For one thing, the DVD specification covers a range of video and audio bitrates, GOP structure, and other arcane parameters. Accordingly, there are any number of possible DVD-legal MPEG-2 encodings of your original AVI that you can generate.
It is not uncommon to get some pixelation at the edges of moving portions of the video image as a consequence of compression into MPEG-2. Minimizing such usually comes at the expense of size of the encoded MPEG-2 file and the time it takes to do the one-time encode from AVI to MPEG-2.
It sounds like you are providing MyDVD with your AVIs and letting it do the encoding. The MyDVD encoder is fairly lame. If you want more control, try using the encoder available from
http://www.tmpgenc.net which is available for free as a trial for 30 days, and for which a license can be purchased for $48.
If you use TMPGEnc in connection with MyDVD, bear in mind that MyDVD insists that the audio be 48 kHz PCM. Given this, you may want to set up TMPG to generate separate video (.m2v) and audio (.wav) elementary streams using the appropriate settings. As long as the audio and video files are in the same subdirectory and differ in name only as to their extensions, MyDVD will pick up the .wav file automatically when you direct it to the .m2v. It is my understanding that if your audio was recorded by the camera as 48 kHz PCM, TMPG will not do any transcoding and will just separate out the audio portion of the AVI.
If you decide to use TMPG and it balks about accepting your AVI as input, you may convert the AVI from Type 1 to Type 2 using the Canopus DV File Converter, which you can get from
http://www.vcdhelp.com/tools. For guidance on choosing the settings for TMPG, see
http://pwp.netcabo.pt/0165394101/TMPGEnc_Template.html.
This may be more information than you wanted or expected, but I have just gotten to the point of using MyDVD with my new Pioneer A05, and I have been collecting information for a couple of weeks on the subject, and there's no point in you doing it all over again.
As for the particular problem with red in video, red is particularly susceptible to bleeding, regardless of whether it is presented in DVD form or not. I will defer to somebody else to explain the physics that underlies this issue.
Regards,
Eric J