-
workflow with a super huge canvas
I see a lot of comps that have big video walls, with each video monitor in the wall being HD quality and the camera zooming between the different monitors. If you have a video wall with only 5 HD screens across by 4 screens high you’ve already got a comp size of at least 9,600 x 4,320 or 41 megapixels per frame. Well that seems like a huge waste if your final output will only be 1920 x 1080 (or anything normal size). So what is the normal workflow for creating such a large canvas as would be needed for a video wall of HD footage?
*Do you even make a comp that humongous, or is there a different workflow such as maybe sending each of the HD monitors backwards in z-space in order to keep your working comp a manageable size? If you send things back in z-space does it help AE render faster or would it still need to “think about” all 41 megapixels-per-frame that the video wall would have taken up if it were closer?
* To send a super huge video wall back in z-space, would you send each HD monitor layer back in z-space in your normal size working comp (where I imagine it would be difficult to manipulate them all together as if they were one wall) or do you go ahead and make the super huge video wall comp first and then place that whole big thing in another comp and size it down?
* If you do really work with a super huge comp to create the video wall, is it possible to make the camera view only a normal 1920×1080 (or other size) instead of the entire comp?
I know these things could be done any one of many different ways, but I was curious which workflow would be the smart way to go.
-
4 Replies
-
Hi Mike,
If you are talking about creating a physical video wall with multiple physical screens, then you will need to create a large comp.
If you are talking about creating a video wall effect on a single screen, then there is no need for the large comp. Just create a comp sized for whatever your delivery requires (i.e., 1920 x 1080). Arrange the “video displays” in 3D space and animate accordingly.
There are a couple ways to do that. You could create a camera and fly around them. You could alternately arrange all the displays, create a null object to control them all, set the null as the parent of each of the displays, and then animate the null in 3D space. All the displays would stay “linked” and animate as if they were one element. Finally, you could pre-comp them in a large comp as you have suggested, and animate the pre-comp in your working comp (with or without collapse transformations on, depending on your needs).
Whatever method you choose, I’d keep the working comp at delivery size.
Walter Soyka, Principal
Keen Live, Inc.
Presentation, Motion Graphics & Widescreen Design
RenderBreak: A Blog on Innovation in Production -
Thanks. I did just realize that I maybe I should have given you some examples of what kinds of situations I was talking about. Better late than never…
https://www.revostock.com/After-Effects-Project/61603/Video-Wall.htm
https://www.revostock.com/After-Effects-Project/25640/video-wall-01.htm
https://www.revostock.com/After-Effects-Project/21409/GIANT-BOX-05.htmBut I do think I understand what you’re saying I should do. To make “huge” scenes like these, work in a normal comp size but use the camera view to see what would otherwise be way way off screen if it were just 2D (right?). Thanks again.
-
[Mike Sussman] “To make “huge” scenes like these, work in a normal comp size but use the camera view to see what would otherwise be way way off screen if it were just 2D”
Exactly.
You can use Red Giant Software’s PlaneSpace to arrange and manipulate all those layers in your video wall — this will make your project a lot easier.
Walter Soyka, Principal
Keen Live, Inc.
Presentation, Motion Graphics & Widescreen Design
RenderBreak: A Blog on Innovation in Production -
I’ve had to do similar projects and found the hardest thing to get around is the “Unable to create 9,600 x 5,400 image buffer” error. The way I’ve got around it in the past is to split anything really big down into smaller parts the size of each monitor, (eg 1920 x 1080), in Photoshop and then reassemble them within a 3D comp at the same resolution (1080). Then, create an array of cameras, one for each screen, and lock them together in a grid. If you align everything correctly, this will give you one master comp where each camera correlates to a specific screen in the Video Wall.
This way, you’ve effectively got a single 9,600 x 5,400 comp, but with no elements (including the final comp itself) being bigger than 2K. If you’re good with maths & expressions, it can all be set up fairly neatly. You can also get back a lot of render time by disabling the parts of the image that a particular camera doesn’t see, which in this example would account for 24 of 25 images (or 96%).
Hope that’s not to confusing.
Tristan Nieto
Visual Effects & Motion Graphics
tristannieto.blogspot.com
Reply to this Discussion! Login or Sign Up