ADOBE AFTER EFFECTS: Forum Expressions Tutorials Creative Cloud

When to interpret footage in AE: Please help!

COW Forums : Adobe After Effects

<< PREVIOUS   •   FAQ   •   VIEW ALL   •   PRINT   •   NEXT >>
Bryce McNamara
When to interpret footage in AE: Please help!
on Oct 15, 2009 at 1:31:24 am

Hi there,

I'm fairly new to After Effects and to these forums, but would really appreciate some help in clarifying this particular issue for me.

*Note the following:

- I work mainly with DV PAL footage
- The final animations are destined for broadcast

If I were working on an animation in a PAL D1/DV Comp (for instance) and I chose to incorporate DV PAL footage into this animation in whatever way (3D, applying numerous effects to the animation, etc.) would I:

1) Need to initially interpret this footage prior to animating with it?
2) Need to interpret this footage according to it's intrinsic format (ie. DV PAL, lower-field first, etc.)
3) Would I need to de-interlace this footage instead?

I want to know what the best way to handle the footage would be so that the final animation is crisp and clear, without any footage STUTTERING or field issues when viewed on a television monitor.

I would really appreciate any assistance in clarifying this issue for me!

Cheers!

Bryce
Melbourne, Australia


Return to posts index

Todd Kopriva
Re: When to interpret footage in AE: Please help!
on Oct 15, 2009 at 4:47:36 am

Here's the first and most important thing to know about footage interpretation: It's all about telling After Effects about what the characteristics of your footage are. If After Effects has guessed right, then you don't need to do anything more with footage interpretation.

For DV PAL footage, After Effects should get everything right, without manual interpretation steps.

One of the things that it will do is separate fields. This is crucial if you're going to do much of anything in After Effects with layers based on the footage.

You should also check to make sure that you render settings are set to render fields as appropriate for your output format.

BTW, the biggest problem that I see with animations shown on devices that use interlacing is twitter---the flicker that comes from vertical detail (like thin horizontal lines) alternately disappearing and reappearing from one field to the next. There are some tips for avoiding this here.

---------------------------------------------------------------------------------------------------------
Todd Kopriva, Adobe Systems Incorporated
putting the 'T' back in 'RTFM' : After Effects Help on the Web
---------------------------------------------------------------------------------------------------------
If a page of After Effects Help answers your question, please consider rating it. If you have a tip, technique, or link to share---or if there is something that you'd like to see added or improved---please leave a comment.


Return to posts index

Bryce McNamara
Re: When to interpret footage in AE: Please help!
on Oct 15, 2009 at 6:53:41 am

Thanks Todd,

That really cleared up a lot for me and those links were very helpful. If I could just ask something in addition:

Usually when I work with graphics in AE, I tend to create in AE and then render the animation out as a Photo-Jpeg (or Animation if exporting Alpha) and then I import that Photo-Jpeg file into a DV PAL FCP timeline.

Firstly, do you see this method as problematic at all - it has been working well for me thus far; however, I'm wondering if using DV PAL footage in an AE comp that is then rendered out as Photo-Jpeg and re-imported into a DV PAL timeline would possibly present any problems?

Also - If I use DV PAL footage in an AE comp and render out as Photo-Jpeg/Animation - would I still need to change the field order settings to lower-field first to match the *final* FCP output?

I really appreciate your helping me grasp this issue!

Regards
Bryce


Return to posts index


Steve Roberts
Re: When to interpret footage in AE: Please help!
on Oct 15, 2009 at 12:37:54 pm

If I may cut in ...

When making stuff in AE, I generally render to the codec of the editing timeline. So if I were eventually going into a PAL DV FCP sequence, I'd render to the PAL DV format. Then the clip just drops onto the FCP timeline quite nicely. There's no advantage in rendering to another format, since FCP would convert it to the DV codec anyway, to match its sequence settings. However, there aren't any problems with your method, just an extra rendering step in FCP: that takes time, and there may be a bit of degradation. But really, it's unnecessary.

If I may generalize, AE people usually render to an intermediate lossless codec if they don't know where the render is ending up, or they don't have the editor's codec on hand, or if the piece may be ending up in multiple formats, multiple editors, etc.

Fields and codecs have nothing to do with each other. You can render with or without fields in any codec in AE ... though for some codecs (like for the web), it would be pointless. So if the final output requires fields, render fields no matter what the codec is.

I hate fields. Gotta say it. Too much trouble. :-)



Return to posts index

Todd Kopriva
Re: When to interpret footage in AE: Please help!
on Oct 15, 2009 at 3:59:16 pm

Steve is the voice of experience. ;-)

---------------------------------------------------------------------------------------------------------
Todd Kopriva, Adobe Systems Incorporated
putting the 'T' back in 'RTFM' : After Effects Help on the Web
---------------------------------------------------------------------------------------------------------
If a page of After Effects Help answers your question, please consider rating it. If you have a tip, technique, or link to share---or if there is something that you'd like to see added or improved---please leave a comment.


Return to posts index

Steve Roberts
Re: When to interpret footage in AE: Please help!
on Oct 15, 2009 at 4:39:19 pm

... unfortunately. ;-)

Todd, being a clever fellow, maybe you could confirm this description of how AE renders each frame of a movie:

1. according to the Render Settings, a high-quality frame is created, with no compression and stored in some kind of buffer. This is like a "pure" or "ideal" image. This is also where fields are created.
2. That frame is then compressed or altered according to the output module. If the Output Module is Lossless, TGA or similar, the "pure" frame is basically left untouched. Basically.
3. That altered frame is stored along with the previous frames, awaiting a Quicktime wrapper if necessary. If rendering to a still sequence, each still frame is finished once it's rendered.

Is that about right?



Return to posts index


Todd Kopriva
Re: When to interpret footage in AE: Please help!
on Oct 15, 2009 at 5:10:11 pm

That's pretty close, Steve. I have a plain-English description of the process in the first few paragraphs of "Rendering and exporting overview".

Here's a somewhat different version that addresses your question more directly:

Each frame is rendered into an uncompressed buffer based on the render settings. This is the hard part that can takes multiple minutes per frame on a heavy composition.

These frames are handed to the output module, where each frame is cropped, encoded, and otherwise processed based on the output module settings. This tends to be the easy part---which is why it's sooooo valuable to note that you can attach multiple output modules to one render item; you can render once and output to multiple formats, etc.

Now here's a wrinkle: If the encoding is done by the Adobe Media Encoder (which happens under the hood for MPEG formats, FLV, and a few other things), the ouput module does some processing and then hands its output to AME for encoding. This is where some mismatches in frame sizes and such come in if you don't follow the instructions in the Comments field of the Output Module settings dialog box.

Regarding wrappers: Yeah, that's basically it. The video data is encoded and written to what is essentially a track inside of a wrapper---whether QuickTime or F4V or whatever. The container is created early in the process, though; it's not as if the video tracks are created and then a wrapper is put around them.

Regarding lossless encoding: The frames are very much "touched". They are just touched in a way from which all of the image information can still be retrieved. For example, I can encode a one-hour movie of a 1000x1000-pixel pure red image by saying this: "Repeat [255,0,0] for each pixel for 1000 horizontally and 1000 vertically for each frame from 1 to 108,000 frames." That's a highly efficient lossless encoding of a movie that would take hundreds of gigabytes to express without compression. The Animation and PNG codecs---and some others---have high-quality settings that do a lot mathematically to the image data, but all of the data is recoverable.

---------------------------------------------------------------------------------------------------------
Todd Kopriva, Adobe Systems Incorporated
putting the 'T' back in 'RTFM' : After Effects Help on the Web
---------------------------------------------------------------------------------------------------------
If a page of After Effects Help answers your question, please consider rating it. If you have a tip, technique, or link to share---or if there is something that you'd like to see added or improved---please leave a comment.


Return to posts index

Steve Roberts
Re: When to interpret footage in AE: Please help!
on Oct 15, 2009 at 7:03:02 pm

Thanks, Todd. Good info.



Return to posts index

<< PREVIOUS   •   VIEW ALL   •   PRINT   •   NEXT >>
© 2017 CreativeCOW.net All Rights Reserved
[TOP]