[Barend Onneweer]

Hi,

Thanks for thinking along.

Actually what you’re doing is basing rotation on the accelleration, but by coincidence I just stumbled upon that example at Base80 and it did help me get to a solution, I think.

I’m using the Position Velocity > Vector2Reals into a Math Divide into the object rotation.

I need to tinker a bit more to see if it does everything I need.

Barend

Raamw3rk – digital storytelling and visual effects

[Barend Onneweer]

Okay, I’ll see if I can add some ideas.

I think what I’d need is some Xpresso that aligns the object to it’s motionpath – but the amount would be triggered by the speed of movement.

On a descriptive level I think it might work like this:

1) Calculate motion direction by comparing current position to previous position.

2) Compare current (keyframed) rotation with the motion direction and calculate the transform needed to align the object fully to it’s motion path.

3) Calculate speed of movement and use it to control the amount of transform that is mixed in.

So the faster the object moves the more it will be aligned to the motionpath.

But I’m lost in trying to build this…

Raamw3rk – digital storytelling and visual effects

[Barend Onneweer]

Just to confirm: I can’t load any of the Arri LUTS in Scratch, because the supplied LUTs don’t have headers.

I’ve actually been in touch with LightIllusion about them and LightIllusion doesn’t load the LUTs either. Seems like Arri placed a mostly useless set of LUTs – to most of us anyway.

Barend

Raamw3rk – digital storytelling and visual effects

[Barend Onneweer]

If the overall color balance looks okay, I’d look at increasing the saturation – this will bring out the color a bit more.

Barend

Raamw3rk – digital storytelling and visual effects

[Barend Onneweer]

First guess would be to check if THE framerates of background and foreground are the same in CS4. It could be that CS3 has slightly different interpretation rules.

Barend

[Barend Onneweer]

Like Scott says: if you shoot the TV without picture on it you get the reflection. Then composite your picture in using “Screen” blend mode and tweak to taste.

If your camera is moving you may want to make sure you’ve got something to track. 4 small dots in the corners of the screen will do.

Barend

Raamw3rk – digital storytelling and visual effects

[Barend Onneweer]

A couple of things come to
mind:

1) Is the footage meant to be 4:3 or 16:9? AE may be doing an anamorphic ‘unsqueeze’ to stretch the image to 16:9. By default the scaling quality is low so you’ll see subtle blockyness. This is just the preview quality and if you disable the Pixel Aspect Ratio correction in the Comp Window the image should look better.

2) Is the footage interlace or not? AE may be Separating fields which will also result in reduced resolution in the Comp Window. If the footage is progressive disable fieldseparation in the Footage Interpretation Window.

3) Is your preview resolution set to 100%? (dropdown in Comp Window)

Hope this helps,

Barend

Raamw3rk – digital storytelling and visual effects

[Barend Onneweer]

It’s a really wide subject, but an interesting question.

Personally: Often I’ll have an idea of the look I want to achieve and the way to get there. I’ll start with a couple of steps that I think may work. Sometimes this leads me directly where I want to be, and sometimes it doesn’t even get me close. Thus starts the tinkering and
tweaking, trying out different combinations of effects,
in different orders etc. It’s a fairly organic process that can be a bit hit-and-miss. But through this process you learn a lot about the possibilities of your tools, and it allows for what Bob Ross used to call “happy accidents”. You stumble upon unexpected results from combinations of effects. And if it’s not what you need right now, it may be something you can use tomorrow.

It’s a bit like cooking. As a cook you should know the basic taste of all ingredients. But cooking is something else. And coming up with new recipes is a long process.

That’s how it works for me anyway.

Barend

Raamw3rk – digital storytelling and visual effects

[Barend Onneweer]

In that case you’ll need to reproduce the camera movement in 3D space and you’ll need a 3D floor layer in AE.

It may seem like a small difference, but 3D tracking and matchmoving isn’t all that straightforward to get right.

If you’re planning a small push just to get a subtle tension in the shot, you might consider shooting locked off and doing a small push-in digitally. It’s not the same but it does have some effect, with minimal effort.

For 3D tracking you’ll need markers on the floor too, and not too few. You’ll track the 3D camera move in an application like Boujou, Syntheyes (most affordable), PFTrack or Matchmover. Then export the camera solve to After Effects and place it in your virtual set.

If you’re interested in these techniques and have time to investigate and experiment, go for it. Otherwise, get someone else to do it or fake it per my earlier suggestion.

If you’re going to do the push-in digitally it helps if you shoot at higher resolution than the delivery format. Shooting Red @ 4k gives you some room when you’re finishing in HD.

Raamw3rk – digital storytelling and visual effects

[Barend Onneweer]

Hi,

For tracking scale you need two markers. In AE’s tracker controls you can select a combination of scale, position and rotation tracking. This should be fairly easy as long as your two markers are never obscured by the models.

This is assuming that you are not digitally adding in the floor, because then you’ll need a more complicated approach since you’ll have parallax. You ‘d need 3D tracking.

Barend

Raamw3rk – digital storytelling and visual effects