The new noise variance infrastructure should make it fairly easy to accept an expression for the variance of each field. We might need to think a bit how this works for vector or tensor fields, but having this for scalar fields would already be a big plus!
The new noise variance infrastructure should make it fairly easy to accept an expression for the variance of each field. We might need to think a bit how this works for vector or tensor fields, but having this for scalar fields would already be a big plus!