*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi Andrew

As previously discussed there are a multitude of issues to consider
when deconvolving the images from your customized microscope such as
choosing an algorithm appropriate for the noise model.  While
investigating that you would also want to understand any noise related
user-adjustable parameters the algorithms may have.

After you've worked through the noise handling of the algorithm you
would want to look very carefully at the PSF generation.  I was with
AutoQuant/MediaCy for many years and worked on (among other things)
theoretical psf and aberration models, so I have some thoughts on
these issues.

As others have mentioned for a novel modality, if no software is
available for calculating a theoretical PSF, you would have to use a
measured PSF.   From being involved in deconvolution in the past and
attending the same workshops as SVI I am moderately familiar with
Huygens.  One possibility is that the differences you are seeing
between the Huygens implementation and the Parallel Iterative open
source implementation could be partially due to PSF processing.  You
should ask SVI more about this part.

As you've probably experienced the measured PSF needs to be
preconditioned.   At the very least it should be centred and made the
same size as the image.  It is sometimes denoised and made (at least
laterally-radially) symmetric.  There are many variations of psf
preprocessing routines.  Some involve preconditioning a single bead.
Some average multiple beads.  Another alternative is to solve for the
PSF using a "reverse deconvolution" on an object of known size and
shape (usually a field of beads which in this case do not need to be
sub resolution).

Huygens has the option of solving for the PSF through a reverse
deconvolution scheme they call "distilling".

So it could be that the PSF pre-processing and/or distilling (if you
are using it) in the Huygens software is making a difference.  I don't
know for sure but it would be something to investigate.  You should
see if Huygens allows you to export the preprocessed/distilled PSF and
if so take a look at it.

Brian