I was wondering, as there are filters coating on pixels of a camera, can I just keep the green channel of the image and remove the need for emission filter?
I want to use a raspberry pi HQ camera for fluorescent microscopy and I want to make it as small and cheap as possible.
“can” will depend on the spectrum of what you’re imaging, the specifics of the bayer filter spectrum, and the signal-to-noise-ratio demands of your experiment.
cheap as possible will always entail some compromises… and the trick is just to decide whether those compromises are worth it for you. So, specifically, try to determine how much light you’re losing by using the built-in-filter. (looks like that chip might have a Sony IMX477 … so maybe start there when looking for specs on the builtin filter?). Then, compare it to whatever you would be using otherwise.
Hello. I agree with what @talley said. I have some practical experience here with fluorescein (what is your fluorophore?) and a key thing with fluorescence is to efficiently and selectively block the excitation light from getting to your camera while allowing the fluorescence emission to pass. Unless you are looking at uber-duber bright clumps of flurescein (and nothing more subtle) then you may get away with some low contrast result. Otherwise, not only is the Bayer green inadequate to block blue excitation light but also cheap yellow filters (like Cokin) will also be inadequate. That is why I opted for a decent (albeit relatively low cost) narrow band-pass filter for the PUMA system. The thing about low cost Bayer colour cameras is that they interpolate from adjacent pixels to each pixel position so even if your Bayer green was precise narrow-band (which is isn’t) then you would still get significant interpolated bleed from the adjacent pixels (like the blue ones) unless you can read each pixel value pre-Bayer (and I don’t think this is possible with these low cost camera chips - the Bayer interpolate is what is fed out of the chip as the R, G and B channels).
So short answer, in my opinion: No (but hey, prove me wrong!).
Thanks everyone! @P_Tadrous @talley
yeah I just didn’t consider the excitation spectrum. I am using Eva green for Real Time PCR purposes and here is the Bayer spectrum of the IMX477 which just passes lots of excitation wavelength even in green channel.
