[henrikf]

Author seems to be looking to generate a blue noise texture for image sampling. I'm not familiar with them but it seems to be blue noise which also has uniformly distributed values in time domain. Generating white noise in frequency domain and multiplying with a frequency shape mask can generate noise with any frequency distribution, but it does not fill the uniform distributed values in time-domain requirement.

If there are no other requirements than the frequency spectrum then generating the noise in frequency domain works fine.

[amelius]

If you view the IFFT as a random number generator, how do you change the seed of it?

I.e., white noise has a flat spectrum. So if you take a flat spectrum back to noise (using IFFT), what determines the seed of this process?

[nitrogen]

It's the seed of whatever source you use to randomize the phases.

[ssfrr]

Yeah I think that uniform time(space)—domain distribution constraint is the one that the author identifies as the main problem here.