sRestore() is for Blended NTSC -> PAL. So is this where the IVTC thing comes in, or is there something about decimate()? I read a tonne of stuff (mostly in what looks like a foreign language), so now I'm as confused as ever? Any words of wisdom?

TFM() matches fields to restore the original progressive film frames. But it doesn't change the frame rate -- meaning you still have 29.97 fps with one duplicate frame in every group of 5, rather than 23.976 fps. TDecimate() selectively removes that duplicate frame. So you script should look like:

MPEG2Source("C:\New Folder\Movie\VTS_01_2.d2v")
AssumeTFF()
TFM()
TDecimate()

You may occasionally have to play around with the settings of TFM() and TDecimate().

[manono]

So is this where the IVTC thing comes in...

Yeah, the IVTC thing. A short 10-second sample might be nice as well, so we can determine whether or not it's hard or soft telecine.

Or open the D2V file you made using DGIndex , scroll down to the bottom and let us know the percentage of film or video.

[The.King]

MPEG2Source("C:\New Folder\Movie\VTS_01_2.d2v")
AssumeTFF()
TFM()
TDecimate()

You may occasionally have to play around with the settings of TFM() and TDecimate().

Thanks for this - and for explaining it too - it all helps. I'll try that script and see how it looks. If everything is ok, I assume this is something I can just load into Staxrip (optional command line input) and/or actually amend the .avs script that Staxrip generates automatically and then run it through VirtualDub to see how it looks before proceeding? I'll also read up a little on TFM & TDecimate tomorrow as I'd imagine that I'll need to find the plugins for these and install them into the Avisynth plugins folder in order for it to actually work. (i.e. this folder is still very bare looking - C:\Program Files (x86)\AviSynth\plugins

Yeah, the IVTC thing. A short 10-second sample might be nice as well, so we can determine whether or not it's hard or soft telecine.

No probs, can you just remind me which program I use to make a short 10sec clip from the .m2v file that DGIndex generated along with the .d2v & .ac3 file?

Or open the D2V file you made using DGIndex , scroll down to the bottom and let us know the percentage of film or video.

At the very bottom of the .d2v file is says: "FINISHED 100.00% FILM".

[The.King]

Ahhh nevermind, I remembered how to cut a short clip with DGIndex.

It's a mad .vob (Floyd) so pretty hard to find a good spot to cut a clip. Shortest I could get was a 7 sec clip, see attached.

Anywhere I have randomly picked throughout the full .m2v file, I'm still getting aaa bb ccc dd eee ff, etc.

Hope this helps...

[jagabo]

That clip is soft telecined. Each frame is encoded progressively and pulldown flags are added to tell the player how to produce 59.94 fields per second from the encoded 23.976 frames per second. If the entire video is like that you can produce a D2V file with DgIndex in Video -> Field Operation -> Force Film mode and skip the TFM().TDecimate().

By the way, when you do use TFM() you can pass it the D2V file for the most accurate processing:

TFM(d2v="D:\Downloads\VTS_01_1.demuxed.demuxed.d2v")

[The.King]

If the entire video is like that you can produce a D2V file with DgIndex in Video -> Field Operation -> Force Film mode and skip the TFM().TDecimate().

By the way, when you do use TFM() you can pass it the D2V file for the most accurate processing:

TFM(d2v="D:\Downloads\VTS_01_1.demuxed.demuxed.d2v")

Thanks for the tip, but I'm not sure that I follow you fully? Are you saying to open DGIndex, reload the original .vob's again, but then on the Video -> Field Operation tab, use 'Force Film' mode instead of 'Honor Pulldown Flags' and that by doing so, I skip the TFM() & TDecimate() process?

If so, what am I then left with (fps) in the resulting .m2v / .d2v files?

If not so, can I just use

TFM(d2v="D:\Downloads\VTS_01_1.demuxed.demuxed.d2v ")

to amend the .avs script that Staxrip generates automatically and then it will be ok?

[manono]

Are you saying to open DGIndex, reload the original .vob's again, but then on the Video -> Field Operation tab, use 'Force Film' mode instead of 'Honor Pulldown Flags' and that by doing so, I skip the TFM() & TDecimate() process?

Yes, that's one of the things he's saying.

If so, what am I then left with (fps) in the resulting .m2v / .d2v files?

You'll be left with an AviSynth script feeding a progressive 23.976fps video to open in VDub or other encoders. The field operations are explained in the DGIndex doc.

If not so, can I just use
TFM(d2v="D:\Downloads\VTS_01_1.demuxed.demuxed.d2v ")
to amend the .avs script that Staxrip generates automatically and then it will be ok?

I have no idea what StaxRip does, but that's the other thing he's saying - that you can keep the 29.97fps D2V (created using Honor Pulldown Flags) and allow TFM to use that D2V to guide it towards a progressive 23.976fps video. That part is explained in the TFM doc.

[The.King]

If so, what am I then left with (fps) in the resulting .m2v / .d2v files?

You'll be left with an AviSynth script feeding a progressive 23.976fps video to open in VDub or other encoders. The field operations are explained in the DGIndex doc.

Ehhhhh, 'an Avisynth script'? All I am left with is a .d2v, an .ac3 & an .m2v file. There's no script? And VirtualDub won't open any of them.

In the resulting .m2v (the one with Video -> Field Operation tab, use 'Force Film' mode) I am left with 29.970030 fps, not 23.976fps?

So if using the 'Force Film' mode avoided the need for TFM() & TDecimate(), then what else am I supposed to do to get rid of this extra frame jagabo refers to? I am assuming your last response was incorrect in that TDecimate() still needs to be used to remove the extra frame in every group of five? Right?

[The.King]

get rid of this - k-lite codec pack

Why?

[jagabo]

Ehhhhh, 'an Avisynth script'? All I am left with is a .d2v, an .ac3 & an .m2v file. There's no script?

If you use Force Film mode your script will be simply:

MPEG2Source("C:\New Folder\Movie\VTS_01_2.d2v")

The output of that script will be progressive frames at 23.976 fps.

So if using the 'Force Film' mode avoided the need for TFM() & TDecimate(), then what else am I supposed to do to get rid of this extra frame jagabo refers to?

In Force Film mode there are no extra frames. But this mode only works with videos that have been encoded with soft pulldown.

[manono]

I am assuming your last response was incorrect in that TDecimate() still needs to be used to remove the extra frame in every group of five?

You shouldn't assume things about which you have no knowledge or understanding.

Right?

Wrong. I already mentioned that this is all explained in the DGIndex doc. I can (sort of) understand you not wanting to read all 243 AviSynth docs but not to read something of vital importance to what you're trying to do? Okay, I'll quote it for you:

Force Film - This option is intended for the special case of film content encoded at 23.976 fps but pulled down for display at 29.97 fps, i.e., 3:2 pulldown. This option will restore the 23.976 fps film rate while keeping audio and video in sync throughout the clip. This option works by ignoring pulldown flags and inserting/removing frames to maintain a constant and sync'ed output stream. Do not use this option on non-3:2 pulled-down material, and, therefore, never use it on PAL or MPEG1. Also do not use this option on streams containing frame repeats.

"This option is intended for the special case of film content encoded at 23.976 fps but pulled down for display at 29.97 fps, i.e., 3:2 pulldown" refers to soft pulldown, as in your sample, or as in the 100% Film D2V you made from the VOB. "Also do not use this option on streams containing frame repeats" refers to hard pulldown where the telecine is encoded into the video resulting in 3 progressive and 2 interlaced frames in every 5-frame cycle (and the Video percentage will show as 100% at the bottom of the D2V), or to a mix of both hard and soft telecine (cases where the Film percentage is less than 100%).

[The.King]

If the entire video is like that you can produce a D2V file with DgIndex in Video -> Field Operation -> Force Film mode and skip the TFM().TDecimate().

1. This is exactly what I did (word for word, letter by letter), I opened DGIndex, loaded all original .vob's, applied 'Force Film' instead of 'pulldown flags'. The result was a dv2 file, an ac3 file & an m2v file which was 29.97fps? This is why I asked what else I needed to do in order to restore it to 23.97fps. I already referred to this and asked the precise question in post #99.

2. manono responded in post #100 to confirm.

3. In post #101 I again confirmed that

All I am left with is a .d2v, an .ac3 & an .m2v file. There's no script? And VirtualDub won't open any of them.

(referring to the files I had left from DGIndex and the fact that no Avisynth script was generated to be played in VirtualDub as manono said.)

If you use Force Film mode your script will be simply:

MPEG2Source("C:\New Folder\Movie\VTS_01_2.d2v")

I think you might be assuming that I am doing stuff in Avisynth or something? I only generate an .avs script initially to check the source file, field order, etc. to determine whether it is interlaced, progressive, 3:2, etc. I am doing what poisondeathray suggested, using my eyes to determine a source. Once I know the source, I am going back to Staxrip, as everyone here knows - I don't know enough about Avisynth to manually write scripts to convert/encode a file yet.

My main goal (in post #93) was to figure out what to do (in Staxrip) to process the source correctly now that I knew it was 3:2 pulldown. You then suggested a handy way to do this via DGIndex, but I was still left with 29.97fps and didn't know what to do then in Staxrip?

In Force Film mode there are no extra frames. But this mode only works with videos that have been encoded with soft pulldown.

All I can tell you is that DGIndex 'Force Film' mode of all .vob's gave me a 29.97fps .m2v?

You shouldn't assume things about which you have no knowledge or understanding.

Everyone else does! You just made a blatant statement based on your own assumptions: ".....you have no knowledge or understanding". The reality is - I have "little knowledge" about Avisynth, DGIndex and many other things I am only learning to use for the first time. You were in my shoes once, yet this is not the first time you've come across extremely condescending to me? I am well aware that assumption is the mother of all f*ck-ups, but I've spent plenty of posts in all manner of forums explaining myself over & over again because the people with the knowledge 'assumed' that I was doing something the way they would do it, when realistically speaking - I didn't have the knowledge to do it "their way" in the first place - hence why I posted. Not having the knowledge is precisely why I said to you "I am assuming your last response...." instead of saying to you; "I know your last response....." Twisting my words is not helping anyone. I give tonnes of help & advice in other forums on subjects which I have decades of extensive knowledge - but I never assume the newbies asking the Q's are even at 1% of my level and I don't put myself across like I'm speaking to a peer. I try to be helpful instead of judgemental and if I thought people should just read tonnes of stuff to get to my level, then would there be a forum to begin with?

Thanks for explaining the DGIndex 'Force Film' and I've now found the HTML you were talking about, so there's lots more reading to do which sucks. Again, as per above - all I can tell you is that I followed instructions to the letter and after loading all original .vobs into DGIndex and selecting 'Force Film' instead of 'honor pulldown', the resulting .m2v file I get is still 29.97fps, not 23.97 as you guys suggest it should be? That's why I asked (assumed) about TDecimate(), etc. because what you guys were saying should be the case - wasn't.

Thanks.

[manono]

I think you might be assuming that I am doing stuff in Avisynth or something?

That's the only reason to mess with DGIndex - to make a D2V file to use, together with MPEG2Source, in an AviSynth script for opening in VDub or another encoder or in a player.

All I can tell you is that DGIndex 'Force Film' mode of all .vob's gave me a 29.97fps .m2v?

It does nothing to the M2V except extract it from the Vob. That M2V still has the RFF and TFF flags in it and still plays at 29.97fps. To get it to play at 23.976fps (why?), you'd need to strip out the flags, easily done using DGPulldown. You don't even need to save the M2V. It's the D2V (and the audio) that is important. And that Force Film D2V when used in an AviSynth script opened in VDub or whatever gives you a progressive 23.976fps video when encoded. Why not try it?

You were in my shoes once, yet this is not the first time you've come across extremely condescending to me?

Only responding in kind. You've shown ignorance (combined with an unwillingness to put in the time you need to learn), arrogance and condescension in the face of a well thought of and extremely powerful frameserving and filtering program which, if you're serious about your hobby, you'd be wise to learn. I was in your shoes once, yes, and I read everything there was available on AviSynth - the documentation as well as forum threads - until I learned enough to get around in it. And I've never stopped learning.

...the resulting .m2v file I get is still 29.97fps, not 23.97 as you guys suggest it should be?

No one ever suggested any such thing.

[jagabo]

Once I know the source, I am going back to Staxrip

Sorry, I don't know anything about Staxrip. If it can read AVS scripts (I thought this was your intention) you can use the simple Mpeg2Source() script I gave. If not, DgIndex and Mpeg2Source() aren't going to help you -- other than as tools to examine your video's properties.

When you output an M2V file from DgIndex the Field Operation mode has no effect on the M2V file. The M2V file retains all the source's properties. Field Operation only effects the D2V file and Mpeg2Source() in AviSynth. And the D2V file references the original source file(s), not the demuxed M2V file.

[The.King]

Sorry, I don't know anything about Staxrip. If it can read AVS scripts you can use the simple Mpeg2Source() script I gave.

Brilliant - thanks for this. Yeah, as Staxrip analyses the source, it creates it's own .avs file. Same with when you apply any other settings such as crop, etc., so it's easy enough to load the source, apply the x264 settings I want and then amend the .avs script at the end - before beginning the process. In this case, I fed it the .d2v file as you suggested as when I checked the .avs script - it had the exact same content as the line you gave me.

When you output an M2V file from DgIndex the Field Operation mode has no effect on the M2V file. The M2V file retains all the source's properties. Field Operation only effects the D2V file and Mpeg2Source() in AviSynth. And the D2V file references the original source file(s), not the demuxed M2V file.

Thank you so much for explaining that. Now the three files I have from DGIndex makes perfect sense. You've really been a lot of help & I appreciate it - thanks.

[The.King]

Only responding in kind.

There you go again with your assumptions.

You've shown ignorance (combined with an unwillingness to put in the time you need to learn), arrogance and condescension in the face of a well thought of and extremely powerful frameserving and filtering program

Who died and made you God? Is everyone who comes to this forum looking for guidance supposed to be fully proficient in the field they are seeking advice on? If that's the case, would there even be a forum? Why even bother replying to the thread?

Ignorance? I've repeatedly said that I knew practically nothing about Avisynth. What part in the process did you miss that? What were you expecting, the antonym to ignorance?

Arrogance & condescension??? Pot --> kettle??? Suffice to say that none of the comments you have made throughout this thread have been helpful and/or useful in any respect. Some of the other comments you made to people (in other threads) have also been completely ambiguous and gone unnoticed too. You should reflect on that. Everyone else here didn't bring an ego to the table or read a tone that simply wasn't there and they actually all gave a lot of insight & good advice. As I stated before, I didn't specifically ask for your assistance at any stage in this process, you volunteered - and I think you should take whatever assistance you assumed you were offering here and apply it somewhere else, because it won't be missed.

[LMotlow]

Looks like you'd best stick with automated software that decides everything for you, and settle for that. I don't know why you've persisted for so long trying to pick up on something you don't like at all. Use what you want. Plenty of folks come here with little or no knowledge of anything, can't even copy a Windows file from one place to another. But most of 'em seem to learn something, while a few just give up and disappear. Every tech and creative forum follows that pattern.

This activity ain't cut out for everyone. Maybe it's not for you. That doesn't imply anything personally, it's just the way it is. All I get is that you came with one set of ideas in mind and wont let go of them. So get what you can from that.