VHS to PC with HuffyUV to MPEG/H.264?
 

Hi everyone, thanks to what I have read on the forum in recent years I am completing my video capture chain from VHS, VHS-C to PC. Right now I have:

- Panasonic NV-FS200
- ADVC-300
I get rid of the ADVC300 on the advice of various forum members, and thanks to lordsmurf who will send me the TBC green AVT-8710 and new acquisition device, I improve the acquisition chain a lot.

When I'll have everything I would like to acquire while losing as little as possible, so in 720x576 with Huffyuv losless codec. I will keep the source files in archive but how to filter and clean up the videos I still have to choose.
The videos are all homemade from a camera, some cassettes instead come from a previous transfer from Super8 to VHS made by a photographic laboratory (very rough).

I know that it is better to avoid the crop and I would also like to keep the interlacing, I see no reason to remove it (at least in the sources to keep archived).

I saw how to mask the edges in virtualdub, but I wonder, if a resize is not done, how to get to a centered image, and to a resolution that is compatible for future encoding (Mpeg2 or h264)?
Should I activate the letterbox and leave the black frame up to 720x576?
Do I go down to a slightly lower resolution? 720x540?
Is it better that the PAR stays in 4:3? Square pixels?
If I left interlaced maybe it would be better to have the final files in mpeg2 Standard DVD?
The vision will be via HTPC connected to the 16: 9 TV, but obviously to be seen in the original 4:3 format
Some advice?

Correct, no reason to deinterlace source captures. Do that in later copies/encodes. And retain source.

Masking will re-center within the matte. If you cut 20 pixels from bottom, 0 from top, the matted image will have 10 pixels black above and below. Understand? So VirtualDub masking "auto letterboxes" to the 4x3 720x576.

For later copies/encode, to encode at 4x3 for a 1:1 pixel ratio, yes, 720x540 is good -- but ONLY after deinterlace. Before deinterlace, you must do 768x576. However, in PAL, it means so loss of detail from upsize. (With NTSC, we have 720x480 > 640x480. So for once, NTSC is better!)

MPEG copies for viewing interlace as 4:3 is good.
H264 copies (in MKV) deinterlaced as 1:1 is good.

The TV will auto-matte down to 4x3, not stretch. All TVs have an "SD squish" mode.

It is clearer, so the most logical way would be to compress the files in mpeg2 so as to always leave interlaced. Then I don't need to upscale and leave 720x576

Otherwise to go to h.264:
Deinterlace first, then should I upsize? Can't leave 720x540?

High bitrate (15mbit+) 4:2:2 MPEG.
Not DVD compression >10mbit, but <15 (20/25 even)

h264 = deinterlace usually required.
After deinterlace, 720x540 is fine, good size.