Observers may wonder if they need to check spectrograph focus before each night. HIRES has a very shallow and stable focus curve so, unlike many other instruments, this is not necessary. Nevertheless, if such a check is desired, observers should ask their support astronomer for assistance. Scripts exist to automate taking the data and analysing it.

The best focus position of the HIRES camera is measured regularly (using a slit which projects to 2.2 pixels) and if necessary, the default value is modified. This plot shows the best focus camera position since the the new detector was installed in mid-2004. The different colored curves show the best focus position for each of the three CCD's in the detector mosaic. At first glance this plot may look alarming since there is a definite focus offset between the red CCD and the other two. In fact, the offset is of the same order as the width of the minimum of the focus curve. Here is a typical focus curve. Note that the FWHM plotted does not represent the actual FWHM of unsaturated, unblended Th/Ar lines. The curve is the auto-correlation function versus camera position, so it includes blends and saturated lines. The true FWHM is roughly a half a pixel less.

Although the offset between the red CCD and the other (blue and green) CCD's is not enough to seriously degrade the focus for the chosen default value, observers may only really care about the focus on one of the three CCD's. In such a case they may wish to choose a non-standard camera position to achieve the very best focus for that one CCD. In such a case, observers should consult their support astronomer. In the future, we may attempt to introduce some tilt to the HIRES detector to alleviate the offset, but since the blue and green curves agree so well, it is believed a simple tilt will not completely remove the offset. The focus offset is probably at least partially due to some piston of the red CCD with respect to the other two.

.