DV to XVID, a few questions :)

[Magister]

Hi, 1st post here

My DV video is 720x480 in 16:9, I want to convert it in xvid and burn them on a DVD to watch on TV.

I did my search before, I know how to use VirtualDUB but some option are still obscur to me.

I know I have to keep it interlaced, bottom first, but I made tests with a 12 minutes video and used a smart de-interlacer for a test and kept the interlacing in another test. First, why interlaced video produce bigger file? Also on TV it seems there's more macro block in the interlaced video (watching on a 52" HDTV), does the DVD player issue a 480p signal even if the video is interlaced? it's a Philips DVP642 using components, who is making the de-interlacing?

And about resolution, a couple of thread on various forum say to resize from 720x480 to 640x352, do I really need this?!? my camera (canon ZR700) has a very nice 720x480 picture so downsizing it is... well... lame? isn't it? This is footage of my daughter, in 30 years when she will watch it again, she will have maybe a 8000x3000 screen, so it's better to keep the high resolution, right? and also if using a de-interlacer *at* compression gives some lose, it's better to keep the interlacing in the compressed video and let a specialized hardware chip which will be available in 30 years do a superb job by itself?

Also I am using xvid with what is known as the "mb1 interlaced DV" optimized quantization matrix with an average bitrate of 4000kbp/s, using 2 pass. Is it ok? I have the standard default I think, I am using AS@L5. Adaptive Q, Interlaced Encoding, no QPEL, no GMC, BVOP 2 max consecutives, packet bitstream, treillis, motion search of 6, VHQ mode 1 for b-frames too, chroma motion All the settings in % in the second pass are default.

And for pixel aspect ratio and display aspect ratio, do I have to put 16:9 in the PAR or the DAR? even if I can change them after with MPEG4Modifier in case of errors, I'd prefer to have them without retouching the avi.

thanks a lot in advance for the answers

[edDV]

I suggest you try an interlaced MPeg2 9000Kb/s CBR encoded DVD as a reference. Make sure you include pan, zoom and object motion under various lighting conditions. Then compare your xvid experiments to that DVD and original DV playback using the same video material.

I'd be interested in your results.