SD from a HD Camera

[travbickle]

I'm trying to set up a workflow for film transfers. My camera only captures HD AVCHD video, but can output 480p on HDMI. For transfers that will ultimately (and maybe only) end up as SD, should I just downconvert the HD that the camera captures, or would I get better quality by using my PC to capture the 480p output? I could use the Intensity Shuttle for this, correct? The camera is a Panasonic AG-HMC40.

[jmac698]

The simple answer to these types of questions is always the same, try both ways and see what you *you* think is best.

There's always some potential to have better quality if the source is HD. Consider for example, that after the transfer you find you can crop a little off the edges. Now say you have 1380x1020 (4:3). You can downsize this to 702x480 with no loss, where if it were only SD in the first place, you would have to zoom it again to fill the screen.

There's also the potential for Ken Burns effect in editing, or making some widescreen version.

As far as straight downconverting, there's definitely better quality available in software, try the gamma-aware resizing in Avisynth. You can mostly notice the difference in sharp highlights, city lights or star backgrounds.

Any method being done in hardware can easily be duplicated in software; it's not likely that the hardware will be better quality. Just think, hardware has to do this in real-time, that's a lot of processing power, and the more processing power the better the quality, so hardware will never ever equal software quality.
The exception to this rule is that the hardware uses some nice-looking techniques that are patented or otherwise not available in software. For example some 3d comb-filters in hardware capture cards may be better than available software.
Also there is the case that hardware has access to more of the signal than software can; in the case of hardware TBC, the image adjustment is based on a raw signal which is not part of what's ultimately captured.