MM-compatible, cooled camera for long exposure

Hello,
This question is not about micro-manager specifically, but maybe micro-manager users and DIY microscopists could answer. Please move/delete this post if it’s off topic.

I often work with FLIR (Point Grey) cameras. They’re very good, even for fluorescence.
Now, I am building a fluorescence module inside a 37ºC incubation chamber, to run timelapse imaging of weakly fluorescent bacterial colonies on Petri dishes. In these conditions, FLIR cameras become limited. Their dark current is low so long exposure (typically 1 second) and heat are not a problem for the background noise, for most of the pixels. However, at 37ºC and 1 second exposure, there are many “hot pixels” that give the image a salt and pepper look.

I’m looking for an alternative for a cheap-ish (I mean not an ORCA-Flash4.0) camera with cooling. Ideally, I need the same specs as a FLIR camera (2/3" or 1" sensor, 12 or 16 bits, 60-70 dB dynamic range, acquisition from 1 ms to 10 sec; I don’t need high FPS), with a cooling of 30-40ºC. Also, the camera needs to be controlled from micro-manager.

I looked into ZWO and ATIK, but it is not clear if there are MM device adapters for these cameras (I’m waiting for answers from them). Is there a manufacturer that I overlooked? Does anyone have recommendations?

Thanks for your help,
Max

Hi @MaxDeforet,

Great question! Looks like you need cooling, or at the very least low temporal noise even at higher temps. I looked at ZWO in the past, and they even have a MM device adapter. It only works with a specific MM version because they compiled it themselves and did not contribute the source code to our repository (which would include it in automated builds). You could ask them if they can share that code with us (may help their sales;)

You can search other machine vision camera manufacturers. I believe that Ximea has cooled cameras, and there may be others. Also, you can attach cooling fins to some cameras, and you could possibly build your own cooling unit (possibly even involving a water cooler outside of the incubator). Let us know what you end up doing!

1 Like

Hi Max,

Nico recently alerted me to your post. I share your views on both the qualities and limitations of machine vision cameras for microscopy. Having previously written MicroManager Spinnaker support for the FLIR range I have recently been writing drivers for a new inexpensive camera that is derived from the astronomy sector, but specifically re-engineered for bio-imaging under MicroManager. This first model “CellCam” has heatsinking and fan cooling, but if this isn’t sufficient for your needs then a subsequent model with TEC is close to completion. The camera has not been launched yet, but it is ready so let me know if you’re interested and I’ll try to get hold of one for you to try.

Best wishes,
Elliot

Hi Elliot,

[Thanks a lot for your Spinnaker device adapter!]

That sounds very promising. Is the first camera (with heatsinking) already available? Which brand is it?

Max

Hi Elliot, do you have any pictures of this camera or pictures taken with it?
Thanks in advance

1 Like

Hi Max,

The cameras are distributed by Cairn Research. A couple of photos have leaked out on Twitter (multi-camera support is one of the things that they’ve been testing quite hard, and I’ve been told it works well):

I don’t think they’re officially released yet but the first batch has been made and when I was chatting with the CEO he said he’s happy to loan you one to see if it works for you. I would suggest that you contact Cairn directly for more information, but I have already asked them to share some images on the forum.

Elliot

(p.s., sorry for the slow reply, I’ve been on annual leave the past couple of days)

Hi Elliot,
It looks very cool. I’m very interested in having it on loan. Do you know the specs of this “CellCam”? I’m going to get in touch with Cairn directly.

Max

Hi Max,

I don’t have the official spec sheet yet but I can tell you some stuff that I know off the top of my head. Main points are:

  • 5440x3648, 2.4um, rolling shutter sensor
  • Supports 1x, 2x, 3x and 4x digital binning
  • 12 bit ADC with 8 and 16 bit readouts on the mono version
  • USB 3 interface
  • I can comfortably get 19 fps on the mono in 8 bit full frame. The framerate does increase with binning or with a smaller ROI selected
  • I don’t think we’ve tested QE yet but other cameras based on the same sensor report a little over 80% QE for the mono and a bit less for the colour. I’d be very surprised if the CellCams were very different

Unfortunately, I can’t comment on noise, I haven’t taken any measurements yet. I suspect Cairn will have some data somewhere though