DSLR Lenses and UHC Filters

A pleasant surprise

Properly dark skies are pretty rare in the North-West of England, never mind in suburban Macclesfield. So I've experimented with didymium filters on my DSLR lenses - definitely better, but not dramatically so. I wanted to try the more sophisticated characteristics of a diffraction-based UHC (Ultra High Contrast) filter. Sadly they're considerable more expensive and typically only available with telescope eyepiece dimensions (1.25" or 2" diameter) - too small to fit at the objective end of a DSLR lens. One company does do UHC filters which fit inside the body of a DSLR, but only for full frame models - no use to me and my APS-C Canon 550d.

Looking at my EF 75-300 f4-5.6 lens (the one I've been using recently), the rear aperture looked a bit over 1" diameter. Could I fit a 1.25" filter in there without fouling the mirror or the lens optics? Time to get a measuring stick...

1.25" filter in 70-300mm lens (550d body).

The filter has a smaller diameter than the size of the rear baffle aperture, so that's OK. A wire cradle will hold it in place. The only issue is that at the short (70mm) end of the zoom range the rear optical group will hit the filter. I don't plan to use this lens at 70mm for astro photography, so that's not a problem.

The filter screws into the spring-loaded cradle which is a push-fit into the lens baffle.

What about other lenses? A 50mm f1.8 Mk1 proved even simpler. The rear baffle aperture is just bigger than the filter thread and the filter body is just big enough to rest on the 'shoulder' in the camera body which stops it intruding into the 'no-go' area swept by the mirror.

Filter in 50mm f1.8 lens.

The baffle aperture is slightly bigger than ideal, so I ended up making a wire gasket to fit round the filter thread to keep the filter centralised.

1.25" filter in 50mm lens (550d body).

Next up was a Sigma 10-20mm - a bit more tricky. The rear baffle aperture is smaller than the filter thread diameter, but it's just that: a baffle. And held in by three screws. It was work of minutes to take out the baffle and grind out the aperture so that the filter thread would just fit the same way as in the 50mm lens, but with a tight fit.

Where things get a bit marginal is that at 10mm focal length, the rear group glass extends beyond the lens mount. In practice, this rear glass is smaller than the filter and extends only into the unoccupied space in the filter. Lucky!

1.25" filter in Sigma 10-20mm.

Finally, does it fit in a Samyang 135mm f2? This was a gamble because I didn't have one but intended to get one. Looking at the online pictures suggested the rear baffle diameter would be big enough, but that the rear lens group was dangerously close to the mount. And too big to rely on the '10-20 trick'.

I gambled and got the lens. Phew, no problem - the filter fits in the baffle much the same way as in the 50mm lens (I used an bit out of an old dead lens to support it) and clears the rear lens group comfortably.

So, was it worth it?

With mainly overcast sky in early December, I was surprised to see some stars one evening after dinner. Too good a chance to miss! It might have been clear, but the air was very humid and hazy which made the light pollution in the garden worse than usual. I tried a control shot with no filter...

Single frame, 135mm f2 30s ISO400, RAW - no adjustments, daylight WB.

30 seconds was enough for the pollution background to limit the exposure. Drop in the UHC filter and now a 2 minute exposure is limited by the brightness of the centre of M42 rather than by the pollution. There's even a hint of the Flame and Horsehead nebulas.

Single frame, UHC filter, 135mm f2 120s ISO400, RAW - no adjustments, daylight WB.

How about if we get stacking? The temperature is dropping and the humidity is horrible, but it's worth a try. First without the filter...

No filter, 135mm f2 30s ISO400, 18 subframes + darks, flats & offsets plus single 10s subframe for Trapezium.
Processed with DSS (stacking), FITS Liberator (colour balance & stretching) & Photoshop (back-end).

... and now with:

UHC filter, 135mm f2 120s ISO400, 32 subframes + darks, flats & offsets plus single 10s subframe for Trapezium.
Processed with DSS (stacking), FITS Liberator (colour balance & stretching) & Photoshop (back-end).

The halos around the bright stars were caused by the last dozen frames starting fog up. I could have left them out of the stack, but they don't really detract from a non-scientific picture.

Finally, a cropped version...

UHC filter, 135mm f2 120s ISO400, 32 subframes + darks, flats & offsets.

... and a full resolution look at the Horsehead nebula region:

Horsehead and Flame nebulas.
Identifiable stars down to m=17.35 (USNOA2 0825-01625473)

Worth it? I think so.

A few final notes:
1. The purpose of the exercise was test the use of a standard 1.25" UHC filter (Explore Scientific) with DSLR lenses, not to get the best possible astrophotograph of M42 or the Horsehead nebula. There are some obvious colour gradients introduced by the bias frames (for some as yet unknown reason) and also longer exposure subframes would have been better. I probably should have used ISO800.
2. The Samyang 135mm f2 is amazingly sharp edge to edge, even wide open. But you better have a Bahtinov mask for focussing.
3. There was no obvious vignetting from the filter, but then I was using an APS-C camera. No promises if you use a full frame body.
4. Getting a 'normal' colour balance in the image with the UHC filter is quite tricky given that it block a great swathe of colours in the middle of the visible spectrum. NASA's FITS Liberator is much better than PS's 32 bit FP to 16 bit INT conversion, but it's not an easy tool to use. The 550d is not 'modded' in any way, so the extended read sensitivity is actually pretty poor.
4. Next mini-project: A dew heater!

______ Closing credits ______

Tracking provided by my home-made barn-door tracker...

Samyang 135mm f2 (with hood) on Canon 550d.