HDV deinterlacing and color space
 

Hey guys,

This is a question that came up on a forum elsewhere and I figured I would post it here in hopes of finding out more. I understand the technical aspects of the HDV format rather well. No problems there but this one stumped me.

Hoe does the Sony FX1/Z1 deinterlace in camera? I ask because I am curious what affect this has on the 4:2:0 color space. Anyone know?

I'm not sure why someone would ask that question.
4:2:0, 4:2:2, 4:1:1, or whatever other luma/chroma sampling scheme used isn't related to interlaced or non-interlaced, or deinterlaced. you can have any flavor of the above in p-scan or interlaced or deinterlaced.

Yes I know. I'm just curious how the deinterlacing handles the chroma from the two fields when blending them. Is there a loss of chroma (ie less than 4:2:0), is there a gain, nothing?

For instance, when down converting you can get full chroma by throwing out one of the fields. While this isn't the same thing, I do wonder what effect the field blending has on the color per any given area.

I just did a test, you can easily repeat.
Shoot colorbars or something of mixed colors. (I used a color chart)
Shoot all modes.
Capture.
Grab stills of all modes.
drop stills into Photoshop.
Use the eyedropper, compare colors.
you'll not find differences between the two modes.
In fact, deinterlace in your app, and do the same thing.

The only thing you should see is some loss in sharpness.

Spot, just a thought, has anyone tested these new cameras against highly contrasting plaids and polka dots (the bane of dv25)?

Polka dots yes, plaids...no. I don't have any, and frankly hadn't considered it. On my way to Walmart....I'll report back later.

Sampling raster doesn't affect colorspace properties in terms of color rendition. There is however a known issue (the "chromabug") when 4:2:0 footage needs to be deinterlaced. Maybe the artifacts are less visible in HDV than they are in SDDV