Resizing comp for output
by Darren5102
on
Nov 30, 2006 at 12:00:56 am
I've got a composition that is 2340x1560 and it needs to be 720x480 for DV NTSC output. I hope this isn't a dumb question, but I've never had to do this before. What is the best way to change the size without losing resolution?
I've already tried precomping the project and reducing the scale, but the final render looked soft on my NTSC monitor. Is this a situation where 'stretch' in the 'output module' dialogue box of the 'render queue' comes in handy?
Thanks for your output, I mean input. . .whatever. :)
Re: Resizing comp for output by Mylenium on Nov 30, 2006 at 12:34:22 pm
Stretch does the same like pre-composing. Regardless of this, your image will never look as sharp. Until you explicitely re-build your entire comp for the lower res, certain pixels will always be mooshed-up and fall between the cracks, resulting in a softened look. If there's just that little time left, try to sort thru your layers and find any potential candidates. The goal here must be to make things look more crude, so get rid of overly detailed vector art, lines that are too thin and the usual suspects. Also make it a point to check your colors - some colors have a tendency to "bleed" more when downsampled (reds, Yellows, intense Greens) and will look more blurry than others. Lastly, on layers where you can be sure to have sufficient headroom, set the layer quality to draft in your big master comp. This wil lresult in different oversampling patterns and in many cases look sharper.
Re: Resizing comp for output by Darren5102 on Dec 2, 2006 at 12:51:38 am
Thanks for the detailed response! It helped to know what I'm getting myself into going down that road. Sorry I didn't reply sooner. You're right, it looks like a pretty big mess to change all those settings for what is essentially a lower quality project.
I wanted to run this by you though, to see what you think. I may post it on the FCP forum as well. Here goes: I decided to render the full 2340x1560 comp out and then I imported it into FCP. Well, it looks great! The blurriness is gone! Only one problem: it squeezes the image thinner height-wise, and fatter width-wise. The image isn't noticably distorted (like I said, it looks great!), but now I've got a blank area above and beneath the image. This isn't acceptable. My instructions were to fill the screen with the image. (The image, incidently, is a two-page spread of pages in a book that I've animated to turn from page to page.)
In AE,I've tried adjusting the comp size to 3:2, changing the pixel aspect ratio to (0.9), and interpreting the footage to DVNTSC. All any of this does, is push the picture around inside the bounds of the image, but the boundries of the image itself never grow or shrink.
At this point, this is probably an FCP import question, but I wanted to send this your way in case there's something I'm missing on the AE side of things.
