Interesting situation...IVTC or Deinterlace?

[angelleye]

I've ripped a DVD of Dave Matthews Band Live. I've done many DVD rips and encoded to Divx so I've ready over MANY different guides and such.

In this particular case I'm using Gordian Knot. DVD2AVI previews the file as strictly NTSC Interlaced. Therefore I'm choosing None under Field Operation.

I then open the d2v file and when I click frame by frame I see that it's only 2 out of every 5 frames that are interlaced. Normally, according to everything I've ready, this could be safely IVTCed.

However, the interlaced frames are not the same as surrounding frames. One of the things I'm doing with these DVDs is studying the chords and fingering that Dave uses while playing. As I move through frame by frame I notice that at times he moves his hand or a finger or something within the 2 interlaced frames. Won't the IVTC be dropping these and in effect leaving out some of what I'm wanting?

I actually went ahead and encoded it using the deinterlace filter and it looks fine...but I was curious about that IVTC thing and wondering if that woudl have indeed been the better route.

Any information would be greatly appreciated.

Thanks!

[jggimi]

Take a look at the first graphic in the IVTC tutorial (http://www.doom9.org/ivtc-tut.htm). You'll note that the "interlaced" looking frames are fields from two adjacent film frames.

To test whether IVTC would do a better job than deinterlacing alone -- and save you 20% of the space, giving you the ability to use a higher bitrate -- create an .avs script that includes IVTC filters. But rather than encoding, just open that script in VdubMod, and scroll through the same scene.

When using Decomb for IVTC, you can sometimes speed up the process when the Telecide filter reassembles frames without needing to deinterlace. To test this, edit the .avs script to have:Telecide(post=false)Then, look at scenes with movement frame by frame in VdubMod.

[angelleye]

as i'm studying this thing a little closer...it looks as though all the frames are indeed interlaced. I think I'm just going to have to deinterlace.

Your reply brings up some questions though. When I open the AVS file in Virtual Dub and scroll through...is that what it's gonna look like when it's done encoding?

That could really come in handy...and I wasn't aware of it.

thanks for the reply...

drew

[killingspree]

Originally posted by angelleye

Your reply brings up some questions though. When I open the AVS file in Virtual Dub and scroll through...is that what it's gonna look like when it's done encoding?

That could really come in handy...and I wasn't aware of it.

well no, not quite. everything like compression artifacts etc are missing. unfortunately vdubmod cannot simulate compression artifacts. :)

anyway, resolution, AR, and deinterlacing artifacts etc will be the same in the encoded video!

steVe

[jggimi]

The AviSynth script will show you everything to the video except the encoding. The "Preview" button from Gknot's "Save & Encode" window does the same thing.