---------- Forwarded message ----------
Date: Sat, 19 Oct 1996 07:49:00 -0700
From: [log in to unmask]
To: [log in to unmask]
Subject: Message not deliverable
----------------------------------- Returned -----------------------------------
From: [log in to unmask] at @UCSD
Date: 10/19/96 8:39AM
*To: [log in to unmask] at @UCSD
Subject: Re: irreproducible z-axis sectioning of Zeiss CLSM
-------------------------------------------------------------------------------
The Z-section inaccuracy and non-reproducibility is also a problem with
BioRAD scopes. They use a friction coupling motor that is free standing on
the desk. A knurled cup presses on the fine focus. In order to obtain
accurate double label on the earlier MRC 600, we used the split screen
mode to capture two channels simultaneously. This assured us that the
matching images were truly at the same depth (ignoring refractive
wavelength errors). If you do this with triple labeling on the MRC 600,
the problem still had to be dealt with.
On the MRC 1000, with Lasersharp 1024 software, you can obtain full frame
images of three fluorophores simultaneously.
But the problem is also in the accuracy of return of the fine focus on the
microscopes themselves, so BioRAD does not bear the full responsibility
for this, as your note correctly points out.
Leica uses a Piezo Z-axis, and claims much better accuracy in the Z-plane.
Their piezo z drive is, I believe, mounted on the stage itself, and moves
the slide vertically. Expensive, but a neat solution. I don't know the
actual numbers, however, in regards to accuracy.
regards, Harvey Karten
On Sat, 19 Oct 1996, Donna Arndt-Jovin wrote:
> I wish to bring to the attention of the newsgroup problems in z-axis
> sectioning which we have in the hopes that we can solve this to the
> satisfaction of our group and others.
>
> The Zeiss CLSM 310/410 equipped with Marzhausen Z axis drive and stages
> have an accuracy of +/-50 nm in the z axis. The positional readout has a
> better accuracy. As the software is presently written for z scan series the
> command to move the z axis drive is given only once and there is no
> feedback loop to check whether the actual z position is reached. This
> means that in making a z series of 0.2 micron steps the actual z sections
> can vary between 150 and 250 nm (that is +/- 50%!). In addition, the z scan
> motor resets the stage to the original position at the end of the scan but
> this can also vary (on our instrument) by more than 0.15 microns from the
> original position. Thus, in making a series of z scans for different
> wavelengths the actual z sections can be over 0.3 microns different from
> one another in the two stacks. These measurement artifacts compromise
> colocalization data and make deconvolution of the stacks impossible (either
> to make an accurate PSF measurement itself or to calculate the
> deconvolution which assumes the sections to be of uniform thickness). We
> feel that this problem with the Zeiss CLSM is of general importance to
> confocal users and have tried to make this clear to the firm
> representatives for almost a year now and with intensive badgering for the
> last month. We offered to write a new z scan macro with a feedback loop
> for testing the z position which would allow the motor drive up to 5
> attempts to reach the proper z position (our experience with this stage and
> an older controller software which we wrote ourselves showed that only 3
> attempts were necessary to reach the desired position). To date, Zeiss has
> not seen fit to make the driver software available for our modification nor
> have they produced a new program on their own.
>
> My questions to the newgroup are the following:
> a. how accurately do other CLSM's make z scan series?
> b. has anyone written a macro to overcome the problem described above for
> the Zeiss instrument?
> c. aren't other groups concerned about measurement inaccuracies and
> shouldn't the instrumentation companies respond to these concerns?
>
> I would appreciate any code which users have developed which may help us
> overcome the laborious manual adjustment for very accurate measurements.
> Would they please send this to me directly? General comments to
> measurement inaccuracies of this and other types should be aired on the
> newsgroups to allow everyone to profit from the discussion.
>
>
> Donna Arndt-Jovin
>
> Donna Arndt-Jovin
> MPI for Biophysical Chemistry
> Dept. of Molecular Biology
> Postfach 2841
> D-37018 Goettingen, Germany
> tel: +49-551-201-1393
> fax: +49-551-201-1467
>
>-- Saved internet headers (useful for debugging)
>Received: from listserv.acsu.buffalo.edu by mail.ucsd.edu; id IAA03115 sendmail
>Received: from listserv.acsu.buffalo.edu (listserv.acsu.buffalo.edu [128.205.7.
>Received: from UBVM.CC.BUFFALO.EDU by UBVM.CC.BUFFALO.EDU (LISTSERV release
>Received: from UBVM (NJE origin SMTP@UBVM) by UBVM.CC.BUFFALO.EDU (LMail
>Received: from franklin.sdsc.edu by UBVM.cc.buffalo.edu (IBM VM SMTP V2R3) with
>Received: from hydra.sdsc.edu (hydra.sdsc.edu [132.249.20.50]) by fran
>Received: from localhost by hydra.sdsc.edu (4.1/1.11-client) id AA02907; Sat,
>Mime-Ve
>Content-Type: TEXT/PLAIN; charset=US-ASCII
>Message-ID: <Pine.SUN.3.92.96101908272
>Newsgroups: bit.listserv.confocal
>Date: Sat, 19 Oct 1996 08:39:22 -0700
>Reply-To: Confocal Micr
>Sender: Confocal Microscopy List <CONFOCAL@UBVM
>From: Harvey Karten <[log in to unmask]>
>Subject
>To: Multiple recipients of list CONFOCAL
>In-Reply-To: <v01510104ae8d2d382789@[134.76.211.69]>
|
