Does anybody know how to maximize the quality of the Mpeg (future DVD) render? I have set video quality on NTSC DVD to "best" already, but am wondering if there are any other areas where overall improvements could be made.
My advice... don't fiddle. Leave the default settings alone unless you have a REASON and KNOIW WHAT IT IS why you want to change any of the advanced settings. My guess 99% of those changing the settings change them in some attempt to get "better" results often resulting in getting worse results. The defaults were set as they were for a reason. They work. Many set top DVD players can stumble if you push the bitrate too high. Also rendering at "best" in many situtations does nothing but increase the rendering time.
This is one of those times where the choice of words is misleading. Picking "best" over "good" doesn't mean you'll get better quality, it means nothing but a change in the interpolation method. Good uses bilinear, while best uses bicubic. The reality is under many conditions the faster bilinear method renders a "better" picture or just as good as the bicubic method which takes much longer. Not my words, right from the mouth of Sony, (paraphased) well the words of Sonic Foundry. I asked this question years ago and that was the answer given.
Thanks for the advice, what you said makes sense. I didn't know about the bicubic/bilinear difference, and I couldn't really make out any difference between the two on tests.
I will stick with the defaults, since if the DVD stumbles the point is moot anyway.
About the only time you'll see any real difference and then only maybe if you use ultra high resolution still images, zoom or pan them then only if they have lots of gradations. Of course some isolated source files may benefit, but I've only seen it a few times, but then my eyesight isn't all that great anymore. Hows that for dancing all around the issue?
The reason I rarely use "best' and why Sony suggests you at least try "good" to see if it meets your needs has to do with the rendering time. With bilinear only ajoining pixels are checked. The pixel to the top, bottom, right and left. So when you zoom, pan, or are down/upsizing or in any way moving pixels around its a fairly straight computation for Vegas. However if you pick best, then bicubic in addition to factoring in the top, bottom and left and right pixels also looks at the pixels on a angle then computes a weighted average. All that extra work takes more time. If you got some extreme closeup where you want to pick up every subtle difference in someone's face, every pimple, blemish, varation, then bicubic will be more faithful. So too, its kind of a be careful what you wish for thing. By going the extra step you may infact bring out more inperfections.