[Stuart Elith]

Ok, when you say that when you apply the key to just one clip, the rest of them get keyed as well, you’re losing me.

If you apply an effect to one clip, it only applies to that one. You must be doing something slightly wrong… to clarify, if you have the whole video in one layer, then split that layer (so that you now have 2 layers, with part on each), both those layers will be separate (so they won’t both react to a keyer or mask. If you are getting results other than that, maybe post us a screenshot so we can see what you’re doing.

[Stuart Elith]

One difference in the camera data is the co-ordinates scale… not sure of the details, but people using Boujou (software for 3D tracking) have to go through a process to use the camera data (in Maya format) in AE.

Here is a link about that which may be similarly useful for you.

https://finthrow.blogspot.com/2007/02/3d-camera-tracking-data-import-into.html

[Stuart Elith]

My method would be to render at full resolution out of AE and then do your compression afterwards. The reason is, often our projects/renders from AE take quite a long time as they start becoming complex, and for web stuff I often find that I want to tweak the compression settings quite a lot to get the optimum quality/size balance.

So if you try and get it right from AE, you will be re-rendering your comp many times. If you render out the full quality one, you can try lots of different settings in your compression/encoder software (i haven’t used Adobe Media Encoder much but it probably works well).

[Stuart Elith]

Hmm that is strange, it’s frustrating when stuff like this happens without apparent reason.

I’m just here to chime in that it IS possible, because i have various dual monitor workspaces, and the shortcuts are maintained. Not that that’s any help to you really, but…

[Stuart Elith]

Ahh right, well that’s a very different problem. Not sure if you can solve that one very easily… certainly, I don’t think Particular respects that Collapse Transformations switch.

One option may be to make your particular bubbles with the original image, then precomp that and have 2 instances of it, one with each of the blending modes you need. Then the blending is done in the final comp rather than the precomp, so they should work. I think..?

[Stuart Elith]

Yep, you can do this… you need to leave the precomp on the Normal blending mode, and activate the Collapse Transformations switch, which kind of passes the precomp attributes such as vector text, 3D layers, and blending modes, through to the main comp.

Read about it in the online help if you need more details 🙂

[Stuart Elith]

Even if the camera (rather than the object) is moving, motion blur is still calculated. I just did a quick test and it does work (i wasn’t actually sure myself). Just make sure you have motion blur enabled for the layers you want to blur, and the overall MB switch for the comp.

[Stuart Elith]

It sounds like you aren’t RAM previewing your work… if you try and play with the spacebar, you will find almost ANYTHING is slow/not full frame rate. AE isn’t designed to run like that, it’s for working with more complex setups so isn’t optimised for basic playback.

If you don’t want to ram preview, you could reduce the display resolution (found at the bottom of your Comp window, it’s a pulldown menu with Full, Half, Quarter and Custom). Basically it reduces the resolution so previewing is quicker (it’s not destroying your actual work though, it will render fine at the end).

[Stuart Elith]

You actually answered your question… it’s a matter of tracking.
If you are looking to do the 3D style stuff including side-to-side rotation you may need to do a 3D track, which is more than the built-in tracker can handle (though perhaps Mocha in CS4 can, i’m not sure). Making something just stick to a person’s movement on a 2D level is very achievable within AE.

For the 3D stuff, i think my workflow would be :

(1) Do a 3D track (can sometimes be done within a 3D program like Maya, or with another piece of software like Boujou).
(2) In a 3D program, create your text, which will follow the tracked camera movements perfectly. Render it out in whatever style you want.
(3) In AE, you will just need to integrate it into the scene with color correction, masking/rotoscoping.

But as you can see this does require more than just AE. Stuff like 0:30-0:40, 0:51 are just 2D so are much easier.

[Stuart Elith]

Not entirely clear about what you’re describing with the green bar. Are you saying that as you RAM preview, after a while it will stop adding to the RAM preview (shown by the green bar) and taking away the old frames that were RAM’ed? This happens when the computer can’t keep everything in the RAM preview, so you have to preview in sections. It happens.

8 GB of RAM is great, but you only have a single processor, it seems (or is that a quad core?). AE can only use a certain amount of RAM per core (i think it might be 3GB?) so the extra isn’t being used. If it is in face a multiprocessor, you can go into the AE preferences and enable multiprocessing. I haven’t done this so i’m not sure how it works, but my understanding is that it will speed up your previews.