I'm doing it right now and it's working quite well for me... Perhaps the problem is your encoding? You want to make sure that you remove any interlacing from the video BEFORE you do any compression...
Also, make sure you use composite or S-Video hookups from your VCR to your Tuner.. Anything less than that will look like garbage..
My process is this:
6 Head VCR reads tape.
Audio goes into the Tuner card's stereo audio input.
Video goes into the Tuner card's Yellow composite input.
I record it to an uncompressed data file or a Lossless compressed data file (like FLAK + HUFYUV)
I use mencoder to remove interlacing and crop the image (some VHS tapes have black borders on the top and bottom since they were from widescreen sources). I have a Linux BASH script that first runs the video through mplayer to identify unused borders and such and then it gets passed on to mencoder..
After you've run a deinterlace, if it was done properly, you could pause the video at any time and will have a clean (albeit low quality) photo without any tearing or rough edges on objects in motion.
I then pipe the uncompressed, deinterlaced, cropped, scaled and very clean looking vidoe through a x264 (lossy) encoder.. That takes the longest amount of time to do (hits the CPU REALLY hard).. But takes a 5 gigabyte file and turns it down to 500 megs and it still looks just as good as it did before..
Once you get the workflow down, you usually can program it all into one script in Linux to do all the steps above automatically for every video you record.. But again, you'd have to be VERY good with the command line..
The best part though is that I didn't spend a penny in any software..
Onboard RAID vs. 3Ware RAIDI never recommend people run RAID-5 with onboard chipsets.