Introduction to 3D Using Flash CS4 - Page 1
       by kirupa  |  7 April 2009

In the previous page, you learned about fun things like the 3D Translation Tool and the Vanishing Point. That largely sums up the features inside the Flash CS4 UI relating to 3d this article will discuss. In this page, I am going to just discuss some issues that you may run into when using the 3D functionality.

Things to be Aware Of
Despite the coolness of all this, there are some limitations to how much you can do in Flash without resorting to writing a lot of extra code or looking at using 3rd-party 3d libraries.

My File Size Increased
When publishing your movie that has 3d applied to it, you may find that the file size of your overall project increases. The reason for this increase is that the 3d effects you see require ActionScript code that is generated for you behind the scenes. The addition of this AS3 library that contains the 3d functionality accounts for the additional file size increase.  All around genius, Trevor McCauley (aka senocular) wrote about this earlier, and you can learn more about that in his post: Flash CS4- The Overhead of 3

All 3D Content Gets Converted into a Bitmap
One of the drawbacks of the 3d functionality in Flash is that whatever element you are attempting to manipulate in 3d space is automatically converted from a vector into a bitmap. Using the 3D Translation or Rotation Tools with even a minor nudge in a specific direction is enough to rasterize your vector into a bitmap.

A perfect vector shape such as the four squares that I've been using throughout this tutorial looks pretty good when viewed at its default size. Once you start zooming in, notice that the classic banding/blurring associated with bitmaps are visible:

[ oh the blurriness! ]

Besides the visual artifacts, as you can imagine from manipulating actual images instead of vectors, performance suffers a bit as well.

Intersections aren't Possible
When shapes move in three dimensions, interesting collisions and intersections are possible. Unfortunately, this is something that isn't possible with the functionality available today. For example, imagine you have two parallel rectangles placed near other as shown below:

If I were to rotate the blue rectangle, in the real world, it would seem like a portion of the blue rectangle's edge will go through the yellow rectangle that is directly adjacent to it. Instead, what happens is that the blue rectangle rotates without taking the yellow rectangle into account:

This looks a little weird. It is almost as if each element has its own 3d universe based on its z-order instead of coexisting in one large universe where interactions and collisions are possible.

Just a final word before we wrap up. If you have a question and/or want to be part of a friendly, collaborative community of over 220k other developers like yourself, post on the forums for a quick response!

1 | 2 | 3 | 4