IVTC - Force Film When No Dupe Frames?

[onesikgypo]

Hi,

I'm relatively knew to IVTC so please bare with me.

I ran a .vob through DGIndex for 10mins and it came out with:

Video Type: FILM 98% and Frame Type: Progressive and frame rate is: 29.970fps

Now in most of the guides i read - it says that if it shows FILM 95%+ then i should use force film to ivtc to 23.970

However, when i opened the .vob in vdubmod and went thru a scene frame by frame - there didnt seem to be any duplicated frames.

Is this normal for this type of video (film 95%+) or does this mean no ivtc is needed or ? Also im assuming no formal deinterlace will be required.

Thanks.

Edit:

Further Information - Frames to be as following as per vdubmod:

BBPBBPBBPBBIBBPBBPBBPBBI

One further question:

Even if this was a legitimate force film case - is it proper/ok to NOT use it, but then use telecide in the avs script - as with a hard interlaced video?

[manono]

However, when i opened the .vob in vdubmod and went thru a scene frame by frame - there didnt seem to be any duplicated frames.

Is this normal for this type of video (film 95%+)

VDubMod ignores the TFF and RFF flags so yes, it's normal for you to see only the 23.976fps encoded video. With the percentage that high the usual advice is to Force Film it. However, that's no guarantee that there won't be any hard telecine in the main video, although it's usually the black frames at the beginning and end that might be hard telecined. You can have a look at the D2V and if you see the 2s and/or 0s only at the beginning and/or end, you're most likely OK. It's not uncommon, though, to find them at chapter breaks sometimes, and then you might have problems.

Even if this was a legitimate force film case - is it proper/ok to NOT use it, but then use telecide in the avs script - as with a hard interlaced video?

Is it proper? Not really, because then you open yourself to field matching errors which can happen from time to time. In cases where the percentage is high but still less than 100%, I'll make the D2V using Honor Pulldown Flags and then use TIVTC and this:

TFM(D2V="C:\Path\To\Movie.d2v")
TDecimate()

That says to treat the soft telecined portions as if they had been Force Filmed and only IVTC the few hard telecined parts. Much faster and safer. You can even disable the post processer if you're sure this is an easy one, for even faster encoding:

TFM(D2V="C:\Path\To\Movie.d2v",PP=0)
TDecimate()

[onesikgypo]

First id like to say thankyou manono - you have been a great help as ive been encountering these different types of videos. It is much appreciated.

Vvdubmod automatically showing it ar 23.976 makes sense as to why i didnt see dupe frames.

Also - you are correct, in the first 3-5seconds of the vob where it was just black space it showed the video as just "NTSC" - so hard telecine - then when th epicture began, it went to a FILM %.

Re: your recommendation to use TIVTC and TFM & TDecimate()

im guessing that would be the best thing to use in this case - since we have those few frames of hard telecine - and the rest is soft telecine - so it seems from your explanation - both areas will be properly ivtc'd, and also you said itd encode faster.

One thing though i was unsure of is when you said "if you're sure this is an easy one" regarding turning the post processer off - what exactly did you mean by that?

[jagabo]

I would look at the main movie, skipping over things like the studio logo, the distributor logo, black lead in, etc. If the main movie is progressive with pulldown flags I just use the force film setting. I usually cut off the logos at the start and end anyway. TFM().TDecimate() does make mistakes.

[manono]

It's usually OK to just Force Film it when the percentage is that high and be done with it. It's definitely not always OK to do that. I've been burned too many times and can cite example after example. The Criterion DVDs, especially, drop to video at chapter breaks and a Force Film will invariable lead to many frames being interlaced. And they're almost always 99%+ film.

And that's why, even with percentages into the 99+ range, I'll still follow that script above. It does not IVTC the soft telecined parts, as I thought I made clear. The only time I'll Force Film a DVD is if the percentage is 100%, or very, very, close, and I see a few 222s in the D2V at the beginning or end and nowhere else.

One thing though i was unsure of is when you said "if you're sure this is an easy one" regarding turning the post processer off - what exactly did you mean by that?

What I meant was if it hasn't been edited on video, and doesn't have pattern breaks, it's usually safe to leave off the deinterlacer for faster encoding. You'd better check it carefully first, though, and be sure who made the DVD and know what kinds of DVDs are safe. For example, TV episode DVDs often aren't safe because so many have been edited on video, and I wouldn't leave off the post processor. Movies with a high film percentage are a much better bet for disabling the post processor.

[onesikgypo]

OK - thankyou

My last question then, is there any time where the use of the script you gave above would not be used, if the video wasnt all completely hard telecined. i.e you mentioned that its usually ok to force film, but will the technique you gave wiht the above script ALWAYS work - for certainty, or are there also situations where it may cause mistakes if used.

[jagabo]

And that's why, even with percentages into the 99+ range, I'll still follow that script above. It does not IVTC the soft telecined parts, as I thought I made clear.

How could it not IVTC the soft telecined parts? If you enable pulldown flags in DgIndex the entire film will be seen as hard telecined by TFM().TDecimate().

[manono]

How could it not IVTC the soft telecined parts? If you enable pulldown flags in DgIndex the entire film will be seen as hard telecined by TFM().TDecimate().

Because it treats the soft telecined parts as if they are being Force Filmed by using only the TFF and RFF flags for the field matching, just as does DGIndex's Forced Film. It does no field matching using the TFM algorithms, except under certain circumstances. This is from the TFM - READ ME, the "flags" section, immediately following the "d2v" section:

0 - Check the d2v file for illegal transitions and set the order parameter if it
is not already manually set. Also, pass on rff flag duplicate info to
tdecimate.
1 - Same as 0, plus use the trf flags for field matching in film sections (sections
where the trf flags follow the 012301... pattern)
2 - Same as 1, but use the trf flags for field matching in all areas (doesn't have
to be in 0123 pattern) (very much not recommended!)
3 - Same as 0, but don't pass on any info to tdecimate (i.e. only set order and check
for illegal transitions)
4 - Same as 1, but d2v matches are checked for being combed. If a d2v match is detected
as combed then tfm uses its own matching routine for that frame.
5 - Same as 4, but d2v matches are only checked for being combed around scenechanges.
.
.
.
Default: 4 (int)

If you keep to the default 4, you'll get exactly what a Force Film would have given you, with the exception that if it discovers interlacing it does its own matching and/or deinterlacing. If you set Flags=1, you get back the exact same as a Forced Film in the soft telecined parts, even if there were mistakes in the original pulldown (I've seen that also, where an entire film's D2V goes 01230123, but there's still some interlacing remaining after a Forced Film). Yes, it still uses TDecimate to pull out the dupe frames, but you're not saying there might be a problem there, are you? Because in the D2V section just above what I quoted earlier:

TFM will also use the d2v info for field matching and will pass info from the d2v on to tdecimate to help aid duplicate detection and hybrid detection.

I'm assuming a DGIndex Forced Film to be different from a full-on IVTC. I hope you are as well.

My last question then, is there any time where the use of the script you gave above would not be used, if the video wasnt all completely hard telecined. i.e you mentioned that its usually ok to force film, but will the technique you gave wiht the above script ALWAYS work - for certainty, or are there also situations where it may cause mistakes if used.

Well, you have to make sure the source was telecined in one way or another - either hard or soft telecined, or a mix of both. If you have pure 29.97fps interlaced or progressive video, or if you have some field blended PAL2NTSC source, then obviously it won't work. If you like to encode anime then it may not work perfectly all the time, because anime DVDs can be messed up in so many ways. And one probably shouldn't say something will always work, because there may be exceptions. But for a mix of hard and soft telecine, I'd say that script is pretty much guaranteed to get back the soft telecined parts accurately. And it'll be faster than using TFM/TDecimate without also using the D2V parameter, and (probably) faster than IVTCing the entire thing using Telecide/Decimate, and might even produce better results when compared to a full IVTC of the whole thing because of the chance of field matching screw ups.

[jagabo]

How could it not IVTC the soft telecined parts? If you enable pulldown flags in DgIndex the entire film will be seen as hard telecined by TFM().TDecimate().

Because it treats the soft telecined parts as if they are being Force Filmed by using only the TFF and RFF flags for the field matching, just as does DGIndex's Forced Film. It does no field matching using the TFM algorithms, except under certain circumstances. This is from the TFM - READ ME, the "flags" section, immediately following the "d2v" section:

0 - Check the d2v file for illegal transitions and set the order parameter if it
is not already manually set. Also, pass on rff flag duplicate info to
tdecimate.
1 - Same as 0, plus use the trf flags for field matching in film sections (sections
where the trf flags follow the 012301... pattern)
2 - Same as 1, but use the trf flags for field matching in all areas (doesn't have
to be in 0123 pattern) (very much not recommended!)
3 - Same as 0, but don't pass on any info to tdecimate (i.e. only set order and check
for illegal transitions)
4 - Same as 1, but d2v matches are checked for being combed. If a d2v match is detected
as combed then tfm uses its own matching routine for that frame.
5 - Same as 4, but d2v matches are only checked for being combed around scenechanges.
.
.
.
Default: 4 (int)

If you keep to the default 4, you'll get exactly what a Force Film would have given you, with the exception that if it discovers interlacing it does its own matching and/or deinterlacing.

Ah, OK. Thanks for setting me straight.

[onesikgypo]

Thanks, Just for the benefit on anybody who is trying to figure this out (and undoubtedly myself incasse i forget) - this is what ive been able to summarise in regard to normal non-anime NTSC retail dvd's. This is perhaps not always certainly going to work, but for at least the large majority the following would apply:

1) If it's hard telecined throughout - that is in DGIndex Itll Show NTSC As the Frame Type Throughout - Then Force Film Will Not Work. IVTC needs to be done via decomb - with the following script:

Telecide(order=0)
Decimate(cycle=5)

With No Additional DeInterlacing To Be Applied - As It Occurs Automatically Through The Process.

2) If its FILM 95%+ throughout - then in that case Force Film Will Be Best when you output the d2v file through DGIndex or alike. Then the process will be as normal.

3) If your video has a mixture of both hard telecine (frame type shows up as NTSC) and also soft telecine (frame type shows up as a FILM %):

3a) If The Hard Telecine Parts Merely Represent The Beggining And End "black" frames (fade in/fade out) - You Can Merely Cut Them Out - Then Proceed As If This Were the 2nd scenario. - You could also do the below.

3b) If The Hard Telecine Parts Are Within The Video in a way that the frames cant be cut out then you will need TIVTC then using the following script:

TFM(D2V="C:\Path\To\Movie.d2v") 
TDecimate()