FEATURE: Independent control of 'maxregshift' for X and Y axis during registration

[FishQian]

Feature you'd like to see:

Hey there.
I've been enjoying using suite2p and am thankful for this.

In the current version of suite2p I'm using, the ops[‘maxregshift’] is set so that "max(Ly,Lx) * ops[‘maxregshift’]" determines the maximum pixel shift that constrains the algorithm. In some scenarios, however, the required pixel shift does not apply to X and Y dimensions the same way. For example, if the X-Resonance-axis is set as 512 pixels, the Y-Galvo-axis as 64 pixels. Meanwhile, the absolute space is 300 by 60 microns for X and Y respectively. Assume 15 microns of motion drift, same for x and y axis. Then I would love to control the 'maxregshift' so that, e.g., in X-axis it's ~25 (15um / 300um*512) pixels, but in Y-axis it's 16 (15um / 60um *64) pixels. So the 'maxregshift' for X is 0.05, for Y is 0.25.

It will be great if the ops['maxregshift'] can be splitted into two variables as ['maxregshift_X'] and ['maxregshift_Y'] so that users can have flexible controls of the constraints on each axis for registration, given specific imaging conditions each might experience.

Attempted alternative approaches:

the feature is simply not there.

Additional Context

No response