Masking 3D planar text

Hi all - was hoping someone could help with this relatively newbie q:

I am doing a property tour and wish to track text to walls. I can do this with the PlanarTracker but am having trouble masking objects out. For example in the image below I want to animate the text coming up from behind the sofa and also to mask the dogs head out whenever it overlaps. If I use the polygon mask tool it doesn't recognise the 3D space - it masks but its only representative of the 2d version of the text so there is an offset. An example of my node tree is included. Any help greatly appreciated!

Screenshot 2022-06-22 at 11.05.13.jpg (153.65 KiB) Viewed 1001 times

Screenshot 2022-06-22 at 11.05.16.png (19.64 KiB) Viewed 1001 times

Hey,

I admit I was lazy here to explain, so here your dog is a car, and you can forget or not the 3d version of the text

Hey,

I admit I was lazy here to explain, so here your dog is a car, and you can forget or not the 3d version of the text

Thanks for the reply. Unless I'm mistaken using the tracker won't create the planes that i need for a camera move such as a pan of a room. If I repeat that tutorial then the text is in 3d space but it's not translated to the wall of the room as no custom plane exists. So I have to manually rotate the x asis and but it wont stick to the wall. Is there not a way to use a mask with the planar tracker?

Hey

What you want to use is the match move fonction of the tracker.
BTW, the video I posted was to show you how to treat your dog's face trick.

You know, there are a lot of ways to make it, depending on how many times you're going to use it, the footage profile, etc etc.
What I would do is basically a planar tracker to track the cam moves > create a planar transform out of it > put the planar transform between the text node and the merge > then add a transform node if you wish to add another translation to the text.

That's the old way of doing it. Create your text layer. Make a compound clip of your video layer, open the clip in a new timeline, go to color and use the object mask. It will be 50x faster and 100x cleaner.

That's the old way of doing it. Create your text layer. Make a compound clip of your video layer, open the clip in a new timeline, go to color and use the object mask. It will be 50x faster and 100x cleaner.

Only if they have the Studio version and they might want to do it in Fusion.

Yeah, fingers crossed they add the mask to Fusion in the next release. They've been slowly sneaking the mesh tracker and the depth map in so maybe next release...?

That would be great.

To the OP, you don't need the Merge node after the Text+ node, they are used to merge two things. you only have the Text+ going into it.

Of course one wants to do it in Fu... How could it be elsewhere
And there's a whole bunch of stuff which is awaited in Fusion Studio, like magic mask, surface tracker, depth map, etc etc + fixes of what's already inside...
A compound clip, go to color, come on...