ICX 618 vs ICX 098 comparison measurements

This is my first post here, let's see how it goes.
Thanks to a fellow astromarter, I acquired a camera with the Sony ICX 618 chip (Flea 3 FW-03S1M mono, no IR filter). I had a hard time finding a direct comparison with the older Sony ICX 098 chip. Since I still have my first planetary camera with the 098 chip (DMK21 AF04.AS) I wanted to find out they compare.
This was my setup:
Both cameras with the detector about 10 inches from a LCD monitor with an Astronomik luminance filter (to filter out IR and UV). Camera was controlled with Astro IIDC. On the monitor I put a uniform color images made in photoshop White (RGB set to 255), Red (R255 G0 B0), Green (R0 G255 B0) and Blue (R0 G0 B255). I took images with a gamma of 1 and gain set to the minimum setting. Exposure times were 6 ms for red, green and blue and 3 ms for white. Histogram analysis of the entire frame was done with imageJ to determine the gray value of the image and standard deviation as a measure of image noise ( with 0=black and 255=white). Following are the grey levels for the ICX 618 and ICX 098 in parenthesis.
White 203(161 - 21% lower), Red 117 (74 - 38% lower), Green 201 (164 - 22% lower) and Blue 104 (91 - 12.5% lower).
Interestingly for the images with the 618 chip the standard deviation was usually 5-10% less, suggesting a lower noise level.

I then put then looked at what exposure time the image was saturated (white). For the 618 chip the exposure times for saturation were about 25% shorter White, red and green, but less than 10% shorter for blue. This was only a very crude test.

When I took the luminance filter off, the differences between the 2 chips became much more pronounced, indicating the higher sensitivity to infrared of the 618 chip (the front of the monitor is warm to the touch).

In conclusion, I expect to see about 25% shorter exposure times with the 618 chip compared to the 098 chip.

P.s. I will update this once I have a UV pass filter and IR pass filter to see how the chips compare for IR and UV.

Any feedback is appreciated.

It would be interesting to compare the signal-to-noise ratios produced by the two cameras for some identical exposure. For example, you could expose both cameras with the same illumination and exposure time, and then compute the average counts in some region near the center of the detector divided by the standard deviation for the same region. This has the advantage that it is independent of the gains (counts per detected photon) in the two cameras. You could also do it for different light levels or exposure times. In the limit of bright targets, the answer will depend only on the quantum efficiency of the detectors. But in the short exposure / faint target limit it will also depend on the so-called "read noise" of the camera.

The number of counts produced by the cameras is not a great indicator of the performance, since the counts per detected photon is somewhat arbitrary. One manufacturer might set the gain to produce 2 counts per detected photon, while another might set it at one count per detected photon -- so just because one camera produces twice as many counts as another does not necessarily mean it is twice as sensitive.