I have tried, and failed to use a perspective image as a texture on a face in SU but corrected for the perspective.
I’m trying to draw a fairly accurate representation of a model railway point motor, which looks like this (manufacturer’s photo):
I know the overall dimensions (x=68mm, y=40mm, z=45mm), and that the boxy part of the body (the bit with rounded corners) is 52 x 28 x 38mm.
Image:
Example image imported on to a face of approximately the correct size:
And here’s the model.
Image position test.skp (117.9 KB)
I can’t see how to correct the trapezoid distortion to get the image to go from trapezoidal to ‘square on’.
The Blue or Yellow Shear buttons in the Texture Positioning operation don’t seem to do what I want.
I know some image editing programs can correct for this kind of perspective distortion, and I thought SU would do it for me. If it can, I can’t see how.
I tried photo matching a different image of the same object and got reasonable results, but after refining the shape from a box to match the object, I wanted to apply separate parts of the image to separate parts of the model.
I’ve got the model shaped and coloured with plain colours (attached below), but can’t find a way of putting textures back onto the separate faces.
Cobalt point motor and angle convertor .skp (948.7 KB)
Any ideas please? I have read the Help article, which doesn’t help me much beyond saying what the buttons do in general terms.
PS. I’ve just got better but not perfect results using Rotate first to get the bottom straight edge of the image aligned to the face, then using Blue and Yellow shear and scale to get the top edge more or less aligned - not so bad, but the last bit of adjustment isn’t helped by the movement ‘jumping’ as it tries to snap to a corner.
I can get the Blue button to scale in the Y direction. How do you scale in the X direction? If I drag the Blue pin in the X direction, it just shears the image, and the Yellow button only shears whichever way you move it.
In Free Pin mode, perhaps that will do it? I’ve just reread the Help, and it suggest that you can set the four pins at ‘corners’ in the image and then move them to the corners (or other relevant points) on the face. Maybe that’s the way to go?
Yes it seems it is - so perhaps I’ve answered my own question (except for how do you scale in X)?
And by extended accidental experiment I find that holding down the option key (on Mac - Ctrl on Windows probably) I can stop the ‘jump’ near a corner of the face.
So maybe tomorrow I can do all the relevant faces (some will look better in a plain colour).