Need a script to Resize Multiplane Camera project

[D.T. Nethery]

Indeed, the circular pointer is a way to input angle, not rotation: you use it to precisely input a 0 to 359 degree ANGLE, never more - never less, . Good point!

The label, "ROTATION", is definitely wrong.

sven

The labeling of the wheel/pointer as "Rotation" was one of the things that confused me about the actual function of the Rotation parameter in the Multiplane FX. I would move the "Rotation" pointer and see the Angle change in response , but nothing changed in the Rotation parameter. The Tool Tip pop-up which appears when the user hovers their cursor over the Rotation parameter should clearly communicate to the user that Rotation refers to the number of times the layer will make a complete 360°rotation (between two keyframes) -- 1 means one complete rotation , 2 means two complete rotations, 3 means three complete rotations , etc. The starting keyframe set to 0 and the next keyframe set to 1, 2, 3, 4 , etc.

However, the Tool Tip pop-up as it is now adds further confusion. Notice that if you hover your cursor over Angle the Tool Tip pop-up says it is for setting the "Rotation Angle" , but then if you move your cursor down to hover over Rotation the Tool Tip pop-up also says "Rotation Angle" . It should say "Number of Rotations" or something like that.

.

[Svengali]

It is worth noting that values for angle and rotation are necessarily combined when representing any angular value greater than 359 or less than -359. For example an angle of 520 degrees would be Angle 180 and Rotation 1. And you just have to kinda "know" this because you'll never see 520 degrees displayed anywhere.
Multiple complete rotations of the circular pointer (CW or CCW) will automatically increment or decrement the Rotation counter.

sven

[D.T. Nethery]

Well, I hijacked my own topic by going off with questions about the Rotation and Blur parameters in the Multiplane Camera FX .

Let me try to get back to the original topic to add a few more thoughts that occur to me as I try to understand this Multiplane tool better. (to me, it is really not intuitive or easy to master ... compared to compositing multiplane camera moves in After Effects or Harmony or Moho). I welcome input from more experienced users like Svengali who kindly correct my misunderstandings of how advanced functions in TVPaint are meant to work. I realize that maybe only a handful of TVPaint users are interested in this topic ... I think most people -- including me -- tend to give up on trying to work with the Multiplane Camera FX in TVPaint except for very simple multiplane camera moves, and will use After Effects to accomplish more complex multiplane camera moves.

I'm bringing up these points because I've recently turned my attention back to trying to make use of TVPaint's multiplane camera again , after having spent some time using After Effects for compositing multiplane camera shots instead. The truth is, After Effects works great for the purpose of compositing multiplane camera shots (especially using scripts like PixelBump Multiplane for AE or pt_Multiplane ) , but I suppose I am coming back around to the thought that if there will be a tool in TVPaint called Multiplane Camera , then that tool should be improved as much as possible. It may be that any significant improvements will come in the next generation of TVPaint , not in the current engine. However, things like the method that Cardin Collins has proposed (which I posted HERE ) can help bridge the gap.

One point I have come to understand better is that the use of the IMAGE LIBRARY for the SOURCE images to be used in composing a Multiplane camera shot is very important. I will no longer use layers on the TVPaint timeline as sources for the Multiplane ; I use layers in the Image Library as sources (or for animated layers , save those anim layers as a separate TVPaint project and add those as a source , as mentioned in Point 3 of the Cardin Collins workaround). What I did not completely understand before was that it appears as if any images added to the Library will always retain their original pixel resolution, even if the TVPaint project size is scaled down to a lower resolution . For example , if I create multiplane camera Layers in Photoshop or in TVPaint at 4x overscale (1920 x 1080 overscaled 400% at 7680 x 4320 ) and add those layers to the Library in a TVPaint project with resolution of 1920 x 1080 , the layers in the library retain their original size of 7680 x 4320 . Those layers in the Library retain their original size even if I downscale the TVPaint project by Modifying the Project from 1920 x 1080 to 640 x 360 to allow for easier composing of the multiplane camera without the lag experienced at larger resolutions (using the Cardin Collins method ) , so the multiplane movement is worked out in the downscaled project , then the multiplane camera coordinates are Added to the FX Bin , and switching back to the higher resolution project the coordinates saved in the FX Bin are used to render the move in the higher res project . The rendering may take a long time depending on the complexity of the movement and the number of planes , but it was faster and easier to compose the shot in the lower-res. version of the project , then render it out in the high-res. version of the project after having seen the lower-res. preview.

Perhaps I'm being redundant in reiterating what I had previously written regarding the Cardin Collins workaround , but think it's important to emphasize the point about using the high-res. overscaled images in the Library as the Sources for the planes in the Multiplane Camera FX because of what I had mentioned previously regarding project resolution needing to be doubled, tripled, or quadrupled to allow for zooming-in close without the images becoming blurry or jagged. I wrote in one of my previous posts:

"... the final output resolution is intended to be 1920 x 1080 , so if the camera is pushing-in closer (or pulling-out from a close to wide shot) in the Z plane , you must overscale the original artwork to prevent the pixels from getting blurry , so if you have a multiplane camera move where the camera is going to zoom-in close , say by a factor of 3x , then your project resolution for the original artwork must be 5760 x 3240. If it's a zoom-in factor of 4x , then the project resolution must be 7680 x 4320 . At those resolutions you will definitely experience lag , to the point where it is almost impossible to work with the Multiplane FX tool, especially if you have many layers and if you are adjusting options such as Depth of Field and Focusing Distance to get rack focus effects. Even 2x overscale at 3840 x 2160 will be laggy. For me, the biggest issue is not the rendering time with a high-res scene (I can set a scene to render and then go do something else , take a walk , make lunch ...) but the lack of responsiveness in the TVPaint FX Stack interface when setting up the X, Y, Z positions of the planes and the laggy (or frozen) preview playback before the scene is rendered."

What I should have written more precisely is where I wrote about the project resolution needing to be set to 2x , or 3x, or 4x of the intended output size of 1920 x 1080 ... in fact, I believe that what I should have written is that the TVPaint project size may be set to standard 1920 x 1080 , but the source layers stored in the Image Library (or source Anim layers stored as separate .tvpp project files) that will be used in compositing the multiplane shot should be appropriately scaled up 2x , 3x, 4x , etc. larger than 1920 x 1080 if the camera will be zooming-in close , or if the camera is starting close and zooming-out wide.

If the camera is planned to zoom-in through multiplane layers and end up in a tight field on part of the scene , it's important to know the end position of the camera , the point at which the camera is most closely zoomed-in. The most extremely enlarged area of the zoom-in must be equal to an area that is 1920 x 1080 pixels (or whatever is the intended output size). So to allow for that, you must calculate ahead of time if your original multiplane layers must be painted at a pixel resolution that is 2x , 3x , 4x, or 5x larger than 1920 x 1080 to allow for the zoom-in without causing the pixels to become blurry.

I hope this all makes sense and isn't just a lot of rambling that simply goes to show my ignorance of how the multiplane and the image library were intended to work (?) . I remember watching the video tutorial that demonstrates using images in the Library as a source , but when I watched it before it did not sink in that the source images in the Library remain at their original resolution, independent of the TVPaint project resolution.

[D.T. Nethery]

Something else that should have been mentioned about the Cardin Collins workaround is that it doesn't work if you are making any changes to the Depth of Field or Focusing Distance parameters. Or rather , if you resize the project to a smaller size , the amount of Depth of Field/Focus Distance will be exaggerated . So it is better not to apply any changes to Depth of Field or Focusing Distance parameters when you are in the smaller project . When you get back into the larger project you can add any adjustments to Depth of Field or Focusing Distance before doing the final render OR another option is to render each layer one at a time to a separate anim layer in the TVPaint timeline (you do this by lowering the opacity to 0% on all layers except the layer you are rendering) , then after having rendered each layer to it's own separate anim layer you can add simulated depth-of-field or rack focus effects to the layers using Gaussian Blur FX .