madbutter

Is there a way to input a matrix into jit.rota, using boundmode 4 and output it to a bigger matrix (with all the folding jit.rota does in boundmode 4 ) so that all (well, more) of the pixel data of the original matrix is retained, but the bigger matrix is "filled out" with copies of the data?

For instance I know that I can take an 800x600 image and scale it .5 in boundmode 4 and I can get a matrix out that is also 800x600 but has the image half resolution folded out to fill out the output matrix.

But I want to take that 800x600 matrix and do the same effect (without scaling the original pixel data so much) and have it fill out a 3200x2400 matrix.

In other words, I want to copy a smaller matrix into a bigger matrix, using the cool look that jit.rota gives you when you use fractional zoom values to make the image tile, messing with theta and anchor point values, but preserving resolution.

Is there a way to input a matrix into jit.rota, using boundmode 4 and output it to a bigger matrix (with all the folding jit.rota does in boundmode 4 ) so that all (well, more) of the pixel data of the original matrix is retained, but the bigger matrix is "filled out" with copies of the data?
For instance I know that I can take an 800x600 image and scale it .5 in boundmode 4 and I can get a matrix out that is also 800x600 but has the image half resolution folded out to fill out the output matrix.
But I want to take that 800x600 matrix and do the same effect (without scaling the original pixel data so much) and have it fill out a 3200x2400 matrix.
In other words, I want to copy a smaller matrix into a bigger matrix, using the cool look that jit.rota gives you when you use fractional zoom values to make the image tile, messing with theta and anchor point values, but preserving resolution.

Is there a way to input a matrix into jit.rota, using boundmode 4 and output it to a bigger matrix (with all the folding jit.rota does in bo

jit.rota into expanding matrix