Recently, I've been working a kids' show in which the actors were videoconferencing quite a bit. Along the way, I developed this little trick that helped tremendously in making the process of putting faces in screens less cumbersome.

This tip has two parts.

Part 1 - Making the composition.

Putting an image inside a screen is a 2-track effect: one for the action, and the other for the TV frame.

The first part of the tip is to drag the 3D PIP effect not on the action, but on the TV frame. I found a lot of editors usually do the opposite, since it's the default for a PIP effect to shrink and extrude the image you put it on.

But since you want to drop the effect on the TV screen this time, you will first need to cut this shot above of the action, on v2. Next, you drag your PIP, highlight the Swap Sources button (located in the Foreground menu), and voilà, your action lies on top of the TV screen shot, ready for you to frame carefully.

At this point, he result looks similar to what you would achieve with a regular PIP placed on the action. But there are two advantages to our new way. First of all, the scene is easier to cut, as you don't need to put up the screens until you are satisfied with the dialogue, and secondly, it's quicker to make changes, especially with the second part of the tip.

Part 2 - Saving the effects to a bin (with source).

This second part is what makes this tip shine, especially if like me you have many choices of angles for any one screen. Once you have tweaked your TV screen frames to perfection, you can save your hard work to a new bin as a 3D PIP (With Src), by option-dragging the effect to a bin. The beauty of this is that the ensuing clip is treated in all respect like a subclip. You can splice it in over your dialogue, review, change your mind and replace it with another angle, all very quickly. It helps if you put your bin in Script mode, as you will see a preview of your compositions.

Those shots being in a separate bin also helps later if you want to steal one for use in another scene. Since the composition is already done, it's only a matter of splicing it in.

This was very handy for me. Hope it works for you too. I have the feeling we will be doing more and more of those videoconference scenes in the future, don't you?

This sounds interesting. I've usually just done corner pin.  Is there any chance you could do a brief video tutorial on this? I'm having trouble visualizing/following some parts of it jsut from reading about it.

Maybe not a video--I don't have the proper setup--but I'll look into putting up some screenshots.

Thanks for your interest.