Archive for April, 2010

“Pop” Wave

Posted in Blogs, Evan's Blog, Projects with tags , , , , , , , , , , , on April 30, 2010 by EAbrams

I call this little number the pop wave. Looks pretty simple right? Just some things popping on across the frame in a controlled wave, and the off, and then on and ungulate before off again right? Wrong! This is some pretty cool stuff going on back there.

I’ll break it down for you. There’s only one thing changing here, and it’s a linear wipe. That’s right only a handful of key frames for this awesomeness! The rest is expressions. Those little lines of joy/frustration that make magic out of mole hills.

So the linear wipe, which arguably has some severe alterations to it (feather, fractal noise and the like) is the control layer. What it controls is the individual objects on top which are in comps so they can be swapped out with ease. The expressions are on the rotation for a random seed, and the scale to get that sweet pop.

The random rotation is just a random see generator frozen over time and a random number expression. That’s all that needs be said about that. Each copy made will have a unique number associated with it so cmd+d away!

The second is arguably the most complex thing I’ve dissected. It’s called the sampleimage(). Basically it takes a point on a layer and spits out the R,G,B, and Alpha. You can link that up to variables, and I prefer to make lots of variables in this, to make the function relevant. here’s how it looks for me:

targetLayer = thisComp.layer(“Control”);
samplePoint = transform.position;
sampleRadius = [1,1];
t =110* targetLayer.sampleImage(samplePoint, sampleRadius);
[t[1], t[1]]

Break it down!

This is applied to the scale parameter of the individual objects. The first line is a variable, you can call it anything but it will be the layer you want the sampleImage to sample or look at, it’s a “what?” qyestion. The next is the samplePoint, again you can call it x or y, and it’s going to be a place holder for where on layer you want to look it’s displayed as [x,y]. The Next is the sampleRadius or how big the sample point should be. I think it’s a circle that eminates from a point up and down the firs character and left and right the second character, so mine was 1,1 but the size is really 2 by 2. Now we put it together calling up an arbitrary variable t. and making it all come into the sampleImage telling it to:

110 times “on this layer”.sampleImage(“here”, “this big”);

Why 110? because the values this thing will make are from 0 to 1, a percentage basically. The 110 makes a value of 1 or pure white (R=1, B=1, G=1, A=1). The value is actually [1,1,1,1] and you can get something from it however you want really but I wanted to make the t 110 when it’s all the way on and 0 when it’s all the way off.

The final line is putting it all to work in the scale parameter that is displayed as [#, #]. In this case it’s the first part, [0] of the four part output. t[0] is the Red value. So the variable reads [“how much red”, “how much red”]

That’s that, the layer now references what’s at the same spot on the control layer and changes it’s scale from 0 to 110 depending on the colour value. Then that layer animates, which is easy to control and preview before you get hundreds of objects doing the “pop” wave.

Also if you just copy paste my code you won’t be learning so be sure you write it out and understand it best you can to make the most of it later on. Ususally when it would break during my coding it was because I was not giving it the right kind of variable, class 2 versus scalar for example as it can get confusing. If you have any questions, comments, or jeers let me know.

Particles or Character Animation

Posted in Blogs, Evan's Blog with tags , , , , , , , , , , , , , , , , on April 25, 2010 by EAbrams

So I was at it again, another two 10sec outputs of some effects work.

The first: an ink spray using expressions, some particles, a little of that fake 3d text and some images of ink splats. The think here is to ramp up the inherited velocity to really throw the ink from a single source when it get going faster by linking the speed to other properties as a controller. It looks kind of like that symbiot suit from Spiderman. I mean the cartoon not the mistake of comic to film adaptations. That was a jewel in the Fox Saturday morning cartoon crown.

Second: I was torn between which I wanted to focus on for the next “do x of y” grinding projects. To that end I made this second video with character animation. If you recall, we made a pretty strange and punchline-less video of a Molar Bear. I might as well embed that here so you know what I mean.

Now this goes under a heading that could either be “Mispronounced Words That Are Still Real Words in Picture Form” or “Rejected Pokemon Names”. However I envision a time when such poorly thought out and non idiomatic phrases like “Made out like bandits” are also give visual form. For those who don’t know, the phrase is supposed to be “made off like bandits” when people get away with something big and are gleeful about it. But I say the other one wherein two thieves make a big score and then score big… with each other… with tongue and hair action. To that end fire up the Laser Bream!

This is about 12 illustrator layers that all come to make the deadliest catch of all. A little key frames, some expressions (mostly time and wiggles) I think the wiggle my be the best expression ever invented. You pair that up with a slide control for both the magnitude and the frequency and you are in business. Both projects used that same technique somewhere can you tell where in each?

More 3d Projections

Posted in Evan's Blog, Projects with tags , , , , , , , , , , , , , on April 19, 2010 by EAbrams

So I thought I would just do 3d projection until I tired of it, then consider it added to my tool kit. I’m taking on my after effects training like grinding in WoW. So to become a better after effects guy I’m going to give myself quests like “create X of Y” and bring them to the blog. To that end I’ve created a few more 3d projections to get a handle on the technique and uploaded them to youtube.

The first I wanted to make a transition that would illustrate that I was making it into 3d. So I took the grid I was using to eyeball the placement and anaimated it on to a duplicate set of surfaces. Then a little transparency strobe, some linear wipes, and a little green tint I think give it a little matrix like feel. Like a computer is building the scene, which it is, only not at all that way. I love how so many movies show computers doing things that computers don’t do. Like coding in a 3d program (I’m looking at you swordfish). And though it is possible to code in almost every damn thing these days, it’s not always the best. Also CSI can eat it for their computer lies.

Next up I thought i would map some text to a 3d surface and try another way of scamming some depth from these shots. So for the text I just copy/pasted the orientation and position properties of a surface to a text layer and pushed the text out a few px. Then moved them around and rotated to fit like they could have been painted. In the future I’ll likely have something animated since it’s as easy as repeating the same process with a nested 2d comp. Now for the doors they are recessed into the building and to get that little recess i would have had to make two walls and a roof for each like a little mini hallway, or done some trickery with a little script. By referencing the layer above I set a few copies of a masked surface to offset by 5 on the z and then just hit command D and made a few till it fit the depth I wanted. You can you that to make some quick 3-d text with depth too if you can’t afford a 3d app or CS5.

Lastly I was noticing something intersting when I had textured planes and a small appature on my camera. By jacking up the blur I found it looked like shooting a miniature with the depth of field being so tight. So I thought I would apply that. I checked out some stock images on Stock Exchange ( and broke it down using the regular technique. Then I lined up some text and did a little move from title to title and a little particle action to boot. Everyone likes particles, especially dusty ones. I don’t think theyturned out like I would have hoped but I really havn’t taxed myself to learn more about the cc particle world yet and I likely should. Particles I find look the best when you layer many systems together and blend them, but that takes rendering power that don’thave when I’m already cranking out an extrapolated scene and asking the camera to look cooler.

So that rounds out my 3d projection phase. I think that should be enough to get everyone damn bored with making photos into 3d scenes. I’ve also included the photos below for your reference so you know where they came from if that’s at all interesting to anyone besides other after effects nerds. Also if you would like to have a go at duplicating my stuff let me know and I’ll see if I can guide you along in making this work for you.

GRMedia Logo

Posted in Blogs, Evan's Blog, Projects on April 5, 2010 by EAbrams

So Tim and I have been going back and forth on some logo ideas. It started out something like this:

Many terrible Proofs

But then I had one good one:

It's better than nothing!

It's better than nothing!

And Tim called me up at my other job to exclaim how he thought it looked like an Elephant, and wanted more elephantine qualities about it. I think when I asked him why more elephant he countered with: “Why is Linux a Penguin?”

To that end new proofs were created:

Elephant proof!

I liked the bottom one but Tim did not and wanted the top one but with the MEDIA underneath and the P removed. At this point I balked for a while, became obsorbed in something else and now it’s the easter weekend and I can present to you the logo as I think Tim wanted it… also with some things I’ve always thought a logo needs. But don’t take my word for it.

I think this is what Tim wanted?

I know this is what I wanted.