Closed jecampagne closed 1 year ago
Hello, looking at def drawFFT_finish(self, image, kimage, wrap_size, add_to_image) I was wandering if in this line
def drawFFT_finish(self, image, kimage, wrap_size, add_to_image)
real_image = Image(breal, dtype=float)
the dtype would be more the "image dtype". I know that np.float32 is the default but... My 1cp.
No, that's intentional. We do the intermediate calculation in double precision regardless of the dtype to store the result.
Hello, looking at
def drawFFT_finish(self, image, kimage, wrap_size, add_to_image)
I was wandering if in this linethe dtype would be more the "image dtype". I know that np.float32 is the default but... My 1cp.