Cineversity

Project here:

https://shared-assets.adobe.com/link/6e4c9975-0c7b-45e0-72db-e6286d60ee44

I’m animating a camera “digitizing” part of an image, and then “ingesting” it into a database. I cloned a photo template and built the feeder animation sequence using the Plain Effector. Then I created an image sequence in PhotoShop and imported the layers with PhotoSmart.

The idea is that as each photo is fed into the scanner, I run the animated sequence in which the object is traced, extracted, and then “sucked” up into the camera lens. All very old-school manual.

This happens 10 times (for 10 photos).

When I first set this up, I expected to be able to use the Multi-Shader to feed in the initial photo through the cloner. The tricky part would be adding the mask area to the photo as it exits the scanner (as it is visible in the background). I would then build the tracing and extraction sequence over the top of the image as a separate manual construction.

Can you point me in the right direction as far as controlling the multi-shader to switch on and off the mask area in sync with the motions? I generally handle the multi-shader in default mode only because every time I try to switch something minor, it breaks.

I also noticed that the alpha in the overlays is partially transparent and I can’t find the checkbox to turn it back on to 100%. I suspect the problem occurred when I swapped the underlying PSD from a massive 100MB source file to a more manageable one for sending over to you. The original PSD was clean and bright. The substitute file is semi-transparent. I’ve had that happen before and was mystified, so if you can shed light on that also, it’d be a big help. Usually I just start over from scratch and everything snaps back to normal. But this sequence promises to be pretty complicated by the end, and that would be a mess.

Thanks.

Hi fordmar,

Would you mind uploading the file again? Please check its size first. I get only 0KB with two different browsers.

Cheers

Is this one better?

https://shared-assets.adobe.com/link/578861d6-009e-4de3-7310-4aedd8955450

My browser says this one is 10MB

Thanks for the extra effort, fordmar.

I tried the first link again. No data, but the second link worked just fine.

You have four things in the procedure. The image arrives, the edge is detected and marked, and the figure is taken out. Figure flies off.

Steps one to three are doable with three MG setups that are the same, except they are each one until higher than the previous, to avoid the “same position problem.”

Clone A1 arrives, and after a moment, it is replaced by clone A2, then by A3, B1 moves in, B2, B3, then C1, etc. In the project, they are just called Cloner.0, Cloner.1, and Cloner2.

The exchange is done via Plain Effector> Visibility/Field animation.

Example
https://www.dropbox.com/s/pra4vb6yx7gnky8/CV3_S26_drs_22_MGti_01.zip?dl=0

I would work with single images, never with layer transparency. (A transparent pixel is never a good idea, better to have a full pixel and an alpha.

I have checked if anything has changed with the Multi Shader. All fine, but of course, it needs to be set up exactly as you need it. If the amount of layers is not equal to the number of clones, things become different than expected. I love to work with ID, though colors (brightness or R, G, or B are sometimes better).

To pick up the “cut out” seems to be not a problem. Clones visible and movement/scale.

If you need to use Photoshop, things get more tricky with the Multi-shader, as you can select the layers, and sometimes that gets screwed up if you change something in the files heavily.

All the best