Subject: | |
From: | |
Reply To: | |
Date: | Mon, 5 Mar 2007 11:11:46 +0000 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
Hi Marc,
The offset control on Orca cameras is an analogue voltage offset - it
changes the voltage from the CCD that the camera considers to be
"black". With newer Firewire Hamamatsu Orca cameras, Openlab
expresses this as a factor from 0..255, with older Orcas that use a
Snapper card, Openlab expresses this as a percentage of the 0..255
range that Hamamatsu provide. Note that the factor in both cases is
(as far as I know) in arbitrary units - it does not correspond to a
particular number of levels.
One thing to note about Hamamatsu cameras is that their DC offset is
set somewhat higher than cameras from other manufacturers. In
practice this means that with no light going to the camera it will
read between 180-220 levels. This is a deliberate feature -
Hamamatsu make sure that the background signal is not clipped, this
allows users to measure the background properly and take it into
account in their analysis.
If you are measuring weak signals this DC offset must be taken into
account either in the analysis or in the imaging by using background
subtraction. It would also be possible to use the offset control on
the camera to clip out the DC offset (and possibly this is what your
user is doing), however I feel you should be cautious about doing
this and ensure that you really are just clipping background and not
signal. If the images are acquired with the background clipped there
is no way to go back and check later!
Finally, with any instrument it is advisable to test the accuracy of
the instrument. There is a paper on our website that discusses
measuring linearity in intensity measurements. It is aimed primarily
at the Optigrid, however the technique could be used to check the
linearity of any fluorescence imaging system:
http://www.improvision.com/pdfs/technotes/TN424.pdf
Cheers,
Steve
On 3 Mar 2007, at 3:20, Marc Green wrote:
> Search the CONFOCAL archive at
> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>
> List,
>
> I am having some difficulty understanding exactly how "offset" as
> used by
> OpenLab talking to a Hamamatsu ORCA is being applied.
> My initial understanding of how this sort of function was applied
> was that a
> fixed amount of current coming off the chip would be deducted from
> every
> measurement, "setting black to a given value".
> In OpenLab it is expressed as a percentage, but a percentage of
> what? The
> range of intensities in a given captured frame? Or a percentage of
> all 12
> bits the camera can produce? In a few quick snapshots with
> transmitted
> light I have failed to convince myself that it is working in a linear
> fashion.
>
> This has all come up because a user imaging YFP has reported that a 5%
> offset (as opposed to 0) is making all the difference in detecting
> signal
> over background, eliminating the need for histogram stretches.
> I am concerned that a blanket hardware offset at collection (as
> opposed to a
> calibrated pixel by pixel table) may be skewing (perverting?) the
> data when
> we are working in the bottom end of the camera sensitivity.
>
> Thanks for your thoughts,
> Marc
>
> ------------------
> Marc D Green
> [log in to unmask]
> Pombe.net
Steve Baxter
Software Development Manager
Improvision
+44-2476-692229
Image Processing and Vision Company Limited (trading as Improvision)
is a company registered in England and Wales with company number
2505778. VAT number GB536188722.
|
|
|