mveyette

Taking this a step further, you could also account for how resolution changes across an order. I am not sure the best way to implement this, but it would be...

Nice! In theory, you could deconvolve the instrument profile as a function of wavelength (or pixel for echelles) from arc lamp lines for instruments that don't provide resolution curves. Then...

You're bringing up some great points today, Gully. I've thought about this too and have implemented Chebyshev polynomials before to correct for bad wavelength calibration. I didn't care about RVs...