VFX History

This is an old post I made on a long dead site I used to host. So in the interest of preserving history:

The Eye

One of the requests to me the other day was to discuss how to “step up”, move my effects look from “Joe-in-the-backroom” to “What-did-you-do-on-that”.

Tall order.

There’s a lot of discussion in my circle about The Eye. “Oh, yeah, he’s got a good eye.” “Who, Joe? Of course it looks like that, he has a tin eye“. We use the eye as a separator between someone who makes stuff that looks good, and stuff that just looks run of the mill.

So, I’m going to take the question as that – What is the eye, and how do I get one?

Honestly, I think the eye is a half-way mystical thing that is mostly focused on awareness. People who have the eye are very aware of images, of what the world looks like. They don’t just look, they SEE, they stare and think about how things really look, what things look good.

And here’s the awful truth about the eye. Some people will never really have it.

It’s not about some set of magic formulae you can chant and get the image. It’s not a cookbook. It really is an awareness, a sensitivity to the world around us.

Here’s the second awful truth about the eye, and one I find even more frightening:

Many people have the eye, but they never train it. A gift that is undeveloped, left alone and forgotten. They don’t train it because it’s hard, it’s something you have to constantly work on. That just saddens me.

Okay, let’s say you have the eye, or at least, you think you do. How do you train it?

Let me say again, what I tell you three times is true – the eye is about awareness. Look at stuff, and really SEE it. Look all the time.

Our business is focused on images, so it’s images you must obsess on. How does the light wrap around that tree? The light hitting that cat, what do the individual hairs look like? And how does that contribute to the total image of the cat? Analyze the things you are seeing. Think about light, moving and hitting stuff, energy flying around and finally coming to rest on some thing, draping it with light. How does that work?

Some things you can do to train the eye is to take up drawing, or photography. I’ve tried both. Personally, I’m not much of a draughtsman, meaning my drawings bear a closer resemblance to 4-year-old sketches, but that’s also about the time I stopped drawing. With more practice, I’d get good. I’m lazy and photography is easier for me, so that’s been my choice (and since I’m into geek photography like stereo & vintage cameras, that keeps my interest).

Alright, having said all that, here are some cookbook stuff. Most of what I’m going to tell you now is just opinion. As soon as I say it, someone out there is going to say, “I’ve tried that, it looks like ka-ka (a technical term you often hear professional image-makers use), what is he thinking?” Another group will write it down like scribes in secret manuscripts and trot it out twice a year to adoring acolytes. The truth is somewhere between that. Take it with a grain of NaCl.

Last thing I want to say about the eye is that you can’t just apply it to things you see, you have to apply it to your own work – and that is the hardest thing to do. When you look at your own image, you see all the stuff you did. No one else see’s that, only you. You have to take out the eye and really apply it to that thing you just made. See it, what does it really look like. Between you and me (and everyone else reading this), I know professionals of many years who still can’t do this. If they’re really professionals, they’ll take steps not to be judging their own work.


We love to soften our images. I think we do it, because some very important VFX houses tend to rely heavily on darkness and softness. We as artists run the risk of staring too long at the work of our peers and not looking at the real world.

Now, I think the blur node is my enemy. I think long and hard about every blur I include in my comp. I don’t like them. Here’s why:

I believe that we humans normally see things in pretty sharp focus (at least, as long as I’m wearing my glasses). I’m used to that. If I see something and I can’t focus on it, I get a little panicky – something must be wrong with me. So, softness, in some deep animal part of my head, is associated with being broken, and it introduces a little bit of tension.

But Rory, I see soft focus images all the time! Artists are always using it!

Yeah. Why? Because they want to introduce some tension, or they want to force the viewer towards something of interest in the image. How many times have you sat in a monster movie, the victim cowering in the foreground in sharp focus, while behind her, just beyond focus, something horrible is moving? Doesn’t it freak you out, you can’t focus on that thing behind them?

So why do you want to make parts of your image soft? I think the eye is going to go right to those soft areas, pick out the pattern and try to figure out why it doesn’t look right. Personally, I claim to be able to spot a certain famous software’s gaussian blur implementation. There is a certain characteristic to it that draws my eye and doesn’t look real.

I hear you blubbering now. “I love my blur! What do you mean I can’t use it!?!” I didn’t say that. I just think that blur is the easy answer. If you want better images, focus more on mixing than blurring.


We often use a blur when we really should be using a glow.

Light is not a straight-line, ruler driven thing. It’s waves. It bends, it folds around stuff.

This is my gag for introducing a “glow”. It’s a gag, not a real physical model, though the concepts are based on physics.

Take an image and apply a blur. Now, subtract (isub) the original image from the blurred image. Don’t swap this, or you’ll end up with a poor man’s unsharp mask.

The resulting image will be very dark. I think of it as the light that is leaking around and contaminating the darkest areas. Usually, I don’t want to apply this back to the image at 100%, so I throw a brightness on to control how much gets added.

I take this final image and add it back to the original image (iadd).

I’ve used variants of this technique to replicate the action of ProMist filters, as edge glows and wraps. I find it pretty useful, and more importantly, I’m not making the image softer, although I am affecting local contrast. All said and done, it’s more a color operation than a blur.

17 Feb 2003

Leave a comment

Your email address will not be published. Required fields are marked *