[Scott G]

so no one knows??

[Scott G]

there’s a fantastic script at http://www.aeenhancers.com that will duplicate a layer’s effects and link them back to the original layer, so if you change it on the original layer, all the “child” effects change as well. brilliant!

[Scott G]

there’s a fantastic script at http://www.aeenhancers.com that will duplicate a layer’s effects and link them back to the original layer, so if you change it on the original layer, all the “child” effects change as well. brilliant!

[Scott G]

similar to what you say, but also kinda, different, at the moment some of the 3d guys render 3 or four mattes into one file, and we take a single channel (r, g, b, or a) for a matte. but that doesn’t cover all objects in a scene, so i was looking for a way to have the 3d render one file (or sequence) and we’d be able to pull any matte we wanted from it.
will look into combustion more, since as above it states there’s no aliasing on the id matte. 🙁

[Scott G]

use the lenscare filter set from http://www.frischluft.com. amazing depth of field effects, with or without a depth matte.

[Scott G]

trapcode doesn’t have a fog preset that i could see, i had to modify te smoke preset and use layers. does anyone have any trapcode presets they’re willing to share? 🙂
i’d put my foggy one up, but it’s not really fog, just puffs that i layered over other things. i didn’t try to create a flowing fog “bank” with trapcode. has anyone? do you think it would work?

[Scott G]

i just made some fog for use over water, but it’s essentially the same thing. my screen was pretty much black, you couldn’t even really tell it was water.

anyway, i used a combination of
a) fractal noise
b) smoke / steam that i filmed then comped in
c) trapcode particular.

A-
the fractal noise gave me “banks” of fog. i made 4-5 comps with different types of fractal noise. i think from memory i used
1) fractal type: basic / noise type: linear / contrast 220 / brightness -15
2) fractal type: dynamic twist / noise type: spline / contrast 220 / brightness -15
3) fractal type: basic / noise type: linear / contrast 200 / brightness -15
4) fractal type: threads / noise type: spline / contrast 120 / brightness -15

different types of noise give the appearance of depth when you layer them. the dynamic twist gives a swirly movement, the basic noise gives patchiness.
try scaling these at different sizes to aid the appearance of depth.

i also played with digieffects delerium fog factory, but only used that in one layer. the near-far fade you can do with this looks nicer than a simple soft mask.

B-
film steam or smoke over black, then comp. it doesn’t have to be pure black, as to make it look like fog instead of smoke i darkened it up and blurred the heck out of it, so no need to be precious when filming.
some footage items i used were stock, and these were actually invaluable as i couldn’t create a similar effect with home made smoke or steam. eg, dry ice behaves much differently to smoke or steam, and obviously i can’t film that myself. that was only one layer in 7 or 8 though, so you can do without it.

C-
using the “smoke” preset of trapcode particular, i modified the properties to get it wafting in the right direction with the right patchiness and turbulence. i could write entire paragraphs about that, but i won’t, you’ll have to figure it out, and every case is different. after half an hour i had several layers of this i could place where needed, for slow wafts, small pufts, or litte displaces.

FINALLY
it helps if you have a subject on a transparent background. whether it’s created this way, or you need to render separate passes or even rotoscope, it helps give the illusion of depth. put some layers behind the character, and some layers over it.

if you want to get really tricky you can add turbulent displaces to some of these layers, or isolate the turbulent displace with a masked adjustment layer.

all this was REALLY render intensive. 6hrs on a g5 for a 30second comp. sheesh! but since i don’t have 3d skills or an inferno, i’ve gotta make do.
hope that helps!

regards,
scott.

[Scott G]

ARGH. i totally missed the “colourize” command in trapcode shine. sorry, was up all night again.
problem solved, many apologies for wasting your time. 🙂

[Scott G]

thanks for the help guys, i suspected afx was holding the whole huge comp in ram, but wasn’t sure. i prerendered quite a few of the layers, which did help, because even though they’re still HUGE, the effects don’t need to be rendered every frame, so much less processing needed. was quite effects intensive – couldn’t collapse because the effects went haywire.

i’d altered the secret prefs before as well in an effort to reduce the number of errors i was getting, but sadly they didn’t help that much. some segments i couldn’t even preview at full res after a purge, let alone render them, so altering the maximum memory usage / flushing the cache every x framers didn’t do much for me.

last night i tried rendering to a tiff sequence rather than a lossless quicktime, and oddly enough, no errors. perhaps because it didn’t have to build a large qt, only individual frames. who knows.
well, the combination of patient pre-rendering and a tiff-sequence worked for me. so if people are getting out of memory or image buffer errors, in addition to all of the above, i recommend trying to render a sequence rather than a quicktime. seems to work.

[Scott G]

thanks for the help guys, i suspected afx was holding the whole huge comp in ram, but wasn’t sure. i prerendered quite a few of the layers, which did help, because even though they’re still HUGE, the effects don’t need to be rendered every frame, so much less processing needed. was quite effects intensive – couldn’t collapse because the effects went haywire.

i’d altered the secret prefs before as well in an effort to reduce the number of errors i was getting, but sadly they didn’t help that much. some segments i couldn’t even preview at full res after a purge, let alone render them, so altering the maximum memory usage / flushing the cache every x framers didn’t do much for me.

last night i tried rendering to a tiff sequence rather than a lossless quicktime, and oddly enough, no errors. perhaps because it didn’t have to build a large qt, only individual frames. who knows.
well, the combination of patient pre-rendering and a tiff-sequence worked for me. so if people are getting out of memory or image buffer errors, in addition to all of the above, i recommend trying to render a sequence rather than a quicktime. seems to work.