🚀 go-pugleaf

------- Start of forwarded message -------
From: cur@rlstar.bnsc.rl.ac.uk (Malcolm J. Currie)
To: fitsbits-request@fits.CV.NRAO.EDU
Subject: Re: World Coordinate System proposal
Date: Sat, 7 Nov 1992 19:39:15 GMT

Steve Allen writes:
>From my experience I infer that the astronomical community has
>pretty much established that the first pixel in an array is the pixel with
>the integer tag 1 which has floating point boundaries 0.5 and 1.5
>(middle of pixel is integral).

In Starlink we use pixel indices that, by default, start at 1.  However,
our definition of pixel co-ordinates has the first pixel (index 1)
between floating-point boundaries of 0.0 and 1.0, since we regard pixels
as binned data.  Isn't it more logical to start at the origin rather
than 0.5? 

Our NDF data structure permits axis widths to be defined in addition to
axis centres.  So for sampled data the widths could be made much smaller
than inter-pixel spacing. 

- -----------------------------------------------------------------------
Malcolm J. Currie   |     Span: RLVAD::CUR
Starlink Project    |     Janet: CUR@UK.AC.RUTHERFORD.STARLINK
                    |     +44-235-821900 x6735
- -----------------------------------------------------------------------
------- End of forwarded message -------