Questions regarding anime DVDRip

[newborn]

Hello everybody. I woul appreciate it if you could help me by answering the below:

I want to rip anime DVD to .mkv

sample from the DVD: http://www.sendspace.com/file/12z7ok

DGindex:

http://img338.imageshack.us/img338/631/46201265318.png

Note: before I load the .vobs in DGindex, the option "honor pulldown flags" was checked by default. When do we leave this option checked and when do we use "Ignore pulldown flags" or "forced film"?

If I understand things a little, this footage is telecined and needs inverse telecine. How can I deinterlace this video and keep 29.970fps by using tfm() and tdecimate()? I mean, there are a lot of modes for these filters and I am very confused. I also think that start credits are interlaced in a different way;

Also, how can somebody check for duplicate frames by vision only? And what does it mean in theory? Is there a guide with photos so as to be understandable by newbies?

After parsing the .d2v file to DGindex I opened the file with .txt and I didn't understand a thing. Here's the .txt:

http://www.sendspace.com/file/aa8vta

Could somebody explain it to me a little what all these numbers mean by using photos of my sample?

Moreover, how can somebody check for ghosting? Can you show me a picture from my sample where ghosting exists?


Let's go to Aspect Ratio. Megui claims that video is anamorphic ITU 16:9 NTSC (1.823169). Is this right? After cropping, if I encode without anamorphic encoding megui suggests resolution 720X400. But if I keep anamorphic encoding megui suggests 720X480. I want to encode and keep the best quality from the video. What should I do?

If I encode to anamorphic 720X480, what would be the exact aspect ratio to keep the video as shown in DVD?

[jagabo]

How can I deinterlace this video and keep 29.970fps by using tfm() and tdecimate()?

By using TFM() only.

Also, how can somebody check for duplicate frames by vision only?

You open the AVS script in an editor like VirtualDub and step through the video frame by frame (using the arrow keys in VirtualDub).

That video appears to be a mix of 30p and 12p. Probably other rates too.

[Guest34343]

Note: before I load the .vobs in DGindex, the option "honor pulldown flags" was checked by default. When do we leave this option checked and when do we use "Ignore pulldown flags" or "forced film"?

That is described in the DGIndex User Manual. For your video, there is no pulldown so Ignore Pulldown will do nothing because there is nothing to ignore. Forced Film also is inapplicable because there is no pulldown.

http://neuron2.net/dgmpgdec/DGIndexManual.html#FieldOp

[newborn]

You open the AVS script in an editor like VirtualDub and step through the video frame by frame (using the arrow keys in VirtualDub).

That video appears to be a mix of 30p and 12p. Probably other rates too.

Could you show me a screenshot where you see duplicate frames?

How did you understand that the video is a mix of 30p and 12p?

[manono]

Could you show me a screenshot where you see duplicate frames?

It's nothing you can't do yourself. Make the D2V using Honor Pulldown Flags and make an AviSynth script with only TFM. Open that in VDub and scroll to various places and begin advancing a frame at a time. The places where there's movement every frame are 30p (the same as the framerate of the video itself) and the ones where you see a pattern of 3 2 3 2 3 2, where those numbers represent the number of times you see the same frame, are 12p. When you see different patterns, then those are other easily figured framerates.

[jagabo]

Could you show me a screenshot where you see duplicate frames?

It would be pointless. All you will see is a picture. The point is, the pictures repeat 2 or 3 times in a row when you step through the video frame by frame.

How did you understand that the video is a mix of 30p and 12p?

When stepping through the video I saw the frames were repeating 2 or 3 times in a rw. So 12 different pictures were repeated like this:

3 2 3 2 3 2 3 2 3 2 3 2

If you add up all the repeats you'll see those 12 different frames were repeated to make a total of 30 video frames. 30 frames of 30 fps video is one second of video. Since there were 12 different pictures in that same one second the underlying frame rate was 12 fps.

In other parts of the video (where the snow(?) is falling) you can see that every frame is different. Since it's 30 fps video that means those sections are 30p -- 30 unique frames every second.

If you look carefully at the character animation during those "show" shots you'll see that the characters were rendered at 12 fps and the snow was overlaid at 30 fps. This is very common for animation where it takes a lot of work to draw the individual character movements, but it's easy to pan images around.

[newborn]

And how should I handle this case regarding the duplicate frames?

And what about the aspect ratio?