Synedra superba diatom with 405nm oblique illumination

[jmc]

A few images of a Synedra superba diatom slide by Arthur Cottam. Imaged using 405nm light with bright field oblique illumination on my modified Olympus BHB and using a monochrome converted Nikon d800. Light source was a 50W mercury xenon lamp. Images reduced in resolution for sharing, apart from the closest crop from the 60x objective, which is still at the original pixel resolution. Objectives were a 40x Olympus DPlan Apo UV (NA 0.85), and an Olympus 60x SPlan Apo (NA 1.4), and the condenser was an Olympus Aplanat Achromat NA 1.4 set to oblique. Oil immersion on the objective and condenser when the 60x was used. Single images (no stacking).

First with the 40x DPlan Apo NA 0.85.

And then with the 60x SPlan Apo NA 1.4.

60x SPlan Apo, cropped in to show more detail.

60x SPlan Apo, cropped tightly (original pixel resolution).

Finally, the slide itself.

[Lou Jost]

Spectacular level of detail. Congratulations. I'm curious how you converted the D800 to monochrome.

[jmc]

Spectacular level of detail. Congratulations. I'm curious how you converted the D800 to monochrome.

Thanks Lou. MaxMax have done the monochrome conversions I've got - https://maxmax.com/
Basically the Bayer filter and microlens array is removed, and then a clear window put back over the sensor to protect it (usually WG280, or these days I request fused silica in case I want to image below 300nm). As for exactly what he does, that's his secret. Even having had a few cameras done for different jobs, I still haven't been given the details.....

[Macro_Cosmos]

A pencil can be used to scrape off the Bayer Array, though I assume his method is less evasive and safer.

[jmc]

A pencil can be used to scrape off the Bayer Array, though I assume his method is less evasive and safer.

It is something more complicated than that (although I have seen that approach give usable results). I'm not sure though if he removes it physically, chemically of by etching (with a plasma or ion source).

[Lou Jost]

Thanks for that information. I looked for reliable companies to do it but was unsuccessful for the brands of cameras that I own. I will look onto MaxMax.

I ended up getting a monochrome astro camera with a cooled sensor. The cooled sensor is very useful; you might want to experiment with that as another way to improve the resolution of images.

[Macro_Cosmos]

I have learned (RIP wallet) that actively cooled cameras requires a lot of care, rendering used copies a bigger gamble than used objectives.
Always check the QE curves, these cameras usually come with a window installed, making them insensitive to UV and IR light.

[Lou Jost]

The window seems thin; I don't think it blocks IR light. Probably does block some UV light, but it will pass the 405nm light used in this post.

[jmc]

MaxMax have converted some very varied cameras, including astro ones and medium format ones. It is not cheap though, so be sat down when the price comes back.

One of the biggest issues currently for UV imaging is the coverglasses that the sensor manufacturers use tend block quite a bit of the UV starting at around 400nm and then blocking most of it by the time you get to 300nm. I recently got a monochrome astro camera with a Sony IMX571 mono sensor, and this had one of those blocking coverglasses on it. I've had that removed now, and replaced with fused silica, an will hopefully be getting it back next week. Nobody seems to be sure about whether the microlenses will block the UV (or whether they block some of the UV but the focusing they provide outweighs the blocking). This will hopefully answer that question.

Internal filters inside the cameras can certainly be an issue, especially around 405nm. Some colour cameras tend to block from around 410nm and below, but others cut off close to 400nm. Ideally check with the manufacturer and as for QE curves as mentioned.

I got into microscopy after doing UV imaging, so already had suitable cameras thankfully.

[Lou Jost]

The thing about cooled astro cameras is that there is virtually no penalty for long exposures. We often use hours-long (or even days-long) exposures to capture faint nebulae. So as long as the blockage is less than 100% complete, and as long as your subject is not moving, you can get good well-exposed photos. Your exposures are only limited by the purity of your UV source; the amount of leakage in the visible spectrum will limit what you can see. Even then it seems like there are things you could do to get more UV signal. You might be able to subtract out the visible light by taking the same photo, with the same UV illuminator and same exposure, but with a good UV filter, and then subtracting that from your UV photo.

[jmc]

The thing about cooled astro cameras is that there is virtually no penalty for long exposures. We often use hours-long (or even days-long) exposures to capture faint nebulae. So as long as the blockage is less than 100% complete, and as long as your subject is not moving, you can get good well-exposed photos. Your exposures are only limited by the purity of your UV source; the amount of leakage in the visible spectrum will limit what you can see. Even then it seems like there are things you could do to get more UV signal. You might be able to subtract out the visible light by taking the same photo, with the same UV illuminator and same exposure, but with a good UV filter, and then subtracting that from your UV photo.

Don't get me wrong, I get the appeal, and long cooled exposures are great. As you say, eventually it comes down to whether other wavelengths are contaminating the image, and if so how to deal with that. Camera sensitivity in the UV is low compared with visible and IR. Even with no glass in my system which would absorb some or all of the UV depending on wavelength, when I do my 313nm imaging (using a mercury xenon lamp), I have to have 2 x 313nm OD4 bandpass filters stacked together as the final step before the light gets into the camera, as one is not enough to block enough of the unwanted wavelengths.

[Macro_Cosmos]

I will ask MaxMax if they offer sensor enclosure resealing, I have unfortunately lost access to my university's lab because the manager got laid off in a round of savage staff cutting. I was going to do it myself and write about it...

Edit: They can do it.

[Lou Jost]

Returning to the question of the spectral response of astro cameras, my QHY163 monochrome camera has a quantum efficiency of 40% at 675nm, 20% at 810nm, and 5% at 925nm.

This is not gas-filled because it has a replaceable silica cell to keep the sensor dry.