HV20 reverse telecine?? Why it works??

[borax]

I have yet to edit any footage shot with the HV20 but I want to be prepared, for when I do. Reverse telecine seems to be the core process and I do have several formulas and applications ready to do this.

Right now I am having a hard time understanding how you can properly convert the HV20 29.97 to 24p without artifacts.

If one of the problems with deinterlacing HV20 footage is if the first frame you start from is not a "key" frame, and instead one of the GOP frames, won't that effect the reverse telecine process?

To further illustrate my point. If I hit record and capture 10 seconds of footage then stop and record another 23 seconds and so on for a full tape, Then capture all the footage into FCP with "capture now", won't the sequence of the Key frames constantly be changing?

In my mind that should result with a improper cadence if Key and GOP frames. So that when you apply reverse telecine to an hour of footage in theory some of the footage will be deinterlaced poorly and some perfectly.

Does that make sense?

I just want to make sure I understand the underlying concept at work

[Barry_Green]

I only tried this with Vegas, not with FCP, but my experience was: yes the sequence constantly changes. You have to do each clip individually and guess the pulldown for each and every clip. It was a tedious pain in the patootie and I ended up donating my HV20 to Isara.

Hopefully the process is more automated now and there's an easier way to do this -- let's hope someone writes in with the answer.

[borax]

Wow Barry,

If you are saying that's how it is, I'm worried!

Seems like an nightmare!

Since you responded to this post, can you make suggestions about my workaround.

Its here:
http://www.dvxuser.com/V6/showthread.php?t=128198

Thanks

[Barry_Green]

That's how it was. I don't know what the situation is today.

[Nektonic]

borax,

I think this thread might have some useful info.

http://www.dvxuser.com/V6/showthread.php?t=98411&highlight=HV20+workflow

Above you stated that you are using FCP. If you have the latest version of Compressor, then you can do the deinterlacing through that.

[borax]

Nektonic,

Thanks for the thread link, it's a great resource.

But still the question of whether or not you can capture an entire tapes worth of footage as 1 clip and successfully remove the pulldown when the cadence of key frames changes from shot to shot?

Simply will a hour of random footage in a single clip deinterlace properly or does it only work on a 1 shot per clip basis?

I know that all footage should be properly logged and captured per clip, but I normally like to be lazy and capture 1 entire tape at a time.

Has anyone tried this yet?

In that thread someone suggested that FCP2 actually seperates the clip automatically if the cadence changes! Is that really true?

Can I be lazy and have FCP2 do the work? (if so ...awesome!!)

[Ian-T]

CIneform can do this. You can capture a whole tape and it will remove the pulldown for you at the same time...it will even split up the footage for you. Ex: If you have ten different clips that sums up to 1 hour long...after capturing you can end up with 10 different clips with pulldown removed from each...in a high quality wavelet codec.