Spectroscopic WCS reference

Frank Winkler wrote on Jun 07, 2011

For a simple onedspec format fits file, does the wavelength for pixel 1 refer to the left (short wavelength) edge of the pixel? or the center? I think the left edge (as for an image WCS), but would like to be sure.

And if I convert that fits file to wavelength, flux pairs, using wspectext, is the wavelength the left, or center, of the pixel with the specified flux?

Thanks,
Frank Winkler

---

Francisco Valdes wrote on Jun 07, 2011

A wavelength function is continuous. But if you mean what happens when a point is plotted or the output of wspectext the convention is always the center of the pixel. IRAF and FITS put the integer pixel coordinate (e.g. [1,1]) at the center of a pixel and the output of wspectext is for each integer pixel coordinate. It is also generally the case that a pixel flux, which is like an aperture flux from a continuous distribution, is taken to have that flux at the center of the pixel.

I hope this helps,
Frank Valdes

---

Last post on Jun 07, 2011