[Nick Haffie-emslie]

Thanks Karl – that’s a great article!

[Nick Haffie-emslie]

Thanks guys.

Chris – Good point. I should say Compressor scales better than Media Encoder on default settings; when Maximum Render Quality is on in Media Encoder, they seem pretty comparable.

Angelo – Based on my tests, I haven’t been able to detect any differences with in-betweening frames during speed/framerate changes (not yet anyway). It doesn’t, for example, appear to use the pixel flow algorithm that After Effects uses on its highest setting (where you’re not blending frames based on opacity, but actually trying to analyze and map where objects in the frame are moving).

Gabriel – I’m still a little confused by what a “moving object” is here. Are we talking about titles and/or other things in Premiere that the software knows is an object? You mentioned it renders them “more sharply” – how does this relate to motion blur? Does it only come into play when changing the frame rate?

Not trying to be nitpicky, just trying to get a sense of exactly what’s happening. I don’t want to waste the significant additional render time in situations where it doesn’t add anything (e.g. if I’m scaling but not time remapping; if I’m not scaling or time remapping but simply transcoding; etc.)

[Nick Haffie-emslie]

In terms of off-the-cuff tips:

– go heavy on gradients when making your bars (notice how the bars in the sample there often fall off completely to black on one edge)
– add the shiny glows that travel along the bars by making a separate layer (just a shape layer, or draw a mask on a solid) in the shape of the glows, blur them out (or feather the masks) and set transfer mode to Add. Maybe toss a glow effect on there (actually, throw a glow on an adjustment layer for the whole comp). Animate them travelling down the edge of the bars.
– for the reflection: precomp everything together – bars, text, logos, whatever – everything that you want to appear to be flat against the bars. We’ll call this comp with everything in it Comp A. Bring Comp A into a new comp and duplicate it (one of them is going to be the bars, one is going to be the reflection). Flip the reflection one by setting Y scale to -100% (leave X scale at 100% though). Drop the opacity on the reflection layer to like 50%, and put a mask on it, feathered, so that it falls off into darkness as you get lower.

[Nick Haffie-emslie]

I’ve just been dealing with a similar problem where QuickTime Animation codec is unable to carry field information from AE to FCP:

https://forums.creativecow.net/readpost/8/1015256

Sounds like it can’t carry PAR information either. The solution is to import into FCP, then in the Browser, scroll over to the right to the “Pixel Aspect” column and right click “Square”, changing it to “HD(960×720)”

[Nick Haffie-emslie]

Could you post what you have so far?
Then we can give you feedback on what you need to tweak (no sense going through a bunch of stuff that you already have right – best to focus on what’s not quite there)

[Nick Haffie-emslie]

This seems like such a simple feature. All the pixel math is right there in front of you on a graph. All you’re doing is some bezier interpolation of points and then amplifying or dampening a pixel’s color value. I know a lot of people that would throw $20 at someone who’d be willing to spend a couple Saturdays writing this plugin up!

Premiere has it, Vegas has it, After Effects has it…

< /rant>

Anyway, what’s the consensus on current solutions? Is colorista the best bet?

[Nick Haffie-emslie]

And while ProRes seems to keep its field order info, QuickTime does not seem to account for its pixel aspect ratio by default (always shows for me as square pixels).

[Nick Haffie-emslie]

You may want to check with the broadcaster about specifications before you get too deep into this project. Many will not accept uprezzed SD footage and are pretty specific about origination format requirements for HD.

[Nick Haffie-emslie]

And I’m pretty sure I’ve had this workflow go fine in the past… strange.

What’s interesting is that exporting ProRes of the exact same comp in the same dimensions and frame rate from AE works fine: FCP realizes it’s progressive without being told. Are we sure the QT file format has no field order flags for any codec? Perhaps additional metadata is included in the ProRes?

[Nick Haffie-emslie]

D’oh!

Thanks Jeremy!

So FCP looks at the frame size and frame rate and assumes it’s NTSC DV material, even if there is nothing in the file to indicate interlacing?