It works as it is supposed to. There is no way to have a camera filmback/ fustrum size different from the comp size. Would be kinda illogical, anyway. Whatever you have in mind will probably require some pre-composing.
My problem is that I'm working with a large artboard effectively. I have seen many animations created in a similar style where a camera moves across a large area, and I'm sure they can't be working the way that I am as it's incredibly hard work for the computer as well as being a little clunky.
The link below is to a video with what seems to have a basic set up like mine, large artboard and lots of layers, with a camera moving around.
Mylenium is correct.
Your comp needs to be 1920x1080. This is what sets the frame size of your camera. You then animate the camera. No nesting should be required.
The size of your artboard is huge. What kind of content? (Photos, text, vector shapes, video?) Does it need to be that big? This will be dependent on how close the camera comes to it if it's a photo. If you can scale it down at all, before importing it into AE, that would improve performance.
The example you link to uses only text and vector graphics. You see how the gradient background doesn't move? That's a 2D, non-moving layer, so it's only a 1920 x 1080 graphic. Only the text and shapes are 3D, being affected by the camera moves.
Here's a video I created, possibly like what you're going for. The largest image is 1920 x 1080. However, these are vector shapes, so I can zoom in close without losing any detail. During editing, I disabled motion blur, depth of field, and worked at half resolution. Rendering out at full resolution, with motion blur & depth of field, took nearly 10 hours.
1 person found this helpful
Your workflow is flawed. There is no problem moving the camera around a large 3D scene. There are a bunch of tutorials available for dynamic text animations that show you exactly how to do this. Pre-composing a huge comp is just going to slow things way down. AE does not consider elements outside of the composition boarder when rendering. You also set in and out points for your layers to reduce the memory requirements.
Sometimes I'll layout a huge artboard in Illustrator, arrange most of my elements on separate layers, then open the huge artboard as a comp. Before I start animating I'll resize the comp using composition settings. You can pick any corner, side, or center of the composition for the new center. This makes it easy to have all of your elements roughly positioned. Set up a 3D camera and start moving things around.
What am I wrong about? Scaling down before bringing into AE? I'm not pre-composing/nesting anything.
I know that content outside the border doesn't render. But doesn't it affect performance when editing the comp, like when you're moving/panning the camera? I had a large collage of photos created in Photoshop, and scaling the collage down before bringing it into AE seemed to make things more responsive.
So even if I scale my comp down to a normal size, it will still let me move the camera to view and render the rest?
I'm sure I tried this as my initial option but I recall it just rendering out black/ cutting the frame at comp size.
Is this how most people work though, with a large comp and small camera? or a small comp and an even more zoomed camera?
Thanks for all the info guys, appreciate it
Glad to hear, Rick. Just wanted to be sure!
_lornaw_: You need to make your image a 3D layer. This will make it so you can move the camera around the entire image.
You're mixing up some terminology. In AE, your comp size = your camera size. At this point, I would create a brand new comp.
- Create a new comp at the resolution & frame rate you want to output to. The resolution you pick will be the comp size, which will also be the camera size (1080p, according to your 1st post).
- Import your image into AE. Drag it into the comp as a new layer. Your comp is still 1080p. Because of this, and since your image is so large, you're only going to see a small portion in the 1080p frame (which is okay for now!).
- Add a camera. Your comp & camera size are still 1080p.
- Now tick the 3D Layer switch (looks like a 3D cube) for your image layer.
- The Position attributes for your camera and image layers have a 3rd value, which is the Z axis. You can either pull the camera back to see more of the image, or move the image forward away from the camera. Also, still at 1080p.
What kind of image are you using? Is it raster/photographic or vector-based? And what file type? JPEG, PSD, Illustrator? The dimensions are huge, so you might be able to do some optimizations to it without affecting the quality of your video.
I think I must have made a mistake then when initially setting up as it didn't work for me then! But trying it now so hopefully it will all work out.
Thanks for the demo, I do understand what I need to do but I must have made a mistake before. I'm working with vector based illustrator files, currently selecting all parts of the action and splitting them into lots of sections of 190 x 1080 windows, so hopefully this will speed everything up too rather than all the empty space im working with currently.
My question now is more general, about what everyone understands most peoples workflows to be, whether people use large graphics and move the camera around them, or if people use very small and scaled graphics on a normal composition size, and have a camera zoom in close on them?
I'd say it depends on the type of content you're using.
For vector graphics, you can use smaller graphics. By enabling continuous rasterization for the layer, the graphic will remain sharp even when scaling up or moving the camera really close.
For photo or video content, if you scale it up past it's original size or you move the camera in closer than 1:1 pixel mapping, it's going to look fuzzy.
My real-world example uses both vector graphics and video content: http://youtu.be/CzNpgC-5U08
All of the graphics were vectors somewhere around 1080p resolution. Every layer is 3D, at a different Z-depth. I pushed the background really far into the distance, and when I did that I got black borders around it. I then scaled it up to fill the frame.
The videos of instrument playing were shot at 1080p. I masked them, scaled them down, and made sure not to move the camera closer than 1:1 pixel mapping.
I hope this makes sense.
It's interesting to know what techniques and work flows others have, as I've created only two of these and have used large graphics and art boards, next time I may try a regular art board and tiny graphics!!