The procedures for focusing ACAM, and for implementing focus offsets, are discussed in the ACAM observing guide.

The offset in telescope (secondary mirror) focus required to compensate for the insertion of a filter in the ACAM light path depends on the filter bandpass, its thickness and, for a few filters, changes in the shape of the wavefront introduced by the filter. These dependencies are different for the filters in the near-pupil-plane wheels (this is where most observers will use them) and for filters in the focal-plane slit slide.

This page summarises information about the focus offsets likely to be required for filters mounted in the near-pupil-plane wheels.

For filters mounted in either of the wheels (1 and 2), the focus offset is primarily a function of wavelength, because the filters are in a near-collimated beam. Zemax modelling (by Tibor Agocs) predicts the focus offset (movement of the secondary mirror) to be:

dfocus (mm) = -0.0045 * filter thickness (mm) + C (wavelength)

where C = 0.016 (U), 0.028 (B), 0.002 (V), -0.027 (R), -0.055 (I) and -0.063 (Z).

Below are tabulated the predicted offsets as a function of waveband and filter thickness, relative to the offset for an R filter with thickness 5 mm (e.g. Sloan r, filter #216).

If, unusually, two filters are in the light path (one filter from each wheel), the net predicted focus offset is not the sum of the two relevant values from the table above, but:

dfocus (mm) = -0.0045 * combined filter thickness (mm) + C (wavelength)

where C is as given above.

Predicted and observed focus offsets (relative to that for the new Sloan r filter, name = 'SlnR', #702) are compared below. This table is being updated as new information becomes available.

The first eight columns give: ING filter name (as used by the observing system); filter number; thickness; predicted focus offset (using the formula above); observed offset and estimated rms error; date when observed offset measured, and seeing. The final column is a laboratory measurement (using a Shack-Hartmann setup) of the focus term (waves) introduced by aberrations in the filter itself.

Note the generally good agreement between predicted and observed offsets, except for filters with focus-term measurements in the last column exceeeding 0.5.

The Shack-Hartmann measurements of the strength of this focus term for each filter can be found on the TWE (transmitted wavefront errors) page. There's a significant correlation between measured Shack-Hartmann focus term and on-sky focus offset, which can be roughly characterised as:

on-sky focus offset (mm) ~ 0.14 * (Shack-Hartmann focus term)

(ignoring two apparent errors of sign in the data).

Note that this prediction can't substitute for on-sky focusing, because the rms errors in the focus term are 0.5 - 1.0 waves rms, depending on wavelength, implying an rms error ~ 0.07 - 0.14 mm in predicted focus offset (compared with 0.02 - 0.05 mm for on-sky focusing). However, the focus-term measurements are useful for (1) identifying those filters which are likely to give significant defocus, and (2) to provide a starting point for the on-sky focusing.

For most ING filters, the amplitude of the focus term is < 0.5, and for only 11 filters does it exceed 1.5.

Default focus offsets
Measured focus offsets are included in the filters database, and by default these are displayed by the ACAM status mimic when the control system is started up.

Filters mounted in the focal-plane slit slide
Observers occasionally mount filters in the slit slide near the focal plane. These lie in the converging beam, so thickness will be more important than bandpass in determining the focus offset. The above focus offsets will not be relevant here, and the current practice is to focus separately through each filter.