Another budget mirror lens, 900/8 this time

3dpan said:

ProfHankD said:

As I said earlier, I think focus touchiness is one of the biggest causes of unsharp mirror lens images -- and something like the TechArt Pro LM-EA7 is the best answer I've found. ...

Click to expand...

Unfortunately for me TechArt don't seem to include m4/3 in their product range.

Click to expand...

They never will. The two extra mm flange distance is enough to mean an M adapter probably wouldn't get more than 2mm motion, which isn't enough for most M lenses to autofocus for real. Same problem with Canon RF and Leica L mounts (although they are wide enough to make a really convoluted answer possibly work).

Basically, this tech can only work for Nikon Z, Sony E/FE, Canon EF-M, and Fuji X.

Of course, if you give up on M mount, you can make it work with even larger focus throws on all those mounts. However, people wanting autofocus of M-mount lenses seems to be a large fraction of the willing-to-pay-for-this market.

3dpan said:

Otherwise a superb idea.

Click to expand...

Before TechArt came out with the LM-EA7, I was working on a similar kind of adapter (which I decided not to market when Sony refused my formal request for lens protocol details). The design I was working on wouldn't have targeted M and would have had a much larger focus throw. We'll see. Now that there are other mirrorless mounts that may be easier to reverse-engineer the protocols for, perhaps I should re-think this...?

ProfHankD said:

Before TechArt came out with the LM-EA7, I was working on a similar kind of adapter (which I decided not to market when Sony refused my formal request for lens protocol details). The design I was working on wouldn't have targeted M and would have had a much larger focus throw. We'll see. Now that there are other mirrorless mounts that may be easier to reverse-engineer the protocols for, perhaps I should re-think this...?

Click to expand...

Sounds like a great idea.
AF is an increasing "necessity", with the interest in birding for example.

And it would give new life to all the MF mirror lenses out there, because there doesn't seem to be any chance of AF mirror lenses being produced.

3dpan said:

ProfHankD said:

Before TechArt came out with the LM-EA7, I was working on a similar kind of adapter (which I decided not to market when Sony refused my formal request for lens protocol details). The design I was working on wouldn't have targeted M and would have had a much larger focus throw. We'll see. Now that there are other mirrorless mounts that may be easier to reverse-engineer the protocols for, perhaps I should re-think this...?

Click to expand...

Sounds like a great idea.
AF is an increasing "necessity", with the interest in birding for example.

And it would give new life to all the MF mirror lenses out there, because there doesn't seem to be any chance of AF mirror lenses being produced.

Click to expand...

Before you get your hopes too high, I should note that mirror lens doughnut bokeh come from a doughnut-shaped OOF PSF, which is not what PDAF algorithms are designed to work with. The key would be to have it use both PDAF and CDAF, which most mirrorless cameras can do, but often not in their fastest focus mode -- and corresponding focus protocol. Yes, focus protocol depends on focus algorithm, at least in Sonys. This is because different focus modes have different expectations of lens behavior. For example, can you assume that the lens is capable of returning precisely to a focus distance it was at before or do you have to overshoot and come back to get backlash compensated or do you have to servo around to find the exact same position?

Alan WF said:

Alan WF said:

Surely shorter focal length Mirror Lenses will still give advantage and it is easier to make them somewhat faster if the focal length is not so extreme.

Initial fancy makes one wonder about a f4.0 100mm mirror lens pancake.

Click to expand...

Mirror lenses are essentially Cassegrain telescopes, and that’s my day job. I’d note that building a fast[*], wide-field[**] Cassegrain telescope is challenging. I’d also note that the secondary obscuration typically gets larger at faster focal ratios, and this tends to reduce the T-stop and contrast. So, no, I don’t think you’re going to see this.

Regards,

Alan

[*] f/4 is fast for a telescope.

[**] Wide-field here means the size of the image circle compared to the diameter of the aperture, rather than the size of the image circle compared to the focal length.

Click to expand...

Perhaps a redesign of the American Optical 81 mm f/0.384 Solid Schmidt would be a good idea? The linked datasheet states that the measured sample was 3.259 inch focal length (82.78 mm. Do people really need such a fast lens, or would they actually settle for a 90mm or 105 mm f/1.0, with variable aperture, instead? it won't stop people wanting one and anything faster than f/0.5 is bound to result in increased bragging rights!

Why stop there, how about a f/0.25, for micro Four-thirds, ideally for less than £399? :-D

https://archive.org/details/USAF_lens_datasheets/01-Section-1/page/n23/mode/2up

Tom Caldwell said:

A longer throw would mean a tiny bit more room at the infinity end and lost more twisting and turning everywhere else.

Click to expand...

I've seen non-linear focus throw. Make the movement quicker at closer focus, and make it progressively slower as you near infinity.

But it seems to me that focus is more critical at closer focusing distances.

ProfHankD said:

Jan Steinman said:

Alan WF said:

ProfHankD said:

I don't understand is why we haven't seen mirror lenses using the kind of offset mirrors you see in binoculars…

Click to expand...

You mean taking a mirror lens, stopping it down with an off-axis circular aperture between the edge of the primary and the edge of the secondary, and then “throwing away” all of the optics that are not illuminated?

Click to expand...

Don't want to speak for Hank, but I think he meant using prisms, like binoculars do.

Click to expand...

Please do speak on this...

Click to expand...

I imagine there are other problems that would then have to be solved, in order to get the desired resolution.

I'm not an optical engineer. I just know that monoculars with prism exist; I don't know why such designs haven't or can't be used for high-precision optics, like photography.

Given you can get a pretty damn fine set of binoculars for ~$500, I'm guessing economics isn't the constraint.

If you're handy, you can use what astronomers call "eyepiece projection" focusing. Simply cobble up a mount of some kind for a monocular.

I know in astronomy, that technique yields higher magnification, but lower resolution — the age-old trade-off of optical design.

3dpan said:

So, my question, why not add a helicoid focuser instead of the normal lens adapter, and use that to fine tune the focus ?

Click to expand...

Go for it! Shouldn't be too hard to cobble together.

My biggest concern is the wobble that is introduced whenever you put a new bit in the stack. This ridiculous experiment resulted in several inches of wobble at the end!


Story here.

--
Jan Steinman

ProfHankD said:

Focus throw on that autofocus adapter is nowhere near enough to autofocus most mirror lenses, but it's perfect for automatically tweaking an approximate manual focus.

Click to expand...

I've often wondered why the superb sensor-shifting engineers at Olympus/OMDS didn't use their mastery to provide some amount of focus tuning. Contax once had a mechanical camera that "auto focused" any manual lens, by moving the film plane back and forth!

But then I realized: that doesn't help sell new lenses at all!

Trebor1 said:

Perhaps a redesign of the American Optical 81 mm f/0.384 Solid Schmidt would be a good idea?

https://archive.org/details/USAF_lens_datasheets/01-Section-1/page/n23/mode/2up

Click to expand...

Hey, I just read on this website that they've retired something similar from U-2 service!

Are you close to Palmdale? Does the Skunkworks have a "surplus store?" That's how I got a lot of electronic goodies when I worked at Tektronix…

Having seen the camera being pulled from the equipment bay of a U-2, using four G.I.s and a hydraulic lift, I'm not sure that would fit in with Maitani's vision of "portable."

--
Jan Steinman

This thread went in my favourite direction...

Note: Alan has more experience here than I do.

Alan's mentioned a lot about tolerances, and doublet with folded flats, I will focus on tilted-component (TCT), off-axis telescope with a 'powered' primary mirror (has optical power as key part of system).

But in terms of tolerances of off-axis designs with good alignment of additional optics- you would be surprised:

Chiefspeigler | Optical Ed's (opticaleds.com)

8 inch F/7 CHIEF group build - ATM, Optics and DIY Forum - Cloudy Nights

The main thing I have not seen mentioned yet is that there are design difficulties unique to TCT designs, compared to on-axis ones (e.g. conventional mirror lenses), that historically have led them to "to small relative apertures and smallish aperture diameters":

With coma and astigmatism being in different proportions throughout the field of the two mirrors said:

https://i.postimg.cc/L81VRjdp/image.png[/IMG]

You can't look through it, but as a lens system, it is extremely simple. And you just couldn't do this with a tilted primary (it's never been done) - so I guess I'm right?

There's also making mirror surfaces concentric ('sharing same center') to a certain point, such as the center of curvature of primary mirror, which allows some sort of aberration correction.

it is important to remember these are very well byproducts, still achievable with tilted components but a lot harder.

.............

.............

.............

Now all the factors I described show historically why TCT designs like the Schiefspeigler have been limited to f/25 or f/27. But that doesn't mean the design is intrinsically limited- I can't find it now but Chief can be made down to f/5.5 without extra lenses. That's pretty fast.

ProfHankD said:

Trebor1 said:

Alan WF, post: 66362169, member: 1700025"]

Surely shorter focal length Mirror Lenses will still give advantage and it is easier to make them somewhat faster if the focal length is not so extreme.

Initial fancy makes one wonder about a f4.0 100mm mirror lens pancake.

Click to expand...

Mirror lenses are essentially Cassegrain telescopes, and that’s my day job. I’d note that building a fast[*], wide-field[**] Cassegrain telescope is challenging. I’d also note that the secondary obscuration typically gets larger at faster focal ratios, and this tends to reduce the T-stop and contrast. So, no, I don’t think you’re going to see this.

Click to expand...

Alan,

I get what you're saying there, but what I don't understand is why we haven't seen mirror lenses using the kind of offset mirrors you see in binoculars -- that way, you wouldn't have the center obscured and could even have a variable aperture.

I have experimented with using an offset Waterhouse stop on a conventional mirror lens to get around the doughnut bokeh, and it works, but is of course rather dark because the biggest aperture you can fit there is usually several stops down from the lens wide open.

Click to expand...

To expand your mind a bit, you can make a slightly obstructed design- just at the edge of the field... abit like those super-fast lenses with 'clipped' rear elements, or the Canon EF 50mm f/1.0 having bokeh obstructed by mirror box (I'm guessing).

https://www.cloudynights.com/topic/...gn-this-is-a-no-discussion-thread/?p=11911654

Click to expand...

Alan WF said:

So, in summary:

- One can fold a simple doublet or a field-flattened doublet for the cost of two mirrors/prisms. However, simple doublets are the cheapest telephotos, so probably the additional cost of the folding elements makes this unattractive.

Click to expand...

Fair enough. I think there are lots of people who actually like the fact that their 400mm lens is at least 400mm long; it's that "bigger is better" notion again. Not clear they'd pay more for a smaller lens with comparable optical performance. Then again, there are a lot of 300mm f/6.3 APS-C mirror lenses now selling new for $100-$200, so maybe it doesn't need to be a huge price premium?

Alan WF said:

- For more complex designs, in addition one needs to design optomechanics to maintain alignment of the elements in the different sections of the folded path,. I strongly suspect that this additional complication makes such designs uncompetitive with conventional telephoto designs.

Click to expand...

Certainly, precisely moving elements that are not within a concentric set of barrels is a huge pain.

Alan WF said:

Let me give you an example from my field. When my engineers can place optics in a cylindrical barrel, they can achieve 10-15 micron alignment between the elements. When they cannot, they typically achieve 50-100 micron alignment. And our optomechanics are hand-built, we do metrology and iterate the pieces bases on that metrology, and they are then coddled and (hopefully!) not subjected to the rough treatment served to photographic lenses.

Click to expand...

10 microns is pretty impressive, but 100 microns is just 0.1mm -- a tolerance I can hold with 3D-printed parts on a well-tuned printer. I suppose as focal lengths get longer, alignment accuracy gets touchier, and there's also potential for problems from thermal expansion/contraction. However, on the small scale of lenses (vs telescopes), I really don't think 0.1mm tolerances should be expensive to obtain... I just don't know if that's good enough, nor if allowing focus "past infinity" would be enough to compensate for thermal changes in alignment/distances.

I suppose it's also worth noting some designs, which I can't find right now, can weather quite high tolerance errors. Maybe not off-axis ones though...

Also- the Makowsky design doesn't count as one of these TCTs as it was merely a cut-out section of an admittedly very fast Cassegrain- obviously making it expensive to produce:



Now a TCT such as the Schiefspeigler may be thought of

Eggplantt said:

as an off axis segment of a Cassegrain, but rather than having the fast primary (typically f/2 to f/4) and high secondary magnification (typically 3 to 4 times) of the normal Cassegrain, the Schiefspiegler has a long focal ratio for the primary (typically around f/15) and a smaller secondary magnification (typically 1.6x).

Click to expand...

Mark T. VandeWettering: The Kutter Schiefspiegler (archive.org)

Jan Steinman said:

Tom Caldwell said:

A longer throw would mean a tiny bit more room at the infinity end and lost more twisting and turning everywhere else.

Click to expand...

I've seen non-linear focus throw. Make the movement quicker at closer focus, and make it progressively slower as you near infinity.

Click to expand...

See my post elsewhere - most of us see Mirror Lenses as cheap telephoto thrills that are not much good and therefore only useful if they are cheap.

Being made cheaply they gat a bad reputation and so the song goes on ....

Even good ones can have some issues because of the type of construction so it is not just a case of throwing huge wads of money at them to make one that was perfect in every way.

Therefore making one with a non-linear throw might be an added expense without an improved recognition from the price-focused buyers in the market.

Jan Steinman said:

But it seems to me that focus is more critical at closer focusing distances.

Click to expand...

I think that I am with you there 100% - I have found mine much better at closer differences where their natural low contrast is not hampered by much atmosphere in between.

Furthermore the distance scale gets much further apart at closer subjects allowing more movement and precision in the focus ring naturally.

I got my cheap 'no brand' (probably Kenko) 500/6.3 out and give it a try again - I had retired it as terminally hopeless. Re-visiting it has left me thinking that it might be the worst new lens I have ever bought. I had to put my Russian Maksutov on to re-prove that Mirror Lenses can be useful. It was a very dull day so the light and weather were not on the Mirror Lenses favourite playbook anyway.

Jan Steinman said:

3dpan said:

So, my question, why not add a helicoid focuser instead of the normal lens adapter, and use that to fine tune the focus ?

Click to expand...

Go for it! Shouldn't be too hard to cobble together.

My biggest concern is the wobble that is introduced whenever you put a new bit in the stack. This ridiculous experiment resulted in several inches of wobble at the end!


Story here.

Click to expand...

That's an interesting extreme, but always worth keeping in mind.

I have been searching eBay for "helicoid focusers" specifically for m4/3, and there are a lot of answers.

However, the most common version listed is effectively a helicoid-adjustable macro spacer, from one lens mount to another.

Just for example.
There is a helicoid focuser, C/Y to m4/3, with an adjustment range of 26.1mm to 40.6mm.
However, the standard lens adapter (spacer) for C/Y to m4/3 is already 26.1mm, meaning there is no "negative" focus range with the helicoid focuser.

The same exists for,
OM to m4/3 -- focus range = 25.9 - 40.4mm (lens adapter = 26.4mm)
EOS to m4/3 -- " " = 25 - 37mm ( " " = 24.5mm)
etc, etc.

Tom Caldwell said:

I got my cheap 'no brand' (probably Kenko) 500/6.3 out and give it a try again… Re-visiting it has left me thinking that it might be the worst new lens I have ever bought.

Click to expand...

Uhm, that would be the object of this thread (the 900/8), in my case. That's ~$200 I'll never get back and never make use of.

Tom Caldwell said:

I had to put my Russian Maksutov on to re-prove that Mirror Lenses can be useful.

Click to expand...

Hope you get a chance to try out the OM Zuiko some day. Beautiful results in a package that is small and light enough to "live" in the camera bag!

It's a "opportunity" lens — if I know I'm going to need super-tele, I have a number of specialists to call upon, but if I come across a super-tele shot that I wasn't expecting, the OM Zuiko 500/8 (usually with a Viltrox focal reducer) is there, ready!

I don't know how many times it's saved me from the "damn, I wish I'd brought the" regrets.

3dpan said:

However, the most common version listed is effectively a helicoid-adjustable macro spacer, from one lens mount to another.

Click to expand...

Yes, I've seen those… and have a few in my junk box that I haven't really put to use, yet.

Jan Steinman said:

Tom Caldwell said:

I got my cheap 'no brand' (probably Kenko) 500/6.3 out and give it a try again… Re-visiting it has left me thinking that it might be the worst new lens I have ever bought.

Click to expand...

Uhm, that would be the object of this thread (the 900/8), in my case. That's ~$200 I'll never get back and never make use of.

Jan Steinman said:

I had to put my Russian Maksutov on to re-prove that Mirror Lenses can be useful.

Click to expand...

Hope you get a chance to try out the OM Zuiko some day. Beautiful results in a package that is small and light enough to "live" in the camera bag!

It's a "opportunity" lens — if I know I'm going to need super-tele, I have a number of specialists to call upon, but if I come across a super-tele shot that I wasn't expecting, the OM Zuiko 500/8 (usually with a Viltrox focal reducer) is there, ready!

I don't know how many times it's saved me from the "damn, I wish I'd brought the" regrets.

Click to expand...

Yes I took the Tokina 300/6.3 (made for M4/3) to NZ and used it once for a shot that no other lens could capture. Now that is a really compact Mirror lens. There is a Samyang one of the same specifications but it is an aps-c lens remounted on M4/3.

The Tokina has some smarts and is chipped to tell the camera that it is mounted even if it is still a MF lens.

Jan Steinman said:

3dpan said:

However, the most common version listed is effectively a helicoid-adjustable macro spacer, from one lens mount to another.

Click to expand...

Yes, I've seen those… and have a few in my junk box that I haven't really put to use, yet.

Click to expand...

The most promising option looks like,

(i) a "universal" m42 to m42 focusing helicoid. They're actually available in a range of focus ranges.
plus,
(ii) add a m42 to m4/3 adapter. There's one with just 6mm spacing, made specially for helicoids.
plus,
(iii) then add a "lens adapter to m42" for the other end, making sure the total length is within the standard lens adapter length.

However, item (iii) is a problem for my C/Y Zeiss Mirotar. No such adapter exists and I'm going to have to custom-make one to suit.
Probably starting with a C/Y macro tube, removing the C/Y male bayonet and replacing it with a m42 male flange.
"piece-of-cake".

3dpan said:

So, my question, why not add a helicoid focuser instead of the normal lens adapter, and use that to fine tune the focus ?

Click to expand...

Do mirror lenses focus by changing the distance between the primary and secondary mirrors?

If so, that would explain why they are so difficult to focus. In a Cassegrain telescope, the displacement of the focal plane is a large multiple of the change in the distance between the primary and secondary. The precise factor depends on the optical prescription, but 10-20 is common in my experience. So, if you move the secondary or the primary by 10 microns, the focal plane can shift by 100-200 microns.

If so, then a helicoid focuser, which would give a 1:1 relation between lens displacement and focal plane displacement, might be a great help.

Regards,

Alan