Dear All,

I'm having trouble with the linearity of the readings of our Zeiss LSM
410.

I have been controlling the PMT and amplifier gain separately, which
seems to be great because of the linearity of the amplifier gain.

However, when I plot the image intensity vs. the amplifier gain
(constant PMT  voltage and brightness), the curves look more or less
like this:

Intens|
      |
      |         /        /
      |        /       /
      |       /      /
      |      /     /
      |     /    /
      |    /   /
      |   /  /
    0 |--/-/--------------------------
      | //            Gain
      |/
      |

That's the case for either mean intensity of reflection images obtained
with two  ND filters of known (measured) transmittance or integrated
intensities of two  calibration beads (Molecular Probes InSpeck 100% and
29%).
The linearities are high (R2 > 0.999)
Moreover, if brightness = 9555 the ratios of the the slopes correspond to
the  expected image intensity ratios (actually, this is what I used for
finding out the  appropriate brightness setting for my system - any
comments?).
The big problem is, of course, that the intensity vs. gain curves DO NOT
EXTRAPOLATE to 0.

That clearly means that the ratio of intensities of two images collected at
a  certain gain will be different from the ratio recorded at a different gain
(same  PMT voltage in all cases). So one never knows which is the true
ratio, unless  the slopes are compared.

In other words, suppose I measure the mean reflection intensities of an
interface using a 2% transmittance ND filter and a 10% one. The ratio of
intensities should be 5 at all gain levels. But due to the fact that the
curves do  not extrapolate to 0, the ratios vary. However, slopes are 5
and, as I said above,  that's what I used for finding the appropriate
brightness settings (brightness <>  9555 will give either higher or lower
slopes).

The way things are, I must assume that I cannot do any accurate ratioing
experiments (except for high gains, where the offset problem becomes
less  important).
I know that there's a lot of people out there that are ratioing LSM 410
images,  so I suppose that either they've solved this problem or never
had it at all.

I contacted Carl Zeiss some time ago, but I haven't got any reply yet.

Does anybody know what may be happening?

Thanks in advance

Carlos