Useful to feed MeGUI-xvid/264 (or codecs in general) full-size frames?

[chconnor]

Hi -

If i'm rendering out of Vegas and using MeGUI to process video with various codecs (in this case xvid or x264), is there any use in feeding it full size frames if I'm going to scale them in MeGUI or is that pointless?

If i understand the pipeline, AviSync just scales the frame before feeding the codec the reduced version anyway, so if that's correct, the question is "is there any advantage to having AviSync scale the frame with a fancy algorithm instead of having Vegas render to that scale in the first place."

Otherwise I'll save the disc space and just render smaller to begin with... I guess I was kind of surprised that with all the fancy math, codecs don't (seem to) take advantage of the full frame when doing their compression in cases where they are going to be scaling it down at the same time. In other words, i take it codecs don't resize frames (dimensionally speaking)? Seems like there must be a usefulness in that somewhere... Anyway:

Part 2 to that question - if the full-size frames are useful, any reason to render those full-size frames at mod16 dimensions? Is it better for scaling for any of the scaling algorithms?

Thanks!
-c

[poisondeathray]

If HD space isn't an issue I would keep full size.

Why? - (1) You have many choices with avisynth to do the resizing of frames. I can't recall the choices Vegas has? For example, you might choose a Lanczos resize for a sharper image, vs. Bilinear resize for a smoother image. (2) It is ALWAYS better to have a higher bitrate, higher quality source (i.e. more information per frame) to feed into the compression codec - the end result will be better because the codec can "see more bits". An analogy would be making a MP3 from a CD source, vs. a cassette tape source; the cassette already has lost information bits, and the resultant MP3 will be lower quality. If you feed it a mediocre input, you will get a worse-than-mediocre output

The mod16 part comes in AFTER the resize usually for maximum compressibility reasons. (i.e. you should to resize TO mod16 dimensions, not FROM)

Cheers

[chconnor]

Cool, I grok the fact that the mod16 is relevant post-resize... i thought maybe the resizing algorithms liked mod16 as well, guess not. And it sounds from what you wrote that the full-size frame is not useful if the resize happens before the codec (as with AviSynth into the codec) except insofar as AviSynth provides more options for resizing than Vegas (i'm sure it does); and your point of having a full-size max-quality source file to render various resolutions: understood.

Thanks!
-c