Eyepiece projection with the ENVIS objective lens is obviously a very popular way to do night vision astronomy.

No. It may seem like a distinction without consequence mathematically, but there is a huge difference in the real world. Here is a snippet from Televue's imaging methods page under the *afocal* section:

An alternative way of determining effective focal length considers the camera lens and eyepiece combination acting as a relay system. Using the example above, the 50mm camera lens with the 10mm eyepiece gives 5x relay magnification. 5x600mm telescope objective focal length yields 3000mm focal length.

The practical difference is that an eyepiece has many elements working together to make a well corrected field, and using it afocally is right on spec. Same with the objective of the NV device. A random pairing of an eyepiece and a camera lens forming an image at some random place usually leads to very small usuable fields with garbage at the edges, largely because the eyepiece is being used waaaay off spec. Garbage in garbage out, especially when it's intensified.

Even when it was still commonly used, eyepiece projection was reserved largely for planets because they are small.

Would you end up reducing the effective f-ratio by 12/55 = 0.21x? And what’s the impact of the 27mm Mod 3 eyepiece on all of this?

Think in terms of exit pupil to make your life much easier. Any lens you put on your NV device has an opening, called the entrance pupil. That opening constrains how much exit pupil it can handle. When the exit pupil matches that entrance pupil, the effective f/ratio of the whole system is the f/ratio of the lens you put on the NV device. When the exit pupil is smaller than the entrance pupil, the effective f/ratio goes UP in the same ratio as the *areas* __diameters__ of the exit pupil to the entrance pupil.

So, if I can handle a 22mm exit pupil, but I only deliver an 11mm exit pupil, I use one-fourth the available area. The effective f/ratio of the system is ~~four~~ two times that of the lens you put on the NV device.

This is all relative so far. To get the actual effective f/ratio, you need to know the actual f/ratio of the lens you put on the NV device. **If **we assume the same entrance pupil, then a 12mm focal length lens is 2.25x times faster than a 27 mm lens.

The f/ratio of a 22mm entrance pupil with a 12mm focal length is f/0.54.

If that sounds absurd, it's becuase it is. Looking at the specs of the lens you linked to, it has f/ratios between f/1.4 and f/16, because of the zoom. The fastest actual effective f/ratio of a system with this lens can NEVER go below f/1.4. If you try to go faster, the exit pupil just won't fit into the entrance pupil, and you reduce the effective aperture used by just enough to reach the lens' f/ratio.

Just like with any ordinary eyepiece, the magnification presented to the NV device is still: scope focal length / eyepiece focal length. Just like with any ordinary eyepiece, the image scale goes up by a factor of 2.25 (27 / 12) while decreasing total light by 2.25^2 ~ 5x.

**Edited by ButterFly, 23 March 2021 - 05:03 PM.**