Parameterize non-HST filter curves

[pllim]

Instead of using FITS tables, store non-HST filter curves as coefficients that describe the math of reproducing the curves. Perhaps astropy.modeling can be used. Example curves are Johnson UBVRI, etc.

However, this does not replace the old FITS files; Just in addition to.

Pros: Save storage space.

Cons: Loss in accuracy.

[pllim]

@bmorris3 plans to investigate parameterization of Bessel filter and will report back.

Update: See https://github.com/bmorris3/tynt

[bmorris3]

Starting an ongoing effort to investigate this in a repo called tynt

[pllim]

#257 is merged and can be used for this purpose if we ever decide to come back to this.