In extracting a windowed image from a larger file,

If I submit:
r2cf A112_C002_100679_001.R3D SeqLeft444.avi -(0,0,1920,1440) -P1.0

I get a 1920x1440 4:3 file with a pixel aspect ratio of 1.0


If I submit:
r2cf A112_C002_100679_001.R3D SeqLeft444.avi -(0,0,1440,1080) -P1.0

I get a 1440x1080 4:3 file with a pixel aspect ratio of 1.333 (the image is stretched)

It doesn't seem to make a difference if I scale the image to 1440:1080 from 1920:1440 or simply window that size. In both cases when I create a 1440:1080 file the pixel aspect ratio is 1.33. I've tried both with and without the -P switch.

My final goal is to be able to window two scaled 4:3 1440x1080 files from a larger 4096x2048 file.

Is this simple arithmetic failure on my part again or am I missing something even more basic?

Like hitting a special case on the decoder. HDV is 1440x1080, and that is displayed as 16x9. The decoder is not believe that there is 4x3 HD. The sample is still 1440x1080, check it is virtualdub.