Does the format that you capture in make a difference?

[Cart Wheel]

Huffyuv is a lossless codec, which means that no quality will be lost. All MPEG-2 compression will cause some quality to be lost, but depending on the bitrate/resolution you use, it may or may not be visible.

You will probably need a lot more free space for a Huffyuv capture. You can use the AuxSetup.exe program that comes with VirtualDub to find the data rate. Just go to the benchmark mode, enter your capture settings, and divide the given data rate by 2 because Huffyuv provides about a 2:1 compression. (FYI, KB/sec * # of seconds = approximate size of video in KB.)

Encoding with a higher quality motion search, softening block noise, and using noise reduction are also good ways to try to eliminate artifacts. Unfortunately, all of these will increase the encoding time. (In general, quality and time are inversely related.)

-Cart
http://www.geocities.com/lukesvideo/index.html

[SingSing]

If we are seeing small block in fast moving sence, then that
is the blockiness that everyone is taking about. It is a fact
of life for realtime mpeg2 capture on ATI AIW128.
Captureing at 480x352 can reduce a bit of those,but that is xSVCD.

Capture to AVI, and TMPGenc to SVCD always get a better result.
But it take a long time. I give up this method for movie length work.

Another problem is the wid-screen top and bottom black bars.
Those normally negatievly impact your capture and encording video quality.