# K-correction vs resizing photometric bands

Is there a difference between a K-correction (see Wiki https://en.wikipedia.org/wiki/K_correction) and simply "resizing" the bands, e.g. the 2-10 keV x-ray band, with the redshift? I've read that K-correction requires the assumption of a spectral shape. Why is resizing strictly speaking not enough?

• Do you mean physically resizing the filter? In principle that would work, except you can't just resize a filter. You can build filters centered on various wavelengths, but redshift and hence K-correction is continuous, so you would in principle need infinitely many filters, and filter are expensive. Moreover, you would need to know the redshift before observing the object through the appropriate filter. K-correcting is easier.
– pela
Dec 17 '20 at 16:11
• @pela no I mean converting observed flux density to intrinsic luminosity, using the luminosity distance and a factor (1+z) because it's a density Dec 17 '20 at 16:33
• The point of the K correction is that you observe a redshifted object in some band, and you then want to know the flux in the rest-frame of that band. E.g. you want to know the rest-frame B magnitude of a galaxy at redshift z = 0.5, observing it in the B band. But the flux you observe is not the flux that was emitted in the B band, since that light has been redshifted into the observed V or R band. [cont. below]
– pela
Dec 17 '20 at 22:26
• [cont. from above] Rather, it is the flux that was emitted somewhere in the UV and then redshifted to your B band. In order to estimate the original B flux, you have to make assumptions of the spectral shape (i.e. the slope of the continuum, and possibly nebular lines which in some cases can contribute quite significantly to the flux, I think).
– pela
Dec 17 '20 at 22:27
• I think the reference for the K correction would be Oke & Sandage (1968), but David Hogg wrote a quite pedagogical (albeit non-refereed) paper on it (Hogg et al. 2002).
– pela
Dec 17 '20 at 22:33