[Brendan Coots]

I used keylight exclusively when I worked on Sky Captain and the World of Tomorrow, and I know that studio used it on Day After Tomorrow, Hellboy and others as well. So to answer your question, YES it is a capable keyer. But I frequently have to divide out the actor into head/body/other as three seperate key layers to get ideal results, and even then some roto and other techniques are needed at times.

As for the grainy issue, I haven’t seen that in high-end footage, it really shouldn’t be doing that. What format/codec is the footage delivered to you in? How has it been pulled from tape? If something is being done that compromises the quality of the footage before it gets to you, that needs to be fixed first.

it sounds like you understand the basic useage of Keylight so I won’t go into that, but Andrew Kramer has a really good overview if you’re interested:

https://www.creativecow.net/articles/kramer_andrew/colorkey/index.htm

If you have large budgets as you’ve indicated, you might want to convince your shop to look into Primatte, it really works great on everything from DV to film plates. Render times are much longer than Keylight, but it’s almost always worth it.

[Brendan Coots]

Try listing it at craigslist.com. It’s free if you choose the Gigs>Creative category.

I tend to find decent contractors for my studio there.

[Brendan Coots]

Try applying Fast Blur to them, maybe 1-3 pixels or so. Once it’s on a TV set this won’t degrade the sharpness as much as you might think, and it’s a commonly used tactic.

[Brendan Coots]

I should have said this in my last post, but I have also noticed the greenish tint to Sorenson encoded files. Between that and the quality-to-size disadvantage compared to other codecs, I stopped using it long ago. Technical explanations aside, Sorenson is a fairly old codec and you’d probably be better served rendering to h.264, MPEG4 or even FLV.

If you have to use Sorenson for some reason, to my knowledge the only thing you can do is what you mentioned, tweaking the color settings. Needless to say that’s not a good solution for anything that will need accurate color rendition.

[Brendan Coots]

Probably not the exact answer you’re looking for, but I always recommend that people NOT render to compressed formats out of After Effects. Across the board AE does not do a very good job with compressed formats, and I have seen side-by-side comparisons that pretty much prove AE’s compressed output almost always results in larger files and lower quality.

A good workflow solution is to render out of AE to the Animation Codec or similar, then use a third-party application to compress the file to Sorenson or whatever you need. There are many free/cheap apps out there that do this. On the PC I have grown fond of a little app called SUPER. It’s free and compresses to an extremely wide variety of formats:

https://www.erightsoft.com/SUPER.html

.

[Brendan Coots]

Yeah, it really seems like the primary spec advantage of the Pro is the graphics card. Is that worth $1,000 more? If you need to to use Studio 2 or Maya (within OSX), you really have no choice.

A bit disappointing, and I can only guess that this was an intentional move on Apple’s part to push sales of the higher priced Pro model, otherwise they would have offered a video card upgrade for the standard Macbook.

[Brendan Coots]

After Effects is most widely used for:

– motion graphics
– compositing
– visual effects work
– color correction
– video editing/manipulation
– time manipulation (slow-downs and speed-ups)
– Basic cartoons/animation
– Titles

C4d is most commonly used for:

– 3D animation
– Some motion graphic elements

Based on this rough and dirty list, there is no comparison between the two in terms of capabilities. C4d is a great 3D app, but it won’t ever be a good replacement for After Effects any more than AE, which has very limited 3D tools, will be a good replacement for C4d.

If any comparisons were to be made, it would more likely be comparing a program like Fusion or Combustion to After Effects.

[Brendan Coots]

You can’t capture with AE, and the way you are working is how most people do it. You are right to be concerned about the quality loss of exporting to AE, then rendering from AE and bringing that file BACK into FCP. That’s a lot of pixel-trashing, and while using a 10-bit codec (like Blackmagic 10-bit) for these transfers/renders will help to minimize quality loss, you pay for it in hard drive space, heavy files taking longer to work with etc. etc.

There is an After Effects plugin from Automatic Duck called Pro Import AE that allows you to import your entire FCP project file into After Effects. It uses the original source footage like FCP does, and basically recreates your FCP edit on the After Effects timeline, including fades, dissolves and most effects. This way you eliminate that render from FCP, and can finish projects in After Effects.

https://www.automaticduck.com/products/piae/index.php

.

[Brendan Coots]

One of the better new features of After Effects CS3 is its support for multiple processors. Adobe recommends at least 2GB RAM per processor for the functionality to work properly, and I can confirm that (working on a quad-core mac with only 3GB RAM didn’t get all cores churning at all).

As such, I wouldn’t get anything less than a dual-core machine, ideally one with 4GB RAM. If you have to use Windows, you might want to look into Windows XP 64-bit which supports more RAM.

If nothing else, I would avoid Vista like the plague. Not only are there all sorts of problems being reported with it, but it is a complete resource hog. Ideally you want every ounce of RAM and CPU power going toward these resource hungry applications, NOT the OS.

[Brendan Coots]

Position properties like this are stored in an array formatted as position[0], position[1] and position[2]. You can apply a basic expression to your 2D effect’s X and Y properties to link them to the null’s X and Y values. The nice thing is that it’s much easier than it sounds.

On the timeline Alt+click the stopwatch icon for the X value of your 2D effect, then enter the following expression:

thisComp.layer(“Null 1”).position[0]

For the Y value of the 2D effect, use this expression:

thisComp.layer(“Null 1”).position[1]

The above expressions assume your null is named “Null 1” so change that wording accordingly.