ACL viewfinder - improving hand held ergonomics

[Gregg MacPherson]

The ACL I with the early periscope VFs sat more to the rear on your shoulder, better balance. The "eyepiece" on the finder was almost of zero length, the diopter adjustment being effected by shifting VF body horizontally. In contrast, the orientable Angenieux and Knoptic VFs have long eyepieces that shift the camera forward....

I'm wondering if a simple forward translation of the VF is possible. The drawings attached suggest it may be. The sideways translation of the VF to accommodate the added mirrors is shown as a guess (the red box). It's the same width as an AZ Spectrum video tap.  Of course the standard tap has to go, so maybe the hard part of this whole mod exercise will be designing the inclusion of the prism etc that a tap will need.

Unfortunately, I don't yet have a grasp on the theoretical optics, to guess on the one or two added elements that may be needed. If anyone wants to contribute, or give me a leg up on that, please do.

The mod housing may be simple, similar to the red box in the drawing, with two mirrors in it.  Easy to make. Depends on where and how the added optical element(s) are placed.

The VF mounting screws on the camera body will get more stress. Thread inserts would help. Visual Products put those in when they mod for S16. So some cameras already have them.

 

[Boris Belay]

Hi Gregg, Interesting question...  This is the issue that Aaton tried to solve with the basic design of the LTR: moving the VF forward so that the camera sits further back on the shoulder to better distribut the weight of the camera.  On the LTR, Beauviala elected to have the VF come out of the top ot the camera and forward from there (and this was the reason for the patents trial by Arri, who claimed the design was a copy of the SR VF).

Short of re-building the ACL completely to take on an LTR VF assembly, with carrying handle, while you're at it (and that may well be the easier route if you don't have access to a metal-working shop), your solution is the obvious one.  It seems theoretically sound, except that you can count on the optical elements to be at least as thick as the ones in the front assembly of your Kinoptik (or Angénieux) orientable VF.  If you look at your drawing, that formard (left-eye/right-eye orientation) part does what you need, except one of the elements should be inverted.  Since the ground glass image is translated at a constant size, you can count on the optical blocks being as thick.  This, in turn, poses the question of the position of your modified viewfinder relative to the body of the camera (and the head of the operator).

Obviously, you would also have a slight light-loss from adding optical elements in the path, but if they are well-built, it should be nominal.

[David Sekanina]

As long as the rays are still collimated in that portion of the viewfinder, so not converging or diverging, this will work. You can try this by pulling out the original viewfinder by a few centimeters while keeping the axis (making the optical path longer) and check if the image stays in focus. If it is still in focus, you're good to go ahead with a prototype. The image is mirrored twice with your add-on, so no image flip.

You can also use a prism, to ensure the angles meet the close tolerances, or two mirrors, like you said. Enhanced Aluminum or enhanced Silver mirrors have an extremely high reflectivity, the light loss will be minimal.

 

[Gregg MacPherson]

An updated drawing...(again on 5th April 2023)

Edited April 5, 2023 by Gregg MacPherson
updated dwg

[Gregg MacPherson]

16 hours ago, Boris Belay said:

....you can count on the optical elements to be at least as thick as the ones in the front assembly of your Kinoptik (or Angénieux) orientable VF.  If you look at your drawing, that formard (left-eye/right-eye orientation) part does what you need, except one of the elements should be inverted.  Since the ground glass image is translated at a constant size, you can count on the optical blocks being as thick.  This, in turn, poses the question of the position of your modified viewfinder relative to the body of the camera (and the head of the operator)....

Hey Boris,
It's been a while. There may be a couple of places to accommodate extra elements easily, but if the optics are better understood it may restrict their position. There is plenty of room inside the VF porthole on the camera body, the first obvious place. See updated drawing.

I've never looked inside the part of the finder that swings over for the "left-eye/right-eye" option (segment E in new dwg). Have you seen what's in there? I don't want to pull it apart without good reason.

Gregg.

[Gregg MacPherson]

6 hours ago, David Sekanina said:

As long as the rays are still collimated in that portion of the viewfinder, so not converging or diverging, this will work. You can try this by pulling out the original viewfinder by a few centimeters while keeping the axis (making the optical path longer) and check if the image stays in focus. If it is still in focus, you're good to go ahead with a prototype. The image is mirrored twice with your add-on, so no image flip.

You can also use a prism, to ensure the angles meet the close tolerances.............

Hello David,
If I had completed my physics/eng. degree instead of escaping to Art School I would have some better ideas to work with now.

Shifting the VF horizontally away from the camera and trying to understand the changed image was one of the first things I did. Up to a shift of about 7mm I could restore focus with the eyepiece diopter. My eyes normally need the eyepiece at +5 to +6, the video tap maybe skewing the scale a bit, and with the 7mm shift the eyepiece diopter was at -6, the crosshairs sharp. 

Shifting the VF out 20mm as in my first drawings, the crosshairs disappeared. The size of the image in this sloppy experiment didn't change much, or at all. 

I'm tempted to set up some kind of test bench and just insert a negative diopter into segment A, just inside the VF mount port.

The original, unmodified length of the optical path from the ground glass prism to (approximately) the lens of the eye is 335mm. The modified path is 375.4mm.

Gregg.

[Gregg MacPherson]

On 4/3/2023 at 9:52 AM, David Sekanina said:

As long as the rays are still collimated in that portion of the viewfinder, so not converging or diverging, this will work.....

The Kinoptik VF has a focusing lens at either end. The rays may be parallel (collimated) in between. So when the VF shifts away from the camera it makes sense that the gg will not stay in focus. My earlier drawing with the neg diopter placed just inside the VF port on the camera, I don't think will work. It's too far away from the VF on the optical path.

But there is a lot of room to place a modifier inside the inner end of the VF, which is where it probably should go. See the updated dwg above.  The ID=18mm.

In the following, and in the dwg,  I'm representing the focusing lens at either end of the VF as a simple positive diopter

The unmodified distance from the gg to the (camera side) "diopter" on the VF is 103mm. So that is approx a +10 diopter.

The modified distance, with the translated VF is 143mm. So that needs a +7 diopter. The +10 will focus too close.

This is where I need to go back and read some on basic theory. Do I simply add a -3 diopter?

Gregg.

Edited April 4, 2023 by Gregg MacPherson
more words.

[David Sekanina]

Here's a simple relay lens with two achromatic doublets. Left end pics up the image from the focusing screen, right end focuses the image in front of the eyepiece, which then picks it up and magnifies it. In such a setup, you can space the two achromatic doublets further apart, without the relayed image getting out of focus as long as the focusing screen distance stays at the focal point of the first achromatic doublet, and the eyepiece at the focal point of the second achromat.

I hoped the viewing system on your ACL was as simple as that, but by what you describe, the beams are not collimated in-between. In these more complex cases I bother an optical engineer, which I am not. Sorry I could not help ?

[Gregg MacPherson]

Thanks David,

That picture is a very useful prod. I wasn't patient enough looking for likely lens configurations within viewfinders.

My guess is that the Kinoptik VF does have a configuration as in your drawing. What may not be clear from my text and drawings is that both of the doublets are mounted within the viewfinder itself. There's no easy way to increase the distance between them without modifying the VF housings, which I want to avoid. 

So my idea was to shift the intact VF and compensate with a negative diopter at the camera end. My drawing shows that diopter as a simple single element, but there may be an axial distance of up to 16mm to place a more sophisticated lens.

I'll update my dwg with an approximation of the achromatic doublets, but I think you get me already.

Gregg.

[David Sekanina]

The two doublets in the eyepiece is the Plössl eyepiece itself probably.

 

https://milwaukeeastro.org/beginners/eyepieces.asp

Edited April 4, 2023 by David Sekanina

[Gregg MacPherson]

I updated the screenshot from the dwg again (Eclair forum moderator privilege).

For now I'm just guessing on what the elements are inside the original VF/eyepiece. My idea of shifting the VF forward requires extending the focal length on the inboard side of the VF. As a notion, I'm doing that with a single element neg. diopter. There is room axially to place a more sophisticated lens there, if that's useful and feasible.

The ID for the inboard end of the VF is 18mm. There is a thread visible further down.

The astronomy forums may be a source of ideas. A glimpse there showed activity and some knowledgeable people. Meanwhile i better read a little.. 

[Gregg MacPherson]

I started reading a little, looking for simple ways to understand the focal length relationships.

With some simplifying assumptions, like treating the doublet or triplet as a single element, then applying ideas.

I'm sure the equations when adding a diopter to  a doublet or triplet will be more complicated, but hopefully this gives me a start.

1) When you add two thin lenses of focal length f1 and f2 together (with no gap between them) then effective focal length is given by F such that:
1/F = 1/f1 + 1/f2

2) When you add two lenses with a gap d between them then effective focal length is given by:
1/F = 1/f1 + 1/f2 - d/(f1*f2)

The d/(f1*f20) is going to be extremely small, ignorable, so equn 1 then...

F=the combined focal length of the lens and added neg diopter in the inboard end of the VF = 0.143m.

f1= the focal length of the lens alone in the inboard end of the VF.

f2 = the focal length of the neg diopter in m, so 1/f2 = the diopter number.

1/0.145 - 1/0.102 = 1/f2

=> 7 - 9.8  -2.8 diopter

I assume a small reduction in magnification. I'll look at that. I would actually prefer the kinoptik VF image slightly smaller.

Edit: Magnification may be = f1/f2 =0.102/0.357...so too much. Something else is needed.

Probably more knowledge...reading.

Gregg.

Edited April 5, 2023 by Gregg MacPherson
more words.