Scaling Clips Vs Scaling up the Rendering – 320×240 – NOT another HD Scaling Question

[Dominic Lamontagne]

Hi Brad,

I’m pretty much in your “coming out of retirement” situation, so here are my two cents:

For DVD output, I would scale up the (movie) layer full screen in a 720×480 composition (NTSC DV preset in your Composition settings) running at 29.97fps before render.

640×480 is close to Standard Definition (SD) video but to get a nice full frame on TV, you should use the standard SD format which is 720×480. If your layer IS 352×240, scaling up 204.5% will give you a 720×491 layer which would be more than enough to fill-up the 720×480 comp nicely. If your layer is 320×240, scaling up 225% will yield a 720×540 layer which you can position how you like in your SD 720×480 comp.

Then there is the frame rate issue. Is your footage 30fps, 30 drop (29.97), 15?
Is it interlaced (lower field first?) or progressive (full frames)?
When rendering SD video (for DVD), frame rate is 30 drop (29.97) with lower field first field rendering enabled.

Enabling frame blending will give you smoother motion on 15 fps footage that is “spread” across 29.97 frames.

Interpreting your footage (file-interpret footage) correctly when you import it in AfterEffects is critical.
Basically, if AfterEffects understands the type of footage you are feeding it, you can play with it in your comps (stretch it, move it, etc.) without any problems.

When rendering time comes, don’t forget to turn on field rendering, but keep the output module to H.264 or Loseless (don’t forget to add sound back in if you use the Loseless preset) and use ANOTHER software to compress your videos for DVD delivery. I use Adobe Media Encoder. Others use Compressor.

So there it is!

[Brad Bixby]

Great stuff,

Thank you so much I’m working with drop frame.

Is it interlaced (lower field first?) or progressive (full frames)?
I have no idea 🙂

225% that’s an even 25% for better compression results?

Thanks, again

[Dominic Lamontagne]

Normally, 30 drop is interlaced. To be sure: import your original clip in AE and select File:Interpret Footage:Main… and then choose Separate fields : lower fields first and then open your clip in the footage viewer (double-click). Using the “page down” key on your keyboard to step through the frames of your clip, you will notice that the timecode moves up one frame for every two keystrokes. This is because each frame contains two fields and the “page down” key moves you forward one field at a time. Now, the important part: if you notice comb-like patterns in fast moving areas of your frame, and the image changes from field to field (from half frame to half frame) your footage IS interlaced. Choosing a fast panning camera move in your footage is best to notice fields.
If there is no combing effect, and you only see a slight jitter from field to field (no movement), your footage is NOT interlaced. If it is interlaced, leave the interpretation (lower field first) as it is. If it’s NOT interlaced, interpret the footage back to “no fields” to Separate fields: OFF.

Now, the 225% issue: SD NTSC is 720×486 and uses non-square pixels (0.91 pixel aspect ratio). You could scale your 320×240 up to 640×480 (200%) but the surrounding “unfilled” area could show on your final NTSC display (especially if it’s a projector). Some people don’t mind this though.

What I prefer doing is working within an AE preset D1/DV NTSC comp (0.91), correctly interpreting the footage I am using (in your case, you’ll have to figure out the fields issue first and also choose Interpret Footage:Other options:Pixel Aspect Ratio: D1/DV NTSC (0.91)) and then fill-up the comp with the footage. In your case, to fill the comp with a 320×240 movie file in a 720×486 environment, you need to scale it up 225% (2.25×320=720).

Some people will tell you that you are softening the image even more going to 225% instead of 200%, but I don’t think this will be noticeable and you’ll know exactly what will be displayed on-screen.