Kevin Camp
Forum Replies Created
-
I think you are looking or the time displacement effect. it can use a grayscale layer map to take samples from different times of a single clip (or comp).
to creative something similar to the ‘eye’ in your example, first create a new comp. add a new layer and add the fractal noise effect and then the mosaic effect (make adjustments to the mosaic ‘blocks’ and the fractal noise contrast so you have a fairly discernible block pattern).
bring that block comp into a the comp with your footage layer and add the time displacement effect to the footage layer. select the block pattern comp as the ‘time displacement layer’ and increase the max time displacement value and go from there.
Kevin Camp
Art Director
KCPQ, KZJO & KRCW -
i would try cutting the play button into 6 strips. then morph each strip.
or possibly morning the play button into a rectangle and animating venetian blinds to create the bars, and then another mask to make the bars different heights.
Kevin Camp
Art Director
KCPQ, KZJO & KRCW -
yep. no way i can think of to do that with radio waves…
you could do it with a particle system, or simply animating a circle (either shape layer or circle effect) and then animating the scale and color of the outline. then you can add the echo effect to create time-offset copies of that animation to make it similar to radio waves.
using the circle effect has the bonus of easily maintaining the line thickness (the shape layer method would need to adjust the line weight as it scales).
Kevin Camp
Art Director
KCPQ, KZJO & KRCW -
if all the layers were 3d, then your check box could control the arm’s z-position, moving it back or forward 1 or 2 pixels in z-space to move behind or in front of the body (or other layers).
assuming the body and the arm are currently at the same z-positon, a simple expression like this on the arm’s position property should do what you need:
if (effect(“Checkbox Control”)(“Checkbox”) == true) z=1 else z=-1;
value + [0,0,z]when the checkbox is checked, the arm will be in front of the body; uncheck, behind.
Kevin Camp
Art Director
KCPQ, KZJO & KRCW -
when doing a scale transition like this, it’s helpful to use a null that has the scale animation, and then parent the other layer (in the case text) to the null. this allows more flexibility in positioning the text separate from the scale animation.
what i’d do is add a null to the comp (layer>new>null object). leave the null in the center of the comp, and do a quick scale animation — like 100% to 1000% over a few seconds (you can adjust the animation later).
the create the text and parent it to the null at the beginning of the animation (where the null is 100%).
the text should now scale up with the null. make any adjustments to the scale animation (if it needs to get larger/smaller or change the timing, etc)
now, as the text is scaling up, if you need it to fly into a letter, you have two options. if it’s already pretty close to landing where it needs to, you might be able to simple nudge the text to get it where it needs to be, but have the text still look centered at the beginning of the animation. if it’s not close enough to make it seemed centered at the beginning of the animation, you can animation he position of the text so that it slowly slides from it’s centered position at the beginning of the animation, to a position that it needs to be in at the end of the animation.
Kevin Camp
Art Director
KCPQ, KZJO & KRCW -
there’s so little bear footage, it could have been rotoscoped in less time and for less money than it would have taken to set up a portable green screen in a bear exhibit, shoot it and key it.
the ‘bad video’ look might have been used to hide issues with the roto or compositing of the shots.
it’s possible that someone involved in the production new someone who had access to a bear so they could set up a green screen shoot, or found stock footage of a bear on a green screen, but i’m betting they roto’ed the bear from a more normal video shot.
Kevin Camp
Art Director
KCPQ, KZJO & KRCW -
the easiest way would be to use a particle generator that could can use a layer map as an emitter/producer, thus only producing pins with the layer map… trapcode’s particular can do this, as can particle playground (though it’s more complicated).
the negative side of these particle systems is that the layer map is really just used as a visibility map, in that particles do get produced outside that layer map, but you can’t see them… so if you truly need 300 pins with the map boundaries, it would be tough to get exactly 300.
you could use the same setup that you have now, but add a sampleImage() expression to the opacity of each pin… sampleImage() can be used to look at pixels in another layer, and base on what it sees you could have the layer turn on or off. this has the same drawback as the particle system/layer map, where you may have 300 pins in the comp, but maybe only 200 are visible because those 200 are within the map layer….
if you really need 300 pins, then I think i’d use your expression to get the randomness of most of the pins, but then manually move any pins outside of the map so they lay within the boundaries. simply adding the pin layer’s position into the random expression will allow you to do that:
seedRandom(3,true);
value + random([thisComp.width,thisComp.height])
you’ll likely need to zero all the pin positions by simply selecting all of them and setting one of the pin’s positions to 0,0. then simply drag any pins that are outside the map to be within the map boundaries.the only other way i can think of would be to incorporate sampleImage() into the position expression to test if it was within the boundary of the map, and if not, then change the random seed until it found a location that was. this could get computationally heavy as these expressions get calculated every frame, and you’d likely just get many pins that have the same random seed, and thus the same location. so i don’t think that’s a good route to take.
Kevin Camp
Art Director
KCPQ, KZJO & KRCW -
cc particle world might also work…
you’d need to import the 1000+ images into ae as either an image sequence or individually and create a comp, add the images and sequence the layers so that each image is a single frame….
importing them as an images sequence is a little easier if all your images are in a single folder. simply drag that folder into the project window of ae, in the dialog box that follows, set the ‘kind’ to footage and it will import the folder of images as a single image sequence.
create a new comp, add the image sequence and hide it. if your image sequence is shorter than the comp duration, you’ll want to do the loop trick. right-click the sequence in the timeline and choose time>enable time remapping. then option-(or alt)-click the stopwatch for the time remap property to enable expressions for that property, and type loopOut() in the expression field. drag the out-point of the sequence to be as long as needed. alternatively, you can right-click the sequence in the project window and choose interpret footage>main and set the footage to loop xx number of times — you’ll need to do the math to figure out how many time to loop. but if you change you comp length, you may need to change that loop setting, so i prefer the expression method…
create a new solid and add cc particle world. in the particle settings, set the particle type to textured square, and then in the texture settings select your images sequence as the texture layer and set the texture time to birth. a little lower in the particle settings, set the birth and death color to white (this will prevent particle world from tinting your images).
after that, it should be a matter of tweaking the other settings (particle size, producer size, birth rate, longevity, physics, etc) to get what you want.
Kevin Camp
Art Director
KCPQ, KZJO & KRCW -
i don’t think you can get the ae cache to go to any other drive other than the main drive… but i’m not sure that matters for speed.
with that 1tb fusion drive, you have a 128gb ssd that is fused with a 900gb spinning hd.
in theory, any writes to that drive should be stored first on the 128gb ssd, and transferred to the spinning hd as the ssd fill up. so (in theory), your ae cache should live mostly on the ssd after you’ve used ae for a bit.
you could use an external drive for your media cache (that location is definable in the ae preferences), and if the ssd drive you are looking at can achieve usb3 or thunderbolt2 data rates, then you very well could see an improvement in performance, though i have not tested that.
look on the internet for ae benchmarks (or external drive benchmarks) for iMacs… barefeats.com has a lot of mac benchmarks, and you may find something there.
Kevin Camp
Art Director
KCPQ, KZJO & KRCW -
if you want to composite this in an NLE (like Premiere), then, in AE, I would apply the matte to the snow layer and remove the video layer. Then render that out to a codec that supports an alpha channel (like apple’s prores 4444, lossless animation, or avid’s dnxhd).
import that rending into your NLE and and you should be able to place that in a video track over another piece of video and it should work the way i think you want it…
Kevin Camp
Art Director
KCPQ, KZJO & KRCW