Karl is not at or with Photometrics and hasn't been for years.
I am a Senior Scientist affiliated with a large organization headquartered in Europe, and I serve as lead investigator on a number of projects concerned with developing a new generation of tissue-based molecular cancer diagnostics. The technology program I'm accountable for is hoped to contribute to more informed treatment decisions and improve the lives of cancer patients on a global scale. I define stringent performance specifications, and require the equipment I select to meet them.
This also means I'm accountable to be able to explain the synergy of technologies that allow us to extract this medically relevant information, to some of the mightiest regulatory agencies in the world. I need to know what my cameras are doing and how to qualify and verify them, and how to communicate, transparently and clearly. There is nothing to be gained by muddying of the waters or baffling with arcane theory in this arena.
My opinions are my own and do not necessarily reflect those of my current or former employers. I'm free to state my mind and say whatever I wish, although I hope most of it is of redeeming value. I'm footloose and fancy free as cannons go. This is precisely why I chose not to broadcast my affiliation, lest my opinions or advice be misconstrued as representing those of any particular organization.
For matters pertaining to legitimate scienctific discussion I can be contacted directly at my professional address:
[hidden email]
My interest here is primarily to help foster responsible exchange of useful practical information and tested theory and help the group to avoid misinformation. Yes, I have in-depth experience with cameras as well as characterization of numerous other analytical imaging technologies, it is my profession to develop and apply such technologies to relevant practice. I work in an environment where tight-tolerances and high-performance results are expected and there are no good excuses for anything below expectations. Technologies that I feel I can trust and refer to have earned it.
The slope of mean-variance plot will indeed be different at 0 gain on the EM port, this is due to addtional noise factors and reflected by increased variance, but the response should remain linear at different exposure times.
Sincerely,
Karl
On Thu, Apr 29, 2010 at 11:30 PM, Andreas Bruckbauer
<[hidden email]> wrote:
So i guess the question is if the mean variance method is working wel when using the EMCCD chip witout the non-EM port, the problem might be aditional noise. I guess Karl at photometrics has tested this extensively. If you want to varify you probably need a calibrated light source.
best wishes
Andreas
Sent: Wed, 28 Apr 2010 2:47
Subject: Re: EMCCD gain, photon scissors, bogeyman avoidance
Karl Garsha wrote:
>
> To simplify and look at the big picture, let us contemplate, for the
> sake of discussion. What if, we first calculate our gain under non-EM
> conditions, using the non-EM read port. (by the way, the read noise
> here on a good EMCCD should be around 5-6 e- which is well in the
> range of a respectable fluorescence grade sony 285 interline).
Which EM chip has this read noise? I'd be interested in getting it for
our next EMCCD. FYI the charge amplifier should age much more slowly
than the EM register so you only need to calibrate when you need to
calibrate (which for most imaging is never, as I said).
Cheers