Lansing raises an interesting point, which is...

[Katie Boundary]

The first problem is that by decimating, you're going to end up with a clip of variable frame rate. How do you determine when the frame rate change from 24fps to 30fps? How do you declare frame rate for sections that's a hybrid of 24fps+30fps contents?

Variable frame rates have been brought up a few times in conversations that I've started, and while I've always been curious about them, I never really bothered to ask any questions... but now Lansing has asked these questions for me, and I figure now is as good a time as any to get some answers. So here we go...

I have some (okay, a few hundred) DVDs that are a mix of film and hard NTSC. Exactly what's in the hard NTSC parts doesn't matter; it could be 30p stuff that got converted to 30i like in Babylon 5, or it could be natively 60i content like in Carmen Sandiego, or really anything in between, it doesn't matter. Let's assume that I'm interested in converting these DVDs to variable-framerate files of an arbitrary container format, switching between film and bob-deinterlaced 60-fps parts at the exact places where the DVDs switch from film to hard NTSC. How is this done?

[hello_hello]

Wow.... everything I said in your original thread on the subject of butchering DVD video, which you poo-poo'd at the time because you refused to listen. It seems like only yesterday you claimed it was all a waste of time due to lack of VCD compatibility. Remember when I mentioned the year, and instead of discussing the subject like a grown-up you posted silly pictures instead?

"While you've always been curious about it"? Comedy gold!

[lansing]

You just started a thread self proclaimed to provide solutions to fix the "problematic DVD" which were predominately consisted of hybrid contents, and now you start another thread asking how to fix hybrid contents?

http://forum.doom9.org/showthread.php?t=173445

[manono]

The easiest solution is just to bob the whole thing to 60p. I'll assume you're not reencoding for another DVD. There are VFR solutions, but I wouldn't bother. To do it, you'd have to go through all your sources looking for where it switches from one to the other.

By making them 60p they'll play the same as the DVDs on your hi-def television. Same fluidity, same jerkiness. I see no advantage to making a true VFR encode as they won't play any more smoothly either. You'll save some bits when making VFR encodes and everything for the same quality, but so what?

[Katie Boundary]

You just started a thread self proclaimed to provide solutions to fix the "problematic DVD"

No. The purpose of that thread is to provide a list of DVDs that do not respond well to IVTC, and the reasons why they don't. The suggested "solutions" are mere afterthoughts, and I'm strongly tempted to just strip those suggestions out so that people who don't understand English won't get confused and start completely irrelevant arguments.

and now you start another thread asking how to fix hybrid contents?

No. This thread is about converting variable-framerate DVD content to variable-framerate files of other arbitrary container formats. It is not about "fixing" anything.

By making them 60p they'll play the same as the DVDs on your hi-def television.

I never said anything about a television...

You'll save some bits when making VFR encodes and everything for the same quality

I'm mostly interested in a theoretical understanding and the option to screw around with new toys, and other forum members might have their own practical purposes for which VFR is a superior option. But thanks for thinking that these forums are all about me.

[manono]

I never said anything about a television...

Television, computer monitor, same thing, 60 Hz.

If you really want to learn how to encode for VFR, then have at it. There are several ways. The TIVTC - Common Usage Examples.txt shows some ways:

5.) NTSC Hybrid, using vfr via mkv (not anime or cartoon). One pass.

mpeg2source("c:\oursource.d2v")
tfm(d2v="c:\oursource.d2v")
tdecimate(mode=3,hybrid=2,vfrDec=0,mkvOut="mkv-timecodesfile.txt")


6.) NTSC Hybrid, using vfr via mkv, anime or cartoon source. One pass.

mpeg2source("c:\oursource.d2v")
tfm(d2v="c:\oursource.d2v")
tdecimate(mode=3,hybrid=2,vfrDec=1,mkvOut="mkv-timecodesfile.txt")


7.) NTSC, two pass (enables use of conCycleTP parameter) mkv vfr for hybrid source.

First pass:

mpeg2source("c:\oursource.d2v")
tfm(d2v="c:\oursource.d2v",output="matches.txt")
tdecimate(mode=4,output="metrics.txt")

Second pass (not anime or cartoon):

mpeg2source("c:\oursource.d2v")
tfm(d2v="c:\oursource.d2v",input="matches.txt")
tdecimate(mode=5,hybrid=2,vfrDec=0,input="metrics.txt",tfmIn="matches.txt",mkvOut="mkv-timecodesfile.txt")

Second pass (anime or cartoon):

mpeg2source("c:\oursource.d2v")
tfm(d2v="c:\oursource.d2v",input="matches.txt")
tdecimate(mode=5,hybrid=2,vfrDec=1,input="metrics.txt",tfmIn="matches.txt",mkvOut="mkv-timecodesfile.txt")

I'm sure others can make other suggestions.

[Katie Boundary]

Whoa. I had always assumed that, because AVI files don't properly support VFR, AVIsynth wouldn't either. This looks like a very promising path. Should I assume that the d2v files are generated with "Ignore pulldown flags"?

[filler56789]

Whoa. I had always assumed that, because AVI files don't properly support VFR, AVIsynth wouldn't either.

That's what you get for NOT reading the documentation...

[Katie Boundary]

I read the documentation... just not all of it:D

[manono]

Should I assume that the d2v files are generated with "Ignore pulldown flags"?
As I mentioned before, I have no use for VFR encoding. So, I'm guessing here but an educated guess says 'no'. Make it using 'Honor Pulldown Flags'. Maybe if someone knows for sure (or you experiment trying both ways), he/she can chime in.

[StainlessS]

Not sure but I think tfm() can be used with Ignore Pulldown flags, works its magic by reading the d2v file.

Either way, the only difference between Ignore and Honor are that the Honor mode sets "Field_Operation=0",
and Ignore mode "Field_Operation=2" '[EDIT: in the d2v.file], rest of file binary identical.

EDIT: start of honor.d2v

DGIndexProjectFile16
1
D:\NTSC_Top_FORCEFILM-ReturnOfTheKing_NTSC_FORCEFILM.VOB

Stream_Type=1
MPEG_Type=2
iDCT_Algorithm=6
YUVRGB_Scale=1
Luminance_Filter=0,0
Clipping=0,0,0,0
Aspect_Ratio=16:9
Picture_Size=720x480
Field_Operation=0
Frame_Rate=29970 (30000/1001)
Location=0,0,0,28000


EDIT: And end of d2v

FINISHED 100.00% FILM

[manono]

You may be right that in this case it doesn't make a difference. It makes a big difference under ordinary usage since using 'Ignore Pulldown Flags' is a recipe for out-of-synch audio unless the D2V shows 100.00% FILM or 100.00% VIDEO.

I was just trying to be cautious.

[Katie Boundary]

100.00% FILM or 100.00% VIDEO.

100.00% video actually shows up as just "NTSC". Things are only labeled "video" if the film content is 0.01-50% of the total.

</autism>

EDIT: after looking at those TFM scripts more closely, I don't think they'd achieve what I'm thinking of. For one thing, I don't see any bob-deinterlacing for the NTSC segments.

[StainlessS]

100.00% video actually shows up as just "NTSC". Things are only labeled "video" if the film content is 0.01-50% of the total.


Actually that may be true for the GUI window but not for d2v file.

From DGIndex:gui.cpp

if (D2V_Flag)
{
// Revised by Donald Graft to support IBBPIBBP...
WriteD2VLine(1);
fprintf(D2VFile, "\nFINISHED");
// Prevent divide by 0.
if (FILM_Purity+VIDEO_Purity == 0) VIDEO_Purity = 1;
film_percent = (FILM_Purity*100.0)/(FILM_Purity+VIDEO_Purity);
if (film_percent >= 50.0)
fprintf(D2VFile, " %.2f%% FILM\n", film_percent);
else
fprintf(D2VFile, " %.2f%% VIDEO\n", 100.0 - film_percent);
}


Shows 'FILM' or 'VIDEO' depends which is greater.

[manono]

From the bottom of a D2V:

900 5 3 670205952 0 1 20 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 5 3 670523392 0 1 20 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 5 3 670836736 0 1 20 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 5 3 671152128 0 1 20 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 5 3 671469568 0 1 20 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2 ff

FINISHED 100.00% VIDEO


EDIT: after looking at those TFM scripts more closely, I don't think they'd achieve what I'm thinking of. For one thing, I don't see any bob-deinterlacing for the NTSC segments.

Here's more information about how to create VFR encodes:

http://avisynth.nl/index.php/VFR

[Katie Boundary]

Actually that may be true for the GUI window

That's what I was talking about

From the bottom of a D2V:

900 5 3 670205952 0 1 20 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 5 3 670523392 0 1 20 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 5 3 670836736 0 1 20 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 5 3 671152128 0 1 20 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 5 3 671469568 0 1 20 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2 ff

FINISHED 100.00% VIDEO


Yeah, I was talking about what DGIndex itself shows you...

Here's more information about how to create VFR encodes:

http://avisynth.nl/index.php/VFR

THX, will investigate.

[Ghitulescu]

The easiest solution is just to bob the whole thing to 60p. I'll assume you're not reencoding for another DVD.

No, God forbid.

It's about VCD. :) :) :)

[hello_hello]

I'm strongly tempted to just strip those suggestions out so that people who don't understand English won't get confused and start completely irrelevant arguments.

Good idea, especially until you understand why the things you were arguing about aren't irrelevant.
Once the suggested "treatment" is removed, the whole thread can go, given you're probably the only person on the planet not using an appropriate tool for the job, such as TIVTC. Well.... you're testing it now, but only a few days ago it was bloated and it's processing was evil and UnComb was the ONLY tool.

If it wasn't so ironic, it'd be quite funny. Could I ask, has the English language changed since you argued about using TIVCT, or could it be the problem wasn't other people don't understand English, but rather it was your refusal to listen? At which point did you spot TIVTC's transformation to relevance? I'm curious, because not so long ago mentioning TIVTC was starting an "irrelevant" argument, until.... and hopefully my English is plain enough for you to understand.... until you were finally dragged kicking and screaming to the realisation it's not irrelevant after-all.

I'm mostly interested in a theoretical understanding and the option to screw around with new toys, and other forum members might have their own practical purposes for which VFR is a superior option. But thanks for thinking that these forums are all about me.

Wow.... the thanks manono gets for trying to help, bundled with the usual Katie attitude where gratitude should be. Maybe it's your inability to understand plain English again.

[hello_hello]

In case Katie is currently in ostrich mode again, could someone suggest (or has anyone suggested) she try Handbrake/Vidcoder in VFR mode?
That's the sort of thing it's designed for and it'd be much harder for Katie to change TIVTC settings and mess things up. There's not really anything to configure for Handbrake, aside from selecting the frame rate and frame rate mode. Although thinking about it, Handbrake doesn't have an mpeg1 encoder.....

In my limited testing, because Handbrake only de-interlaces the frames where combing is detected, it has a tendency to miss fine combing in just a small area of a frame, and if the video is interlaced, because Handbrake switches between no de-interlacing, blend deinterlcing and Yadif (depending on the combing detected), sometime there's a noticeable quality difference between frames if you look very closely (if for example, frame A wasn't de-interlaced because no combing was detected, and frame B was de-interlaced because there was). I'm not sure if Yadif de-interlaces a frame when no combing is detected, but I'm certain I've seen a quality difference between frames when de-interlacing with Handbrake and I'm pretty sure the de-interlacing caused it.

If Handbrake does leave residual combing behind now and then, which compared to Katie's current method would still be a massive improvement, the decomb filter might be tweak-able via it's "custom" setting in the GUI, but I'm not sure if that's a true story or what options are available. I'd research it myself, but I don't want to risk giving Katie the impression I think the forums are all about her, or however she thinks that works...

[johnmeyer]

I was not aware that DVDs supported variable frame rate. However, perhaps this term is being used in this thread in a unique way. Every definition I've ever seen of VFR is that it is video that can be any arbitrary frame rate at any moment in time. Streaming video is usually VFR.

A DVD is locked into a frame rate. For NTSC, it is either 29.97 or 23.976. You can add pulldown in order to accommodate material that was shot at some other frame rate, like 16mm amateur silent film, for example. Even when mixed with true 29.97 video, the DVD is still delivering only one frame rate.

So, the basic premise for this thread is based on a poor description of the problem. If what you are instead trying to do is deal with DVDs which have different forms of telecine, there are lots of solutions to that problem, and I suggest searching this forum for the answers.

[hello_hello]

I tend to think of NTSC DVDs as being a combination of different constant frame rates rather than being variable.

TIVTC has been mentioned for creating a VFR encode. I've never used it for that myself so I don't know exactly how it works, but my basic understanding of the way the Handbrake/Vidcoder decomb filter is that it can output something more closely resembling VFR, because it always drops any frames it detects as duplicates. You can force a constant frame rate output, but the decomb filter still works the same way. It drops any duplicated frames, so Handbrake then duplicates them again..... as required if a constant frame rate output is specified.

Handbrake's decomb filter has a Bob setting, which I think means it'll IVTC where required and de-interlace the interlaced parts to 60fps, dropping any duplicate frames along the way, and that may very well be the simplest method way to handle hybrid sources, or at least the most Katie proof method. I'm not 100% it works exactly that way but it seems logical and it's how I remember it. I don't use Handbrake myself though.

[Katie Boundary]

I was not aware that DVDs supported variable frame rate. However, perhaps this term is being used in this thread in a unique way. Every definition I've ever seen of VFR is that it is video that can be any arbitrary frame rate at any moment in time.

I've always interpreted it as "anything that isn't a constant frame rate". If a DVD switches haphazardly between 23.976 and 29.97 fps, I'd call that "not a constant frame rate" and therefore under the umbrella of "variable frame rate"

EDIT: the wiki agrees with my interpretation: https://en.wikipedia.org/wiki/Variable_frame_rate

[manono]

If a DVD switches haphazardly between 23.976 and 29.97 fps, I'd call that "not a constant frame rate"...
But it doesn't and it isn't. Every NTSC DVD ever made plays at a constant 59.94 fields per second. If you have a progressive scan DVD player and a progressive display, then the output is 'intercepted' and output at 59.94 frames per second.

What you might call the 'base' framerate (your 23.976fps and 29.97fps) within that 59.94 fields per second can be about anything and is padded out to 59.94 by repeating fields/frames (or by field blending).

[Katie Boundary]

Every NTSC DVD ever made plays at a constant 59.94 fields per second.

Field rate has nothing to do with frame rate, which is what we're discussing. Also, what it "plays" at is irrelevant when we're ripping and converting.

If you have a progressive scan DVD player and a progressive display, then the output is 'intercepted' and output at 59.94 frames per second.

That doesn't change the fact that film DVDs are encoded at 23.976 progressive frames per second.

Also, players and displays are completely irrelevant to a discussion about ripping and converting.

What you might call the 'base' framerate (your 23.976fps and 29.97fps) within that 59.94 fields per second can be about anything

Wrong. Last time I checked, DVD specs only permit 23.976 fps, 25 fps, and 29.97 fps.

You can fake lower or intermediate frame rates by just re-encoding the same frame over and over, or by rapidly switching between these three frame rates, but that doesn't change the actual frame rates.

[poisondeathray]

Field rate has nothing to do with frame rate, which is what we're discussing. Also, what it "plays" at is irrelevant when we're ripping and converting.

Wrong. Field rate has everything to do with frame rate if you want to discuss VFR, various pulldown patterns and cadences. Not all DVD's come "nicely" in perfect cadences. You can have bizarre patterns, and crazy stuff that some studios do for their releases


Wrong. Last time I checked, DVD specs only permit 29.976 fps, 25 fps, and 29.97 fps.

Wrong. (well 25.0 is ok for PAL DVD), but 29.976 is illegal , and 29.97 is an approximation. The actual NTSC frame rate is 30000/1001


You can fake lower or intermediate frame rates by just re-encoding the same frame over and over, or by rapidly switching between these three frame rates, but that doesn't change the actual frame rates.

Wrong. You need to talk about fields because of the repeat field flags (RFF) for DVD's . For example, that's what soft telecine uses. It's repeat FIELD flag, not repeat frame flag. For hard telecine / interlaced encoding it's the various arrangements of fields that you need to understand in order to get the underlying content frame rate

The bottom line for VFR discussion is content frame rate, that is what is really important. You need to distinguish between content frame rate, and how it's organized as fields for DVD. "Fields" wouldn't even be part of the discussion if it weren't for "DVD" - it would be a non issue. That's what manono was getting at (at least for NTSC DVD), the only reason everything is "packaged" into 60000/1001 fields per second is to make it compliant for NTSC DVD standards. You can have different underlying content frame rates, but various repeat FIELD combinations and cadences are used to fit them into that 60000/1001 fields/sec standard. And that's what VFR does - you want to get rid of all the repeat fields to recover the original content frame rate, so the various sections run at their native FRAME rate that they would have run at if not for DVD (and the 2nd common reason it's used is really bandwidth related - so you don't have to physically encode repeated frames. For example, some animations might have a base rate of only 12 or even 6 FPS in some sections which consist of encoded repeated fields on DVD. Timecodes are used to control the display time per frame so you encode fewer frames)

[manono]

Field rate has nothing to do with frame rate, which is what we're discussing.
Now you're just being silly in a failed attempt to prove your point. Okay, you don't like 59.94 fields per second? Another way of putting it is that all NTSC DVDs output interlaced 29.97fps (or 30000/1001 in pdr's more precise notation). But NTSC DVDs do really output a constant 59.94 fields per second. It's not a matter of interpretation. It's not a difference of opinion. It's a fact.
That doesn't change the fact that film DVDs are encoded at 29.376 progressive frames per second.
Well, most are, and they then have 3:2 pulldown applied to output 59.94 fields per second. But many films were made at different frame rates other than 24fps. Maybe you're not too familiar with silent films which may have been intended to play at framerates between 15 and 46 frames per second and which can be put on DVD in a number of ways. You can actually have 19.98fps stored on a DVD with 3:3 pulldown applied to output 59.94 fields per second. I know, I've done it. All perfectly legal. Animations are drawn and then put onto DVD at all kinds of framerates - 8fps, 12fps, 16fps, 24fps, 30 fps both progressive or interlaced. All are legal and this obsession of yours that anything not progressive 23.976fps with 3:2 pulldown is some sort of an abomination is ludicrous.

Wrong. Last time I checked, DVD specs only permit 29.976 fps, 25 fps, and 29.97 fps.

What? Maybe check again. Anyway, if, as you say, only 29.97fps is permitted for NTSC, isn't that a constant framerate?

You can fake lower or intermediate frame rates by just re-encoding the same frame over and over

By following that line of reasoning, for a progressive 23.976fps film, to get it to 29.97 every fifth frame should be a repeat of the one before. Which will make for a much more jerky-playing mess than what is really done - the application of 3:2 pulldown. So what is repeated is fields which is done in an attempt to smooth out the playback. It's explained in the article to which I'm about to link.

Have you ever come across this? It's worth reading four or five times until you understand it:

http://hometheaterhifi.com/volume_7_4/dvd-benchmark-part-5-progressive-10-2000.html

The middle part about the chipsets is less interesting, but the final section about what can go wrong with DVDs might be right up your alley, so-to-speak.

[Katie Boundary]

Oops, I made a typo. I'll go fix it before continuing further.

[Katie Boundary]

Wrong. Field rate has everything to do with frame rate if you want to discuss VFR, various pulldown patterns and cadences. Not all DVD's come "nicely" in perfect cadences. You can have bizarre patterns, and crazy stuff that some studios do for their releases

When you explain how field rate is in any way relevant to the definition of variable FRAME rate, and the fact that VOB files are often encoded in a way that qualifies as VFR, then you can continue with this insane gibberish.

Wrong. (well 25.0 is ok for PAL DVD), but 29.976 is illegal , and 29.97 is an approximation. The actual NTSC frame rate is 30000/1001

The typo has been fixed (I DID say 23.976 in a previous post, so you should have been able to recognize the typo as a typo and not a sincere belief that anything was encoded at 29.976 or 29.376... but whatever). Now address the actual point, please.

Wrong. You need to talk about fields because of the repeat field flags (RFF) for DVD's . For example, that's what soft telecine uses. It's repeat FIELD flag, not repeat frame flag.

So you admit that soft telecine does exist, and is encoded at 23.976 fps with flags. That's great. Do you ALSO acknowledge that this kind of content can exist interspersed with hard NTSC as part of the same video stream on a DVD?

For hard telecine / interlaced encoding it's the various arrangements of fields that you need to understand in order to get the underlying content frame rate

The frame rate of hard telecined content, or any kind of hard NTSC, is ALWAYS 29.97 fps. End of story, end of discussion. Even if it's the exact same frame repeated 29.97 times per second, that's still 29.97 frames per second; even if it's 59.94 distinguishable images per second, each of which contributes only a single field to the resulting stream, it's still 29.97 ugly combed frames per second.

The question is whether or not a DVD can contain a video file that switches between hard NTSC and soft telecine, and whether or not this counts as VFR.

The bottom line for VFR discussion is content frame rate, that is what is really important.

YES!

Finally, we're getting somewhere.

You need to distinguish between content frame rate, and how it's organized as fields for DVD.

Not for this particular chain of discussion.

You can have different underlying content frame rates, but various repeat FIELD combinations and cadences are used to fit them into that 60000/1001 fields/sec standard. And that's what VFR does - you want to get rid of all the repeat fields to recover the original content frame rate

Your syntax here is confusing. When you say "that's what VFR does", does the word "that" refer to the preceding sentence or the following one? Both are wrong but for different reasons. Both are also irrelevant to this particular branch of discussion.

Now you're just being silly in a failed attempt to prove your point. Okay, you don't like 59.94 fields per second? Another way of putting it is that all NTSC DVDs output interlaced 29.97fps (or 30000/1001 in pdr's more precise notation). But NTSC DVDs do really output a constant 59.94 fields per second.

What they output when hooked up to a DVD player and a TV set is NOT RELEVANT HERE. We're discussing how the discs are ENCODED.

Well, most are, and they then have 3:2 pulldown applied to output 59.94 fields per second. But many films were made at different frame rates other than 24fps. Maybe you're not too familiar with silent films which may have been intended to play at framerates between 15 and 46 frames per second and which can be put on DVD in a number of ways.

Irrelevant...

You can actually have 19.98fps stored on a DVD with 3:3 pulldown applied to output 59.94 fields per second.

What the...

(consults Internet)

...well, I did find evidence to support the idea that official DVD specifications only care about the final display rate, and are less picky about exactly how this is achieved. So, you are correct. However, this is all still irrelevant to whether or not DVDs can support VFR.

this obsession of yours that anything not progressive 23.976fps with 3:2 pulldown is some sort of an abomination is ludicrous.

WTF?

http://www.quickmeme.com/img/0e/0e26f4bf405225765f825ca28fac53759ff53a84a776a8572ee14fc50991d534.jpg

if, as you say, only 29.97fps is permitted for NTSC, isn't that a constant framerate?

I never said anything about what is or isn't permitted for NTSC. Congratulations on victory over the strawman you have constructed! Nobody is impressed.

By following that line of reasoning, for a progressive 23.976fps film, to get it to 29.97 every fifth frame should be a repeat of the one before. Which will make for a much more jerky-playing mess than what is really done - the application of 3:2 pulldown. So what is repeated is fields which is done in an attempt to smooth out the playback. It's explained in the article to which I'm about to link.

It doesn't really matter whether you're duplicating every fifth field or every fifth frame. The result is visually indistinguishable on playback. And that STILL has nothing to do with whether or not DVDs can support VFR!

Have you ever come across this? It's worth reading four or five times until you understand it:

http://hometheaterhifi.com/volume_7_4/dvd-benchmark-part-5-progressive-10-2000.html

I'll read it when I get the time

[poisondeathray]

When you explain how field rate is in any way relevant to the definition of variable FRAME rate, and the fact that VOB files are often encoded in a way that qualifies as VFR, then you can continue with this insane gibberish.


You're correct, sorry for the confusion - it's poorly worded - but it's the the field pattern which is important. The field rate will always be double the encoded the frame rate - but that is irrelvant to the discussion. What you really want is the actual content frame rate. You're just "undoing" what was done to make the original compliant with DVD. That's all. That's your the whole topic in a nutshell. All this other garbage you're spewing is extraneous



The typo has been fixed (I DID say 23.976 in a previous post, so you should have been able to recognize the typo as a typo and not a sincere belief that anything was encoded at 29.976 or 29.376... but whatever). Now address the actual point, please.


What point?


So you admit that soft telecine does exist, and is encoded at 23.976 fps with flags. That's great. Do you ALSO acknowledge that this kind of content can exist interspersed with hard NTSC as part of the same video stream on a DVD?

Yes it can, but not on the same title


The frame rate of hard telecined content, or any kind of hard NTSC, is ALWAYS 29.97 fps. End of story, end of discussion. Even if it's the exact same frame repeated 29.97 times per second, that's still 29.97 frames per second; even if it's 59.94 distinguishable images per second, each of which contributes only a single field to the resulting stream, it's still 29.97 ugly combed frames per second.


Yes. Absolutely right. But the content is what matters. If you had the same frame repeated for (let' s just say 29.97 times in 1 seconds for simplicity sake, the effective framerate is 1). So the VFR decimated encode would only use 1 frame instead . But, if it was an interlaced video section, that 29.97 frames per second section would actually have 59.94 different pictures per second. That effective frame rate is 59.94


The question is whether or not a DVD can contain a video file that switches between hard NTSC and soft telecine, and whether or not this counts as VFR.


It can, but not on the same title. Hard is more "flexible" and can hold more variety of patterns. So far you only seem to be concerned with 23.976, 29.97, 59.94 (or those approximations), and those are the most common but there can be others




Your syntax here is confusing. When you say "that's what VFR does", does the word "that" refer to the preceding sentence or the following one? Both are wrong but for different reasons. Both are also irrelevant to this particular branch of discussion.


It means that's what creating VFR videos based on timecodes and decimation does. ie. your original topic. When you make a VFR encode , you're just trying to "undo" the field patterns to make it compatible with DVD. It doesn't matter if it's hard or soft, those are "red herrings" . The bottom line is you want to get back the original content

There are different types of VFR; again - the one you should be concerned about is the actual content frame rate.

[poisondeathray]

...well, I did find evidence to support the idea that official DVD specifications only care about the final display rate, and are less picky about exactly how this is achieved. So, you are correct. However, this is all still irrelevant to whether or not DVDs can support VFR.


There are different types of "VFR". Maybe that's what is confusing you.

DVD's do not support varying timecode VFR. Ie. the type where you decimate and control display time per frame with timecodes

DVD's do support the type of VFR that has different underlying content framerate packed in 59.94 fields/sec to make it compatible with DVD standards. This is done through pulldown (either hard or soft) in various cadences. You want to "remove the pulldown" or basically undo it to get back the original. You have examples of this in your various sci fi DVD titles. The encoded frame rate of NTSC DVD is always CFR 30000/1001, and the field rate of NTSC DVD is always 60000/1001. But that says nothing of what the actual content frame rate is. Hope that makes it more clear

[manono]

It doesn't really matter whether you're duplicating every fifth field or every fifth frame. The result is visually indistinguishable on playback.
Nonsense. That's what 3:2 pulldown is all about - an attempt to have progressive 23.976fps video play as smoothly as possible within the 59.94 fields per second DVD output. The page to which I linked earlier discusses that in some detail, but you're not interested in learning anything. And that's also what the 120Hz and 240Hz televisions can do - have DVDs with progressive 23.976fps content and soft pulldown play absolutely smoothly, with none of the pulldown judder.
However, this is all still irrelevant to whether or not DVDs can support VFR.
And that's how they support it - with hard and/or soft pulldown. Or repeated frames. Or field blending. Or all of them. It's already been said several times that the 'base' framerate or content can be just about anything. Yet you insist your questions aren't being addressed?

By the way, if you want to make VFR MKV encodes, and have the original DVD's interlaced 29.97fps parts be played as 59.94fps, I think your best bet would be to bob the whole thing to 59.94fps to begin with, followed then by removing the duplicate frames to create a true VFR encode.

[hello_hello]

I tend to think of NTSC DVDs as being a combination of different constant frame rates rather than being variable.

I now feel the need to clarify, although chances are the only person who'd misinterpret it has me on her ignore list, but obviously NTSC DVDs are constant frame rate at 29.970fps, or probably more correctly they're 59.940 constant field rate.
I was referring to the original frame rate(s) of the content. Possibly not the best way to look at it in the context of this thread, but it seems that's the perspective my brain was using when I first learned about IVTC and de-interlacing etc.

The frame rate of hard telecined content, or any kind of hard NTSC, is ALWAYS 29.97 fps. End of story, end of discussion. Even if it's the exact same frame repeated 29.97 times per second, that's still 29.97 frames per second; even if it's 59.94 distinguishable images per second, each of which contributes only a single field to the resulting stream, it's still 29.97 ugly combed frames per second.

I know some people have no clue, but has Katie managed to somehow achieve negative clue?

A few weeks ago, back when Katie knew everything, soft telecined NTSC was "film" with every fourth frame repeated using pulldown flags, "real NTCS" was film with hard-coded telecining for 29.970fps, and interlaced video was somehow created by recording at 60p before a field is removed from each frame for 60i.
When Katie changed her "typo" today to include 23.976 in her list of legal NTSC frame rates, she demonstrated she still doesn't get it.

The typo has been fixed (I DID say 23.976 in a previous post, so you should have been able to recognize the typo as a typo and not a sincere belief that anything was encoded at 29.976 or 29.376... but whatever). Now address the actual point, please.

Maybe..... and only maybe..... some time in the future.... it'll occur to Katie if 23.976fps was a legal frame rate for DVD, it wouldn't have to be telecined.... hard or soft, and then maybe she'll realise if you can shoe-horn "film" into NTSC using 3:2 pulldown, you can shoe-horn other frame rates into NTSC using different pulldown patterns too. Although I doubt Katie could then work out it'd either mean any frame rate is legal for DVDs, or 23.976 can't really be a legal frame rate. I suspect a new reality will have to be created to accommodate that one, unless she can be dragged there kicking and screaming while she tells everyone else they're wrong.

[hello_hello]

Nonsense. That's what 3:2 pulldown is all about - an attempt to have progressive 23.976fps video play as smoothly as possible within the 59.94 fields per second DVD output. The page to which I linked earlier discusses that in some detail, but you're not interested in learning anything. And that's also what the 120Hz and 240Hz televisions can do - have DVDs with progressive 23.976fps content and soft pulldown play absolutely smoothly, with none of the pulldown judder.

I submitted my last post before seeing yours, so sorry for the repeats, although I attempted to explain pulldown vs repeated frames a couple of times in Katie's now deleted thread, but all I proved was Katie is a virtual black hole for any form of clue.

I explained why 23.976 viewed on a 60Hz display doesn't look anywhere as horrible as repeating every forth frame (as you know, for 23.976fps, the frames display for a duration of two screen refreshes, the three refreshes, then two ect, for a kind of "progressive pulldown" on a 60Hz display), but at the time Katie was still struggling with 3:2 pulldown not using repeated frames, and she retreated into baby mode labelling everything irrelevant and refusing to listen. That's what got me onto her ignore list, as apparently sustaining her next reality required the aid of the forum ignore function.

How on earth she can claim 3:2 pulldown looks the same as repeating every forth frame is beyond me. I live in PAL-Land and I'm not really bothered by 23.976fps running on a 60Hz display, but when every fourth frame is repeated it's fairly obvious, and quite annoying.

[filler56789]

Katie's now deleted thread,

:confused:

Anyway, the moderators can "undelete" a thread, IF they think the topic is worth reading...

[Sharc]

....How on earth she can claim 3:2 pulldown looks the same as repeating every forth frame is beyond me.......
For comparison I uploaded a .zip with 3 files (http://www.mediafire.com/download/9zr564oi9d7mb1v/Judder_Test.zip):
File 1 is 24fps film (well, 24 slowed down to 24/1.001)
File 2 is file 1 but 3:2 telecined to 29.97
File 3 is file 1 but with every 4th frame repeated

Watch the moving white block .....

[huhn]

of cause 2:2:2:4 is worse than 3:2.

but they are both terrible.

[Katie Boundary]

Okay, it seems like we're all getting back to being on the same page in terms of what we're talking about, even if we aren't quite there yet.

You're just "undoing" what was done to make the original compliant with DVD.

Or, in the case of soft pulldown, just not doing it in the first place.

That's all. That's your the whole topic in a nutshell. All this other garbage you're spewing is extraneous

Johnmeyer's blurb about DVDs not supporting VFR did lead us down quite a rabbit hole, didn't it...

What point?

The range of frame rates supported by DVDs. Which brings me back to something Mano said, which I need clarification about:

You can actually have 19.98fps stored on a DVD with 3:3 pulldown applied to output 59.94 fields per second. I know, I've done it. All perfectly legal.

Just to be absolutely, 100% clear... you're talking about soft 3:3 pulldown, right? I have to ask because if this sort of thing is possible, then I can see no reason why the Region 1 Reboot and War Planets DVDs couldn't have been encoded at 25 fps with soft 2:2:3:2:3 pulldown, but if you're talking about hard 3:3 pulldown, then you're completely not understanding the conversation at all.

Also, if it's soft pulldown, then what the hell would DGIndex report it as when previewing or saving as a project?

Yes it can, but not on the same title

Title? I've seen it done on the same chapter. And it's stupidly common.

Yes. Absolutely right. But the content is what matters....

Let's agree to disagree on that. Motion can sometimes be slow or subtle enough that "smart" software can mistake genuine 29.97 or 23.976 fps content for a much lower frame rate, and let's not even get into the nightmares that might happen when 23.976 and 29.97 fps content get composited together at 60i and then "smart" software starts deleting the "duplicates" that are actually supposed to be there...

It can, but not on the same title.

So when I rip a single chapter of Darkwing Duck, Andromeda, Captain Planet, Star Trek, Stargate SG-1, Gargoyles, Farscape, etc. to my hard drive and create a .d2v project file from it, and DGIndex reports that it's a mix of film and hard NTSC, DGIndex is lying to me?

It means that's what creating VFR videos based on timecodes and decimation does. ie. your original topic.

Original topic mentioned nothing about timecodes OR decimation.

When you make a VFR encode , you're just trying to "undo" the field patterns to make it compatible with DVD.

Now you're really not making sense. VFR encodes can be generated from a wide variety of sources, including Cinelerra.

If by "make a VFR encode" you mean "make a VFR encode from a DVD source", then we need to straighten out the "VFR within a single DVD title" issue before we can proceed on that.

There are different types of "VFR". Maybe that's what is confusing you.

No, it's what was confusing Johnmeyer. I was always very clear on this point.

DVD's do not support varying timecode VFR. Ie. the type where you decimate and control display time per frame with timecodes

I never said they did.

The encoded frame rate of NTSC DVD is always CFR 30000/1001, and the field rate of NTSC DVD is always 60000/1001. But that says nothing of what the actual content frame rate is. Hope that makes it more clear

"NTSC DVD" is not a relevant concept. The relevant concept is DVDs that do the following sorts of trickery:

1) The ones that are mostly film but will switch to NTSC for literally half of a second in order to do a 60i credit fade-in, switch back to film for as long as the text retains a constant brightness, switch to NTSC for another half-second for the 60i fade-out, and then switch back to film until it's time for the next 60i or 30p effect (Gargoyles, Kim Possible, any sci-fi show made between 1982 and 2002...);

2) the ones that are literally pure film except for the first 1 or 2 frames, which are solid black anyway so you wouldn't even know that they were hard NTSC unless DGIndex told you (Battlestar Galactica, Defiance, Gotham... okay, Battlestar Galactica is more like the first 2 seconds than the first 2 frames, but it's still all black...);

3) the ones that could have been done as mostly film, but are instead done as something like 90-98% hard NTSC with random film parts thrown in just to f*** with anyone trying to decode them (I'm talking very specifically about Darkwing Duck here, but Disney probably does it with a lot of their shows);

4) Star Trek: the Motion Picture, which, despite being made for theater screens, had its opening credits hard-telecined on the DVD release for no reason (the rest of the movie is soft pulldown)

I'm NOT talking about stuff like Babylon 5 and Birds of Prey, where the live-action content was shot at 24p and the effects were done at 30p or 60i but then every episode was encoded as 100% hard NTSC because somebody was lazy. Those are a whole different discussion, and if I ever want to convert those to variable-framerate MP4s, I'll start a different thread for them

And that's how they support it - with hard and/or soft pulldown. Or repeated frames. Or field blending. Or all of them.

Uh, no. None of those have ANYTHING to do with VFR. When I talk about DVDs that support real VFR, I'm talking about the ones that switch between soft pulldown and hard NTSC as often as the author wanted them to, wherever the author wanted them to, numerous examples of which I gave months ago and have kept repeating on a near-daily basis ever since then.

By the way, if you want to make VFR MKV encodes

I'd prefer MP4 for its industry-standardness, but right now, we really need to iron out this confusion about the existence of blending soft pulldown with hard NTSC in a single DVD chapter before we go any further. Either you're all wrong and I'm right, or DGIndex has an extremely serious bug and I'm the only person in the world who has noticed it, or I've been having an oddly specific hallucination for over a decade.

For comparison I uploaded a .zip with 3 files (http://www.mediafire.com/download/9zr564oi9d7mb1v/Judder_Test.zip):
File 1 is 24fps film (well, 24 slowed down to 24/1.001)
File 2 is file 1 but 3:2 telecined to 29.97
File 3 is file 1 but with every 4th frame repeated

Watch the moving white block .....

Yeah, sorry, can't tell the difference. Also, what the hell is this unholy ".ts" format?

[Katie Boundary]

By the way, Disney is probably the biggest bag of dicks on the planet when it comes to finding new and increasingly sadistic ways to give people like us a hard time. They're even worse than Paramount.

[poisondeathray]

Also, if it's soft pulldown, then what the hell would DGIndex report it as when previewing or saving as a project?

Title? I've seen it done on the same chapter. And it's stupidly common.



Yes, you're right, I'm going to eat my words. You can have just about anything - mixed progressive / soft flags and interlaced encoding /hard in the same section along with interrupted cadences. You're not supposed to - it's bad practice to mix, but it commonly occurs when bad editing is done.

But the treatment is the same. Whether it's hard or soft is irrelevant if your intent of this thread was how to produce a VFR video from a DVD . You need to identify the underlying content rate and treat it accordingly to create VFR video.


Now you're really not making sense. VFR encodes can be generated from a wide variety of sources, including Cinelerra.

If by "make a VFR encode" you mean "make a VFR encode from a DVD source", then we need to straighten out the "VFR within a single DVD title" issue before we can proceed on that.


Yes, of course, from DVD. You can have VFR from camera phones. That type of VFR is timecode driven. Not the "VFR in CFR" stream that you get with DVD.


You already know you can get VFR content within a single DVD title. You already have examples.




The relevant concept is DVDs that do the following sorts of trickery:

1) The ones that are mostly film but will switch to NTSC for literally half of a second in order to do a 60i credit fade-in, switch back to film for as long as the text retains a constant brightness, switch to NTSC for another half-second for the 60i fade-out, and then switch back to film until it's time for the next 60i or 30p effect (Gargoyles, Kim Possible, any sci-fi show made between 1982 and 2002...);

2) the ones that are literally pure film except for the first 1 or 2 frames, which are solid black anyway so you wouldn't even know that they were hard NTSC unless DGIndex told you (Battlestar Galactica, Defiance, Gotham... okay, Battlestar Galactica is more like the first 2 seconds than the first 2 frames, but it's still all black...);

3) the ones that could have been done as mostly film, but are instead done as something like 90-98% hard NTSC with random film parts thrown in just to f*** with anyone trying to decode them (I'm talking very specifically about Darkwing Duck here, but Disney probably does it with a lot of their shows);

4) Star Trek: the Motion Picture, which, despite being made for theater screens, had its opening credits hard-telecined on the DVD release for no reason (the rest of the movie is soft pulldown)

I'm NOT talking about stuff like Babylon 5 and Birds of Prey, where the live-action content was shot at 24p and the effects were done at 30p or 60i but then every episode was encoded as 100% hard NTSC because somebody was lazy. Those are a whole different discussion, and if I ever want to convert those to variable-framerate MP4s, I'll start a different thread for them


Ok, for 1-4 what are you trying to do ? Don't you just want to create VFR video, obtain the original content framerate? Or are this just curiosity ?

For something like a few black frames does it really matter ? Why are you obsessing over 99.98% when you can't even tell the difference between something like sharc's test ? You probably shouldn't care about creating VFR encodes from DVD either

Hard/soft doesn't matter - they are just different methods of putting the content into a 59.94 fields/sec stream. It's the content framerate that is important.

[wonkey_monkey]

It doesn't really matter whether you're duplicating every fifth field or every fifth frame. The result is visually indistinguishable on playback.

This is absolute nonsense.

The relevant concept is DVDs that do the following sorts of trickery:

1) The ones that are mostly film but will switch to NTSC for literally half of a second in order to do a 60i credit fade-in, switch back to film for as long as the text retains a constant brightness, switch to NTSC for another half-second for the 60i fade-out, and then switch back to film until it's time for the next 60i or 30p effect (Gargoyles, Kim Possible, any sci-fi show made between 1982 and 2002...);

What? DVDs do not do this. They do not switch back and forth. Every individual video stored on a DVD has a single base framerate. It does not "switch to NTSC" because the whole thing is already "NTSC" (if any of it is). I assume at least this has already sunk in, so what you're doing above is a really bad description of the reality.

4) Star Trek: the Motion Picture, which, despite being made for theater screens, had its opening credits hard-telecined on the DVD release for no reason (the rest of the movie is soft pulldown)

It cannot be a mix of hard telecine and soft pulldown, unless they are separate clips on the DVD in which case there will be a pause while it seeks. It seems far more likely that you're mistaken. Again.

I'm NOT talking about stuff like Babylon 5 and Birds of Prey, where the live-action content was shot at 24p and the effects were done at 30p or 60i but then every episode was encoded as 100% hard NTSC because somebody was lazy.

Not because it was lazy, but because it is the only way to encode mixed content to DVD! How should they have done it?

[manono]

What? DVDs do not do this. They do not switch back and forth. Every individual video stored on a DVD has a single base framerate.
When pdr made a similar statement earlier I resisted saying something, figuring I was misunderstanding him. He's backtracked since then. Now you've said the same and you're just as wrong. NTSC DVDs can and do switch back and forth all the time. It's perfectly legal. As long as it's 59.94 fields per second that's being output, they can do anything they want, use any base framerate they damn well please. Unless I'm misunderstanding you. :)
It does not "switch to NTSC" because the whole thing is already "NTSC" (if any of it is).
She's using NTSC in the same sense as does DGIndex - anything not progressive 23.976fps with 3:2 pulldown is NTSC. NTSC can mean hard telecined film, or field blended crap, as well as true interlace (shot using interlaced video cameras). And no, it can be all NTSC, none of it NTSC, or any percentage it wants NTSC. Her examples are valid. It would be nice if it were one or the other but, sadly, it isn't.

Her example 3) includes a common mistake by the great Criterion Collection. They have this nasty habit of dropping to video for a few frames at chapter points. It's so brief that the DVD player doesn't adjust to deinterlace (or IVTC it if it's a cadence-reading player) and so you get a few frames of interlacing and it's both jarring and annoying.

It cannot be a mix of hard telecine and soft pulldown,

It most certainly can.

Not because it was lazy, but because it is the only way to encode mixed content to DVD! How should they have done it?

They could have encoded the film parts as progressive 23.976fps with pulldown, and the video parts as they did. I'm not saying they should have done it that way - plenty of DVDs are done that way. But it wreaks havoc when played on a DVD player. Most DVD players are flag readers and will deinterlace it where if the film parts had been done differently, a viewer would get the full and better resolution in those parts.

I can show you examples of these things to back up my points, if you like.

[manono]

Just to be absolutely, 100% clear... you're talking about soft 3:3 pulldown, right?
Yep.
I have to ask because if this sort of thing is possible, then I can see no reason why the Region 1 Reboot and War Planets DVDs couldn't have been encoded at 25 fps with soft 2:2:3:2:3 pulldown,
It could have been, if the NTSC DVD was made from a progressive 25fps source. Any framerate between 19.98fps (3:3 pulldown) and 29.97fps (2:2 pulldown) can be encoded as progressive and put on an NTSC DVD with the appropriate pulldown flags. However, I have yet to see any kind of pulldown on a retail DVD other than 3:2 (or 2:3) or 2:2.
Also, if it's soft pulldown, then what the hell would DGIndex report it as when previewing or saving as a project?
Video Type: NTSC
Frame Type: Progressive

Here's the D2V from a short one I made. Notice no 0 1 2 3 0 1 2 3. Just 1 3 1 3 1 3:

DGIndexProjectFile16
1
G:\Buzdil\Menus\Menu.19.98.m2v

Stream_Type=0
MPEG_Type=2
iDCT_Algorithm=6
YUVRGB_Scale=1
Luminance_Filter=0,0
Clipping=0,0,0,0
Aspect_Ratio=4:3
Picture_Size=720x480
Field_Operation=0
Frame_Rate=29970 (30000/1001)
Location=0,0,0,2882

d00 5 0 0 0 0 0 d3 f1 e3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 71380 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 180812 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 320560 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 485928 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 652304 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 817812 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 986492 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 1152724 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 1320248 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 1485224 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 1647352 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 1810676 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
900 5 0 1974784 0 0 0 73 71 d3 f1 f3 e1 f3 f1 e3 f1 f3 e1
.
.
.
.
900 5 0 21144160 0 0 0 71 73 d1 f3 f1 e3 f1 f3 e1 e3
d00 5 0 21215952 0 0 0 d1 ff

FINISHED 100.00% VIDEO

[hello_hello]

For comparison I uploaded a .zip with 3 files (http://www.mediafire.com/download/9zr564oi9d7mb1v/Judder_Test.zip):
File 1 is 24fps film (well, 24 slowed down to 24/1.001)
File 2 is file 1 but 3:2 telecined to 29.97
File 3 is file 1 but with every 4th frame repeated

Watch the moving white block .....

If Katie can't spot the difference she should put away her DVDs and go back to video games.

#1 and #2 look the same to me. I assume #2 is being de-interlaced or IVTC'd on playback, but I'm not sure I can tell the difference.

#3 is obviously far less smooth and that's just movement in a small part of the frame. When it's a camera pan or there's far more movement it's horrendous.

I'm still not certain I can see an "judder" in the first two samples, and I'm connected to the TV at 60Hz. I think I'm predominantly seeing a "sample and hold" effect (I think that's the cause) due to the low frame rate and the high contrast between the moving image and the background. If there is "judder" then either I can't see it or it's being masked by the sample and hold effect.
I haven't tried using your samples, but I've done so enough to know if I remuxed your 23.976fps sample at 25fps and switched to 50Hz, it's still not going to look smooth. It would at 50fps/50Hz.
I have a sample somewhere I tested at both 23.976fps and 25fps a while ago. The camera is slowly panning across a dark object and the background is quite bright. The edge of the dark object "jitters" across the screen at 25fps/50Hz, and it may "judder and jitter" at 23.976fps/60hz, but it's a fair degree of horrible either way.

Would viewing this stuff on a Plasma make much difference? I'm not sure if there'd be a logical explanation for 3:2 "progressive" pulldown (23.976fps/60Hz) being harder to spot than it would using an LCD, but if I'm seeing it then maybe my brain isn't too offended by it for some reason.

[wonkey_monkey]

As long as it's 59.94 fields per second that's being output, they can do anything they want, use any base framerate they damn well please. Unless I'm misunderstanding you. :)

Even if that base framerate varies? So I could have a single DVD title which, for the first minute, had, encoded in the stream, 1500 frames at a base frame rate of 25fps (but played out at 59.94fps thanks to a pulldown flag), and then for the next minute has encoded in the stream 1798 frames at a base frame rate of 29.97fps?

How would such a hybrid clip rip, for example, to .mpg, and then subsequently how it would play in (for example) VLC? Badly?

NB: Katie, manono's post is a textbook example of the right way to correct someone (it also helps if you're right, of course).

[manono]

Even if that base framerate varies?
Yes, documentaries do it all the time. You might, in the same documentary, mix up progressive 23.976fps with pulldown, 23.976fps with hard telecine, 18fps 16 mm film sources (which will duplicate frames to make it possible to hard telecine), some PAL 25fps stuff (usually field blended), progressive 29.97fps and stuff shot on video (interlaced 29.97fps). I'm not saying I've ever seen all that in the same docu, but I've seen it in different docus. And you could do it all in the same docu, if your sources were as diverse as that.
How would such a hybrid clip rip, for example, to .mpg, and then subsequently how it would play in (for example) VLC? Badly?
If you're talking about 2 different framerate videos, both progressively encoded, they should play properly on both a hardware DVD player as well as any software DVD players. If they don't, then something's wrong with the player that it won't play a properly made DVD. And VLC player isn't really known as a very good DVD player but it should play a DVD made to those specs correctly.
So I could have a single DVD title which, for the first minute, had, encoded in the stream, 1500 frames at a base frame rate of 25fps (but played out at 59.94fps thanks to a pulldown flag), and then for the next minute has encoded in the stream 1798 frames at a base frame rate of 29.97fps?Yes. As I said, all it has to do is play at 59.94 fields per second and it's perfectly legal. I could even recreate that example quite easily and author it in Muxman. As you know, Muxman is very strict about keeping to the DVD specs.

[wonkey_monkey]

Yes, documentaries do it all the time. You might, in the same documentary, mix up progressive 23.976fps with pulldown, 23.976fps with hard telecine, 18fps 16 mm film sources (which will duplicate frames to make it possible to hard telecine), some PAL 25fps stuff (usually field blended), progressive 29.97fps and stuff shot on video (interlaced 29.97fps).

Yeah, but the ultimate output from an edit suite would be a single video with a single framerate, which you could encode to an AVI, MPEG, etc, and wouldn't require a VFR-capable container.

Katie's talk of DVD players switching between this and that sounded, to me, like she was talking about varying soft pulldown, not hard.

If you're talking about 2 different framerate videos

Nope, I'm talking about one video clip (or title on a DVD). That's where the confusion is, I guess. My point was that DVD players do not (or so I understood it) as Katie's post suggested, "switch back and forth." The 59.94i NTSC video which they display may have content which was originally some other framerate, but it no longer is.

Yes. As I said, all it has to do is play at 59.94 fields per second and it's perfectly legal. I could even recreate that example quite easily and author it in Muxman. As you know, Muxman is very strict about keeping to the DVD specs.

OK.. that more sounds like the kind of hybrid thing you had me imagining. I'd be interested to take a look, if you could do that.

[manono]

Nope, I'm talking about one video clip (or title on a DVD). Oh, sorry, me too. I apologize for not being more clear.
Yeah, but the ultimate output from an edit suite would be a single video with a single framerate, which you could encode to an AVI, MPEG, etc, and wouldn't require a VFR-capable container.
Right. If it's not for another DVD, the easiest and most obvious thing would be to bob it to 59.94fps and be done with it. She wants to learn how to make VFR videos so that solution won't work for her. I do think, though, she'll wind up bobbing her sources initially, before then converting them to VFR.
I'd be interested to take a look, if you could do that.
Sure, but it might be a while (several hours at least) as I'm busy at the moment.

[Sharc]

.....Yeah, sorry, can't tell the difference. Also, what the hell is this unholy ".ts" format?
.ts is a transport stream container format, similar to .m2ts for Blu-ray.
https://en.wikipedia.org/wiki/MPEG_transport_stream

When you can't see a difference between 2) or 3) against 1), then your player does IVTC, or you don't care about the combed frames or jerky playback.

[Sharc]

#1 and #2 look the same to me. I assume #2 is being de-interlaced or IVTC'd on playback, but I'm not sure I can tell the difference.
Your player does IVTC if #2 looks the same as #1. If it would jus deinterlace #2 you would see a jerky playback of the moving white bar. If you would watch #2 progressive you would see 2 combed frames out of 5.

[manono]

OK.. that more sounds like the kind of hybrid thing you had me imagining. I'd be interested to take a look, if you could do that.
I made one that's 25fps at the beginning and then switches to 29.97fps for the final 60% or so. I used the same 23.976fps source for both so the first part plays slightly too fast and the second part quite a bit too fast. You can easily tell the difference by comparing the ways the chorus girls 'flap' their arms - noticeably faster the second time than the first. Because I speeded up everything, the audio went out of synch so I didn't include it.

I'm also enclosing the D2V file. If you open it in a text reader (like Notepad), the opening part goes 0 1 2 2 3 0 1 2 2 3 before switching to 2 2 2 2 2. It's progressive all the way through, if you open a VOB in DGIndex and run the Preview on it. I zipped it because there were several DVD files, as well as the included D2V file.

https://mega.nz/#!vkomGC4Z!BkgZtAZg5WeIeAuDZDKpxFzDcZoS0GfmjfU7iuuFPZQ

If you can't find it when the download is complete (10.6 MB), check your desktop. That's where my Mega downloads end up.