It’s been a long time since I posted – I’ve been kept busy with many things here, one of which has been demoing a new spinning disk confocal for the shared instrumentation grant we were recently awarded. Last week we demoed the Borealis CSU-W1, which performs very well. The Borealis upgrade really does result in impressive light delivery to the sample – with 150 mW lasers, we were getting about 25 mW of light at the sample. Even with the large field of view of the CSU-W1, this is way more light than we need to image samples. This is good because it means we can save on hardware by buying cheaper low power lasers.
While we were doing the demo, we took advantage of the dual camera ports on the CSU-W1 to compare different cameras. In particular, we compared the Zyla 5.5 to the Zyla 4.2. I’ve known that the Zyla 4.2 is the more sensitive camera on paper (going from 5 to 4 transistors per pixel improves the quantum efficiency from ~60% to ~72%) but I didn’t realize just how much a difference this makes in practice. Below, you can see images from both cameras. These were acquired one after the other with the same laser power and exposure time, and the images are autoscaled to saturate the brightest and darkest 0.01% of the pixels).
As you can see, the image from the Zyla 4.2 is of substantially higher quality. This appears to be due not only to the higher quantum efficiency of the Zyla 4.2 but also to lower fixed pattern noise from the Zyla 4.2, although I have not carried out any quantification of these images. As expected, we did find that for the dimmest samples an EMCCD camera outperforms the Zyla 4.2, although the field of view of the EMCCD is substantially smaller.
Given that the performance of many of the objectives we have looked at is poor at the edges of the field of view, I would generally choose one of the 4 megapixel sCMOS cameras over the 5.5 megapixel ones.