The naming convention for these FITS files will depend on the save-as format. For
monochrome formats the extension is changed to ‘.fit’. For colour formats, 3 separate files
are output with the ‘ending’ is changed to ‘_R.fit’, ‘_G.fit’ and ‘_B.fit’ for each primary
colour channel.
This option to save as FITS with double precision binary float values is provided so you
can take advantage of the increase in quantisation resolution afforded by multi-frame
averaging. Currently FITS files can only be saved, they cannot be read by PARD Capture
as input for dark field and flat field corrections (for those input files you can use the raw
doubles option described above). This FITS output option is provided for those who want
to capture images with PARD Capture and then do further processing in other software
that can read FITS files. The FITS files produced by PARD Capture are compliant with the
FITS standard.
Optional pre-processes
The remaining custom settings relate to optional pre-processes that may be applied to
frames as they are being captured and before they are output.
All these pre-processes are performed in double-precision floating point internal frame
buffers so are not restricted by the 8 bpp or 24 bpp ranges of the original captured
images.
All these pre-processes are only applied within the support of the current corrections
mask (see below) so that only pixels under the non-zero pixels in that mask will be used in
the calculations and corrections.
Figure 6.6 shows the algorithm used to apply the optional pre-processes.
The ‘
Scale mean of each frame to first?
’ option, when selected, is a pre-process
applied to multi-frame averaging sequences. It therefore has no effect if multi-frame
averaging is not being done. If this option is selected then, when a multi-frame average is
first started, the first image frame that is captured has its mean pixel value evaluated and
that mean value is stored. All subsequent frames in the multi-frame average also have
their mean pixel value calculated and compared to the stored value of the first frame. Prior
to being accumulated into the averaging buffer, the current frame is scaled by a constant
multiplicative factor (applied to each pixel in the frame) that transforms its original mean to
be equal to the mean of the first frame. The multiplicative scale factor used is calculated
as:
scale_factor = Mean_of_first_frame / Mean_of_current_frame
Thus this option scales the mean ‘brightness’ of each image to be equal to the mean
brightness of the the first image in the averaging sequence and so compensates for any
fluctuations in illumination that my occur over the multi-frame averaging period. This
ensures that no one frame dominates the resulting average result.
The feature may be particularly useful for the purpose of capturing a multi-frame average
to be used as a ‘master flat’ image for flat field correction (see below). Note that during
this process, all frame averages are calculated only from pixels that are in the support of
OptArc AF51 Camera Page 80 of 99 User Guide v1.02
