Hello Confocalists!
I would like to thank everyone who responded to my question about digital
video-rate cameras. I have compiled a list of the responses, with names and
addresses removed, although most of these were already posted to the
general list. At the least, this list provides names and companies for each
of us to do our own investigating for our individual needs. For my needs in
particular, we were looking for video-rate aquisition, not just focusing,
color, and possibly fluorescence capture.
Thank you again for everyone's input, and I hope you find this list useful.
Kathy Spencer
I found very useful Olympus DP 11 (2.11 million pixels - C-mount) or DP 50
(1.50 millins pixels - C-mount), the second is cooled down by Peltier cell
and is total computer controlled, that's a good and bad things at the same
time, while the DP 11, not cooled down, can be used on different microscopes
because the camera not depends on computer and has an hand switch to control
the camera, if you want I can send you specification and data sheet.
----------------------------------
We use an Ademic 40 Mhz camera 1024 X 1024, 12 bit.
It operates in two modes quantitative capture or digital video. Greyscale
only
---------------------------------
Most of the digital cameras, with the exception of the slow scan, hi res
members of the families, now have near video-rate, at least for focusing.
Try:
Optronics (Magnafire)
Diagnostic Instruments (the new Spot RT)
DVC
Pixera Penguin
-----------------------------------------------
Stanford Photonics XR series cameras (which we do sell), can provide 10 or 12
bit images at full resolution 1300 x 1030, 1024 x1024 at rates of 15 and 30
ips, for higher rates you need to either bin or partial scan to get to 60ips
at12 bits (1024 x512) or 120 ips 10bit (1300x106) all from the center of the
CCD. Since the photons are few at these imaging rates (unless you blast your
specimen with illumination) we use a GEN III+ image intensifier to amplify
the photons. This also works well for just plain very, very, low light
conditions.
In fairness, Roper and others also have similar
cameras,
For some examples see the ICCD camera manufacturer's web sites of your choice.
----------------------------------------------
I get the impression that there is a bit of confusion regarding digital
cameras. The term has been used to refer to cameras that are based on
various popular still cameras (like the Nikon Coolpix) and also to refer
to high end scientific imagers. For people interested in rapid framing
at high resolution, the former are out of the question. The latter are
available in various models that provide 12 frames/sec, and more if you
use binning (lowering the spatial resolution but ehancing sensitivity).
I have had nice images with the DVC 1310 run at 12 frames/sec at high
gain. You can do the same for almost any camera based on the Sony IX085
CCD.
PCO makes one called the PixelFly that does the same. You can get them
for color or monochrome. All this does not require an intensifier. If
you can see it by eye then you can probably see it with the camera at
high gain.
If you have really weak signals and you need to see them at video rates,
then you can use an intensifier. But I wonder if the new Marconi chip
will take up some of that market and replace many intensified systems.
Then there are fast framing system that can operate at 100s or even
1000s of frames per sec. Frame sizes here run from about 256x256 and
down. They are expensive, but they can be purchased and will do thing
like motion analysis and such. With an intensifier I bet that they can
be used to image action potentials (using potential sensitive dyes) in
real time.
------------------------------------
A CCD-based system designed to do what he proposes ("... to image
action potentials (using potential sensitive dyes) in real time.") is
commercially available. For information about it, go to
<http://www.redshirtimaging.com>
-----------------------------------
> While these cameras have near video rate focusing, that is usually
>not the issue. The cameras in this class that I have tested
>actually capture at rates more like 2-4 frames/sec for full frame
>data. If you bin, you can go faster, but then you are at the
>resolution of an analog camera, and those can capture at video rate,
>and are significantly less expensive. This is one of the unstated
>disadvantages of digital cameras. If you want speed, and don't want
>to spend lots of money, you probably want an analog video camera and
>frame grabber, or an intensified video camera for more sensitivity
>with speed. I continue to be amazed at how effective my Dage video
>camera and Scion capture board with free NIH IMage software can be
>in this "digital camera" age.
---------------------------------------
As discussed by several respondents to this list,
there are multiple camera manufacturers who offer digital camera based on
the Sony interline CCDs, including Roper Scientific (DVC, PCO, etc). When
the Sony ICX085 or its windowless brother ICX061 are run with a high speed
ADC (like the 20 MHz ADC in the CoolSNAP FX) then you will get about 10 FPS
at full resolution. If you run with binning at 2x2 (reducing resolution and
pixel count) you can achieve 20 FPS. With further binning or subregion
readout, you can achieve higher rates.
I would suggest, however, that the new CoolSNAP HQ type cameras (and those
cameras which other companies are starting to make) with the Sony ICX285 CCD
will replace all the Sony ICX085 and ICX085 based systems in the near
future. That is due to the high red QE that can be obtained with this
device and the very low noise readout that the CCD allows. In the end, it
does not make sense to run fast unless you can actually see the faint signal
you are trying to measure. This new CCD, however, should
allow better data collection in the red region of the spectrum while
preserving the good sensitivity in the green that interlines normally have.
-----------------------------------------
Anyone interested in a new CCD for microscopy should look at the new
CAM 65 from Marconi (used to be EEV, call Ian Moody
<[log in to unmask]> for more details) which was shown at the SPIE
meeting three weeks ago.
MORE INFO AT
<http://www.marconitech.com/ccds/lllccd/technology.html>
It is a TV-rate chip with a (real!) difference. While working at
about 50 electrons of noise/pixel may sound useless (though not bad
for Video rate), the clever idea is that there is a GAIN section in
the horizontal register, just before the readout node, that can
operate as a series of very low gain avalanche diodes. When the
signal is low, this section multiplies small charge packets by about
200x, lifting even single electrons well above the noise floor.
So you have a normal CCD, that reads out so fast that there is no
point in cooling it, making focusing etc. quick and easy, that has
the QE of a CCD, the full resolution of a CCD because the charge
packet from one pixel is always handled as a discrete entity (unlike
intensified CCDs) and which can be made as sensitive as you need. Of
course, if you readout fast, there may not be enough time to collect
much signal but this isn't the camera's fault. On the other hand this
system breaks the old rule that you must get more readout noise if
you readout faster.
There IS one snag.
The charge amplification is not quite noise free. Because the exact
amount that the packet will be amplified depends on probability, the
camera as a whole operates as if it is noise free but has one half
the QE listed for the sensor alone. This reduction in effective QE
occurs only when you use the amplifier. If you turn it off or work in
the photon counting mode, the QE is as advertised.
When the sensors are available in back-illuminated models, they will
have QE of about 85% or 42% when used in the charge amplifier mode:
still very good.
The models now being introduced have pixels that the pixels aren't
square ( 20 x 30 microns) and cost between $10 -20k (analog vs full
digital). Later models will be back illuminated and 512x512.
I don't have much more info than this but I was very impressed at SPIE.
This seems like it should be a really good idea.
Kathy Spencer
The Scripps Research Institute
10550 N. Torrey Pines Road
IMM 24
La Jolla, CA 92037
(858) 784-9372
[log in to unmask]
|
