[Greg Neumayer]

I’ve had good results by simply extending the mask as earlier suggested. Don’t modify your mask. Let them be dead-on perfect, but then in your mask settings, add maybe .5 pixel to the “mask extension” and see if that cleans up your edges without adding any visual overlapping when viewed.
-Greg

Antifreeze Design
https://www.antifreezemotiongraphics.com

[Greg Neumayer]

Oops. Yes. Sorry.
My brain is still using Cinepak encoding.
-Greg

Antifreeze Design
https://www.antifreezemotiongraphics.com

[Greg Neumayer]

It’s important to remember that any time you render with a method that incorporates lossless compression, you’re losing data. If you’re working on DV anyway, you don’t want to lose any more. If you have the ability to import it back in, render in a lossless format. This will allow at least that part of the chain to not lose additional data. As far as compression goes, I like to use quicktime using the PNG compression, since at it’s best quality, it’s lossless, but can still save you 40%-60% depending on your imagery.

-Greg

Antifreeze Design
https://www.antifreezemotiongraphics.com

[Greg Neumayer]

I’ve wondered the same thing for a project of mine. I’ve been meaning to run down a couple options:

1. Look into the “alpha add” transfer function. I think this may have been developed to help with this kind of situation where you have numeric masks butted against each other, but need to have seamlessness when they’re all lined up.
2. Is this what Trapcode’s Echo plug is about? If not, I’d love for someone to post a lead to a plugin or method for creating card flip, slide, etc. transitions built from a vector “puzzle” that I create in Illustrator. Do we really have to have a separate comp for each band? Is there a way to do it with expressions?
3. You might try using an animated displacement map to move the elements: Create an image that has the vertical stripes you want as bands, starting with a white band on one side, and a dark band on the other. This will be your mid-point. Then, animate this image with something like curves so that it starts completly white, then fades to it’s original look, then fades to completely black. Put this animation in your comp with your “girl” image, turn it off, but add “displacement map” to the girl and specify your bands for your displacement image. Then, displace vertically but not horizontally. The only caveat to this method is that you probably won’t get the nice easing that they used for each band.

Please post your solution. Thanks!
-Greg

Antifreeze Design
https://www.antifreezemotiongraphics.com

[Greg Neumayer]

I’ve wondered the same thing for a project of mine. I’ve been meaning to run down a couple options:

1. Look into the “alpha add” transfer function. I think this may have been developed to help with this kind of situation where you have numeric masks butted against each other, but need to have seamlessness when they’re all lined up.
2. Is this what Trapcode’s Echo plug is about? If not, I’d love for someone to post a lead to a plugin or method for creating card flip, slide, etc. transitions built from a vector “puzzle” that I create in Illustrator. Do we really have to have a separate comp for each band? Is there a way to do it with expressions?
3. You might try using an animated displacement map to move the elements: Create an image that has the vertical stripes you want as bands, starting with a white band on one side, and a dark band on the other. This will be your mid-point. Then, animate this image with something like curves so that it starts completly white, then fades to it’s original look, then fades to completely black. Put this animation in your comp with your “girl” image, turn it off, but add “displacement map” to the girl and specify your bands for your displacement image. Then, displace vertically but not horizontally. The only caveat to this method is that you probably won’t get the nice easing that they used for each band.

Please post your solution. Thanks!
-Greg

Antifreeze Design
https://www.antifreezemotiongraphics.com

[Greg Neumayer]

If I were doing it from Flash, I’d work at 30fps, and then try to use Flash 8’s new alpha abilities to export a quicktime with the alpha still intact. I haven’t actually done it, but Flash now supports import of qt with alpha, so hopefully they support export as well.

That said, if you have Illustrator or some other vector app, creating your artwork may be a lot easier in one of these apps. Flash is powerful, but their drawing tools aren’t really trying to compete with seasoned vector packages, so fortunately they’ve made import easy with the ability to send AI layers to Flash layers, etc.

-Greg Neumayer

Antifreeze Design
https://www.antifreezemotiongraphics.com

[Greg Neumayer]

I think you may be adequately compensating (with the scaling) for something that you don’t need to compensate for if you set your comps up correctly.

1. AI files are going to come in as square pixels. Once you’ve imported your file, go to your footage interpretation (right-click on the footage in the proj window) and interpret the art as square, if it’s not already. (photoshop now has the ability to work in non-square mode.)

2. Use your artwork in either a square or non-square (.9) comp.

3. If you want to work in square pixels until you’re ready to output, that’s fine. Work at 648×486 (1.0). (I like to set myself up with a template in AI that is 648×486 so that I can see a basic outline of my comp window.) Then make your master comp for your final render that is at .9 (720×486). Your square comp(s) that you put into it won’t need adjusting, and AE won’t be doing scaling and resizing on your art because it was oversized at 720×540.

—One other note about the line art from AI. Make sure that if you’ve used the collapse layer button on your comp that contains the art, that you also use the collapse button on subsequent layers until you get down to your source ai art. If you don’t have them collapsed all the way down, you’ll get rasterazation somewhere along the line. (I think there also may be a pref that also affects sub-layers so that you don’t have to walk all the way down the nested-comp-line to make sure they’re all collapsed.)

-Greg Neumayer

Antifreeze Design
https://www.antifreezemotiongraphics.com

[Greg Neumayer]

Cool. I got Knoll about a year ago, so I may need to upgrade. I’ll have to try it.

Just out of curiosity, I’m using a fractal as a luma matte for my (Trapcode) Lux layer, and offsetting the turbulence by about .5 of my camera distance to the subject to create varying levels of opacity in my light cone (i.e. smoke-look). It’s a gimmick for sure, but works pretty well if I don’t move the camera too much.

Is this the best way to approach creating varying levels of opacity in Lux? Is there a better way to get 3D light cones that have some substance?

Thanks,
Greg

Antifreeze Design
https://www.antifreezemotiongraphics.com

[Greg Neumayer]

Cool. I got Knoll about a year ago, so I may need to upgrade. I’ll have to try it.

Just out of curiosity, I’m using a fractal as a luma matte for my (Trapcode) Lux layer, and offsetting the turbulence by about .5 of my camera distance to the subject to create varying levels of opacity in my light cone (i.e. smoke-look). It’s a gimmick for sure, but works pretty well if I don’t move the camera too much.

Is this the best way to approach creating varying levels of opacity in Lux? Is there a better way to get 3D light cones that have some substance?

Thanks,
Greg

Antifreeze Design
https://www.antifreezemotiongraphics.com

[Greg Neumayer]

Ah, perfect. Thanks. I was actually almost there with the same basic process from the tutorial from Trapcode:
https://www.trapcode.com/in_action/shine_gobo.html

but hadn’t yet considered the cool distance/brightness idea from your link.

Here’s my line of code I used:
src = thisComp.layer(“null for flare1”); src.toComp(src.anchorPoint);

Thanks!
-Greg

Antifreeze Design
https://www.antifreezemotiongraphics.com