I don't think that I am alone in thinking that the questions that Xuejun
asks are of general interest to many on this list, so I hope responses come
back to the list.

I think that if you want speed, UNIX (SGI) is the way to go.  The speed
comes not from just the CPU, but also from how fast the bus can move data
around, something the SGIs excel at.

We are currently evaluating deconvolution systems as well.  Inovision is one
of the systems we are seriously considering.  In terms of image quality,
speed, and price, they seem very competitive.  I think the user interface
could use some additions and clarifications, which could come very soon.  At
present, I'm a little reluctant to recommend this for a multi-user facility.

I evaluated Deltavision more than a year ago.  The user interface impressed
me with how organized it was.  One big con is the price.  The other con, at
least for us, was the lack of adaptability to datasets from sources other
than their acquisition module, particularly if one wants to import PSF data.
Inovision has this problem too, but not to the same extent.  This seems to
be a general problem for (non-blind) deconvolution systems.

We also looked into the Windows-based MetaMorph/AutoDeblur system, since we
already had MetaMorph.  As some on this list have stated, MetaMorph is
powerful and user-friendly.  The main drawback with AutoDeblur is the
practical xy size limit of images, which is 256x256.  Anything larger has to
be processed in chunks, so I've heard.  I don't think this is the system to
go to for speed.

As to cameras, Kodak has recently introduced a new chip that is not
back-illuminated, but has a QE that is twice that of front-illuminated CCDs
in the green.  I read that Roper (Princeton and/or Photometrics) is
developing a camera with this chip.  I don't know anything else about it,
but if the price is right, it should be popular.

Someone else suggested you try before you buy, and I second that motion.


Jeff Reece
Biomedical Engineer
National Institute of Environmental Health Sciences (NIEHS)
Confocal Microscopy Center
P.O. Box 12233, MD F2-02
111 Alexander Drive, Bldg 101, Rm. F219
Research Triangle Park, NC 27709
Ph: (919) 541-0311
Fx:  (919) 541-1898
Email:  [log in to unmask]






> ----------
> From:         Xj Sun[SMTP:[log in to unmask]]
> Reply To:     Confocal Microscopy List
> Sent:         Tuesday, February 23, 1999 5:37 PM
> To:   [log in to unmask]
> Subject:      3-D software/computer and CCD camera
>
> Hi, Everyone:
>
> We are looking into a standalone 3-D image processing/analyzing
> workstation
> and a back illuminated CCD camera. The primary usage we need to perform on
> the workstation are:
>
> 1) De-convolution.
>
> 2) 3-d image processing/measurement. More specifically, we need to get
> integrated density measurement of 3-d/4-d (3-D time-laps) experiments.
> Also,
> possibly, we need to measure of co-localization of multiple stains in 3-d.
>
> We do have a PC based de-convolution package now. The problem is speed.
> Even
> with the fastest computer we can get and maximum amount of RAM we can put
> into, it is still fairly time consuming with the data set we have.
>
> The questions I have are:
>
> 1) Any recommendations about particular software and computer platforms
> for
> all above applications? I know it is a difficult question. I am
> particularly
> interested in the experience of those who use a Unix platform for the
> above
> purposes. It is intended  for use in a multi-user facility.  Consequently,
> ease of use and flexibility of software is a main concern.
>
> 2) If anyone has compared PC based and Unix based system for such
> applications, I would particularly like to know the pros and cons of the
> two
> platforms relative to each other.
>
> 3)  If anyone has opinions on particular back illuminated CCDs.  Our key
> requirements are a) high sensitivity in the green wavelengths, b) small
> pixel size (no greater than 10 micron pixels), c) relatively fast frame
> acquisition with a greater emphasis on low dark current and readout noise
>
> I thank, in advance,  those who respond. Please respond to me directly
> instead of to the list (unless you think it is of generally interesting).
>
> Commercially motivated responses are welcome. Nevertheless, we are in the
> early stage of thinking whether it is worth to pursue further. So there is
> no definitive plan for this yet.
>
> Best wishes to all.
>
> Xuejun
>
> ***********************************************************************
> Xuejun Sun, Ph.D.
> Dept. of Experimental Oncology
> Cross Cancer Institute
> 11560 University Ave.
> Edmonton Alberta T6G 1Z2, Canada
>
> Phone:  (780) 432 8898 (office)
>         (780) 432 8468 (lab.)
> Fax:    (780) 432 8892
> Email:  [log in to unmask]
>