I'm quessing you will answer the second option, but I'm thinking about it like this: If I shoot HD with my camera and downscale it to become SD in Premiere, the footage tends to look better than shooting SD to begin with. Would the same principle not apply to animated lower thirds/text
No, if anything it will look slightly (very slightly, hopefully) worse. On a related note, though, I understand Premiere Pro recently got an option for improved, higher quality scaling (in the 4.1 update, if I recall correctly),.
I think the HD to SD downcoversion example sometimes is more about subjective qualities than anything. If you stick a Letus adapter on an SD camera to use a 35mm lens on it, most viewers will prefer it over downconverted HD because of the "perception" of resolution and the shallow depth of field. This has little to do with scaling interpolation math
Thanks for clearing that up!
1 person found this helpful
Would the same principle not apply to animated lower thirds/text?
No, not really. What, however happens, is that your text's pixels are blended with different pixels of the underlying footage and because scaling samples based on the source dimensions, the result is usually a little more pleasing than the fixed size antialiasing kernel used to blend at native dimensions. so basically by making it a bit softer, you make it look better. I agree, though, that just like films a lot of the perceived quality boost of HD footage hinges on other factors.
Thanks for your insight Mylenium!