zoom in on hdv footage for sd use

[Marc Hangl]

Hi
I usually shoot in hdv on a sony hd1000 and use the camera to downconvert to sd for editing in fcp. Sometimes when i look at the footage I realise its not framed well or i would prefer to zoom in a bit.
What I was wondering is at what point in zooming in on the original hdv footage would there be a noticeable lose of quality after converting to sd compared with the original sd footage.

In terms of number of pixels you have approximately 4 x the number of pixels in hdv which should allow you to zoom into 50% but then you have to take into account that hdv is far greater compressed and also GOP.

I was wondering if anyone has experimented with this and could other some guidelines
- is there a difference with when the camera is static or panning zooming

Thanks for any advice
marc

[Pete Cofrancesco]

I've thought about the same thing but haven't had the time to experiment. The other thing you have to take into account is the lighting. Any noise from gain will also be magnified. I'd assume 10% zoom would be fine. I know with still photography that you lose detail and quality when you zoom in a lot. Just think of it this way, the camera sensor is 1/3 in (HD or SD). Zooming in 2x is like your shooting with a sensor that is half that size.

Regardless I'd be curious of test results rather than my speculation.

[Martin Chab]

I´ve done this some weeks ago going from 1024 to full 1920 (I mean close to 45% zoom in). also I reframed several takes. It worked great.
In one take the camera was wondering in a crowd stopping here and there. I reframed on the camera movement so when the frame stabilizes is already reframed. Is impossible to see any reframing going on!! is very natural. Again, works great!!

[Jon Goulden]

If you want to zoom in on the HD footage, you need to leave it as HD (don't down res). Set the FCP sequence to SD, then drop in the footage as usual. On the HD footage you can zoom and pan (it has roughly 4x the amount of information as SD so you can do a lot).

The FCP sequences allow intermixing of various resolutions. It will look a little 'fuzzy' until you render.

However, if you are doing multiclips, I'm not sure of the best workflow for best quality (I just posted a question on this which is how I found your post).

[Aric Mannion]

I do this all the time.
I would never, do this in final cut though. HDV is still interlaced (assuming this is 1080i) with upper fields. when you zoom in those fields are noticeable (and ugly) horizontal lines, especially during movement. Also final cut like to add a shift fields filter to HDV clips in a SD timeline automatically, which I think is a super crappy filter. A lot of people complain about there HDV footage looking worse then SD when they downscale, and these are some of the reasons.
The way I do it is in After Effects. When you import footage in AE, it usually interprets footage automatically. But you can do it yourself by control clicking the clip in AE (or apple+F) to "separate fields-upper first" and check "Preserve Edges". It's now de-interlaced, and dropping it into an SD comp will give you a high quality 480p image to play with that you can make 480i on export if you want.
If you don't have AE, quicktime does a pretty good job de-interlacing. I guess you could open the video from your capture scratch and hit apple+J, select the video track and you can de-interlace in the visual settings, then you will be prompted to save when you close it. I suppose that would look fine in a SD timeline in FCP, but I never tried. Just make sure FCP is not adding some filter to your HDV clips automatically.

[Christopher Drews]

Good suggestion Aric - I was about to mention AE but you beat me to it. AE is really the best de-interlacer program around. My workflow always is Shoot -> AE -> FCP.
-C