Capturing for DVD2SVCD

I have an AverDVD capture card based on a BT878 chipset. I'm wanting to do a capture from a VHS and convert it using DVD2SVCD. I'm wondering what the best format for capturing would be. I assume it's probably AVI, but what compression should I use, and what audio settings? Anything else to consider? I'll probably use Ulead DVD maker as it's one of the few progs that will let me capture in any format (my card is picky). What software do you use? I admit, I'm a newbie, and the numerous options are a little overwhelming, so I'm wondering what formats people have had luck with. By the way, if this has been covered before, I apologize. I did extensive searches and didn't come up with much.

Post back to let us know what you find out. I am interested as well. I think you may have the best luck with a card that captures straight to Mpeg2.

If high quality is important to you, as Mark said, capture to a (low) lossless codec avi like Huffy AVI and then use DVD2SVCD. The Huffy avi file will be very large. If you can accept a lower quality, then you can probably use the avi codec that came with the card. As for capturing to mpeg using a codec provided by the card (does it?), I believe you will find the quality to be less than when using the 3 mpeg codecs that DVD2SVCD is based on. If you are capturing to a NTSC Interlaced format (very good chance you will be), then consider using the Canopus ProCoder mpeg codec in lieu of CCE - which is the main reason the Canopus codec support was added to DVD2SVCD. But as Mark said, it is probably best that you visit the Capture forum for a while to find out how get the best avi file that you can first and then load into DVD2SVCD. Also, consider capturing at the format size/ratio that you intend to have at final output (SVCD or VCD) to avoid any artifacts due to resizing during the avi to mpeg process.

No, capture at full D1. What you've suggested would seem to make sense but it does not yield the greatest probability of accurately sampling the analog signal. You're thinking in digital terms, not analog. Each analog pixel is a constantly varying signal. For video tape, CVD is really good because then you've got two samples per cycle, minimum. 4x oversampling would be best but that's only available with very expensive systems. Sampling and editing should always be done at the highest possible resolution, preferably at a multiple of the base frequency, and the last step is mixing down. One sample per cycle is the worst option. You can't really be sure at what point in the cycle the sample was gathered.