O.K., so we won't kill the thread. I guess if anyone truly gets sick of
this they could just keep deleting messages with this subject line. I
would like to respond to Eric's last message, but that will take longer
than I have right now to compose. In regard to Aryeh's message:
 
I think your idea would work, except I think that I would take a very
careful look at the illumination system on the confocal and the accuracy
of the focus motors.
 
The illumination that we use incorporated a "scramble" type apparatus that
homogenizes the light coming into the microscope. This then gives us a
very uniform field of illumination (~5% shading error) compared to more
traditional methods. This also very effectively fills the entire back
apperature of the objective. Shinya Inoue presents an argument for DIC
that this greatly increases the axial resolution of the system. This is
consistent with Hiraoka's findings in "Hiraoka Y., Sedat J.W., Agard D.A.
(1990) Determination of three-dimensional imaging properties of a light
microscope system: partial confocal behavior in epifluorescence
microscopy. Biophys J 57:325-333." and is also consistent with our
observation that the the axial resolution that we are have in our raw
images (pre-deconvolution) is as good as we have seen on any confocal,
including the latest Leica.
 
Second of all, I would suggest that you take a good look at the focus
control on your MRC1024. Most stepper motors behave with about 100 nm
resolution +/- 100nm. This error, even on a step of 1 um is big (10%, I
calculated this all by myself :)) This is a pretty sizable error for any
sort of measurements that you might want to make and it will have a
deliterious effect on any deconvolutions that you may wish to do (personal
communication, Carl Brown, API). I would at least try to test the
accuracy of the axial steps before just buying off on their performance.
 
More later.
 
________________________________________________________________________________
 
 
Paul Goodwin
Image Analysis Lab
FHCRC, Seattle, WA
 
On Thu, 6 Jun 1996, Aryeh M Weiss wrote:
 
> ok -- I have finally been tempted to conribute to this discussion. We have an
> MRC-1024, with air cooled 100mW argon ion laser, so no UV. I am fairly new at
> the game, but noone has yet brought me a sample which I could not image (given
> that the dyes will excite with my laser. But then again, I have not yet done
> anything really tricky, and no fast kinetics or ratio imaging yet.
> Ideally I would like to have both methods available, but if there is to be only
> one system, I cannot imagine only having a system where it could take many
> minutes to many hours to get the images. So I think that the first system has
> to be confocal. Of course, in a couple of years the computing will be 10 times
> faster and 5 times cheaper and then the equation may be different.
>
> But I would like to suggest how I would modify my system to get the best of
> both worlds, and then see if anyone out there decides to do it. Because I
> cannot for the life of me figure out why both technologies don't exist on a
> single instrument.
>
> I put a digital CCD imager on the trinocular head of the Zeiss 135M which is
> coupled to our MRC-1024. I also have a Uniblitz shutter in front of the Hg
> lamp. The Z-motor of the MRC-1024 can be controlled independently through the
> control unit. (BioRad said they would provide the serial port commands to do
> that, and one day I will probably get them). Those are the ingredients we need
> for a digital deconvolution system. The rest of the process is software, and
> not even complicated software for acquisition (perhaps I am missing something
> on this point?). The data acquisition program can acquire an image series, and
> offload it to your favorite workstation for processing.
>
> I dont know how simple this would be to do on other confocal systems, but I
> know that on the MRC-1024 it would be pretty straightforward.
>
> If there are fatal flaws in this dream I would be glad to hear about them.
> --aryeh
>
>
>
> --
> Aryeh Weiss                          | email:     [log in to unmask]
> Department of Electronics            |
> Jerusalem College of Technology      | phone:     972-2-751146
> POB 16031                            | FAX:       972-2-422075
> Jerusalem, Israel                    | ham radio: 4X1PB/KA1PB
>