Reversed wide angle vs. telephoto for w/a macro relay

[Thagomizer]

Hi all;

All of the fish-eye/wideangle macro relay set-ups I've come across online (including here on the Forum) have used reversed wide angle lenses as part of the assembly. This evening, in looking at what people have done with Raynox achromats on Tamron 90mm lenses (having just bought a used 72B earlier today), my Googling came up with this rig put together by NickyBay:

https://twitter.com/singaporemacro/stat ... 31264?s=20

It uses (from camera out) a Nikon D800, three sets of extension tubes, a Tamron 90mm f2.8, a pair of Raynox 250s, and a 2.1mm CCTV lens. I'm curious as to how this works, how it is that this combination of optics and extension does the same thing (presumably?) as the reverse wide angle lens I've seen use on every other macro relay rig? Is one way better or easier than the other? His results look comparable to other fish-eye or w/a macro shots I've seen, so I'm really wondering about this hitherto unknown (to me at lesst) path to this type of result.

Any light that could be shed on this would be greatly appreciated. Thank you in advance for your attention.

Bruce

[rjlittlefield]

I can't tell for sure without having the lenses in hand, but...

My best guess is that the "2.1mm cctv lens" in front is creating a small aerial image close behind itself, and then that small image is being magnified by the other lenses and extension, acting as a relay. The overall result will have a viewpoint established by the entrance pupil of the front lens, hence a wide angle perspective, with a net magnification that is equal to the fractional magnification of the front lens as focused, multiplied by the relay magnification of the rear optics.

One oddity: I notice that the in-use photo at https://pbs.twimg.com/media/EQbFLmOU0AE ... name=large shows some more extension that is omitted from the setup shown at https://pbs.twimg.com/media/EQbK4LAUYAM ... ame=medium . I'm referring to the roughly 2 inches of straight tube that Nicky's left index finger is touching, in the in-use photo. Changing the spacing between the cctv lens and the rear optics will change the relay magnification, which allows matching the image circle of the cctv lens with the sensor size of the DSLR.

If the optics are working as I suspect, then the image in the viewfinder will be inverted, like looking through an ordinary compound microscope. That might be one of the reasons why he's setting up the shot "without looking through the viewfinder".

--Rik

[Thagomizer]

Good catch! I'm thinking that the extra extension is the Tamron's focus travel. In the set-up image, the lens is at its minimum length, focused at infinity. In use, it is focused closer.

Assuming that the extension tubes between the camera and the Tamron are automatic, then he still has aperture control and automatic diaphragm, unlike relays where reversed w/a lenses are used.

[patta]

Hello
I'm playing with a similar setup now
yes the extra spacing is the Tamron extension; and autofocus and iris should be working; this is a mayor advantage against the other relay system with inverted wide. The system is as Rik wrote, the Tamron picks the aerial image of the 2.1mm lens in front. The Raynox likely work to further shorten the focus of the Tamron. The small 2.1mm fisheye has likely an image circle of about 8mm (a little more in close focus) so you want to enlarge it by 2x-4x so to cover your sensor, and that's the job of Tamron + Raynox + spacers.
If you can get a macro with internal focusing (not like this Tamron, that extends) that will be much better as the lens can autofocus well without the front moving; the front lens is very close to the subject (1mm)

[Thagomizer]

Gil Wizen http://gilwizen.com/wide-angle-macro-1/, after using reversed w/a lenses in his relay system, found another solution which was easier to use. I'll quote a bit from his website:

In the previous post I mentioned that one can achieve a satisfying wide-angle macro effect using a relay lens system. It is probably the cheapest way to go about it too, as the different parts can be obtained in garage sales and thrift stores. However, using a relay lens has its own drawbacks, for example stepping into full manual gear territory, extreme chromatic aberration and soft focus. This is not, by all means, the end of the world, and after a period of trial and error using my relay system I managed to get some interesting results with a very unique perspective. I was happy with those photographs, for a while.

But after some time I grew tired of the cumbersome system I built. Each photo took me over 10 minutes to plan and execute, rendering most animal subjects uncooperative. I was frustrated with the lack of auto aperture control. And most annoying – my relay lens system was very long, with a tiny front element. This means that it did not let too much light enter the camera, resulting in a dark, upside down image in the viewfinder. In addition, the photos I got using this system all had soft focus and a strange halo around the subject, and I suspect this was a result of chromatic aberration, diffraction and the way I was lighting the scene.

It was back to the drawing board for me. I started to think what kind of look I wanted for these wide-angle macro shots, and then I remembered that some years ago people experimented with attaching a peephole lens to a point and shoot camera to get a fisheye effect. Incidentally, some of these combinations had wide-angle macro capabilities. Once I had an idea of what I wanted, I verified that it was indeed plausible, and went hunting for the suitable parts. The problem with peephole lenses is that they vary in image quality, and also some lens combinations “play well” together while others result in a photographic catastrophe. It took me over a year to come up with the right combination of optics to get the desired look I was after, but I think I got it now (at least until I find something that works better).

This current relay system is shorter, lighter and has auto focus and aperture control. The results are much sharper and there is no loss of detail. Almost perfect. Wait, almost?? Yes, although this lens combination perform better than others, the final result also depends on the camera settings, subject magnification and lighting conditions.

I'm wondering if this is similar to Nicky Bay's set-up, which also looks like it would have aperture control. Wizen has a photo of an earlier relay system, but doesn't include an image of his peephole lens solution, so I can't say if it's the same or similar to Bay's or not. There can only be so many ways that these things can go together and still work, right?

I'm guessing that a rig like Bay's used on APS-C rather than full frame would require less extension? I'm also wondering if two Raynox 250s would be required crop sensor coverage, or if one would do. Does anybody have thoughts or experience with this?

This solution using a short telephoto macro lens instead of a reversed w/a as the relay for the extreme wide angle CCTV front element sounds a lot more practical. I'm going to have to see if I can put together something similar.

[Thagomizer]

Hello
I'm playing with a similar setup now
yes the extra spacing is the Tamron extension; and autofocus and iris should be working; this is a mayor advantage against the other relay system with inverted wide. The system is as Rik wrote, the Tamron picks the aerial image of the 2.1mm lens in front. The Raynox likely work to further shorten the focus of the Tamron. The small 2.1mm fisheye has likely an image circle of about 8mm (a little more in close focus) so you want to enlarge it by 2x-4x so to cover your sensor, and that's the job of Tamron + Raynox + spacers.
If you can get a macro with internal focusing (not like this Tamron, that extends) that will be much better as the lens can autofocus well without the front moving; the front lens is very close to the subject (1mm)

Thanks for that. None of my macros are internal focus, so I'll be going long!

[enricosavazzi]

Years ago I did some work with a relay system, which I described in my 2011 book "Digital photography for science". While the discussion in the book contains additional details, the essential components from front to rear are:

- A short-focal-length "pinhole lens" in CS mount (I used both a 2.8 mm f/1.6 and a 6.5 mm f/1.8 - the shorter the focal length, the wider the angle of view). Note that a pinhole lens is not the same as a pinhole. A pinhole lens is a lens with a small entrance pupil (1-2 mm in diameter) located in the air well in front of the physical lens barrel. This is a wideangle hidden surveillance lens that can be mounted e.g. behind a mm-wide hole in a wall or a similar inconspicuous arrangement.
- Adapters and a focusing helicoid.
- A 25 mm f/2 Leitz Photar bellows lens (a conventional wideangle lens is generally not useful here, because its entrance pupil is too far recessed into the lens - a reversed wideangle may sometimes work, but not as versatile as a short-FL bellows lens).
- A second focusing helicoid and adapters between second lens and camera.

The first helicoid is used mainly to set up the right distance between the two lenses. The second helicoid is used mainly for choosing the desired magnification. The angle of view can reach roughly 90 degrees while still well into the macro range, depending on the ratio of focal lengths of the two lenses. Effective aperture is rather high and reasonably priced pinhole lenses are not designed for high resolution, so image resolution as a whole is less that what most forum members are used to. Since this is a compound optical system (just like a microscope), the image in the camera viewfinder is reversed.

The so-called Frazier lens works in a partly similar way. It has been in the news a few years ago because of controversies that caused its patent (US 5,727,236) to be declared unenforceable, but the principle itself does work. The claims about exceptionally high DOF that often accompany discussions about this lens are however unfounded, since DOF versus diffraction in this optical system has the same limitations as all other optical systems (I am mentioning this to avoid unnecessary repetitions and pointless discussions).

[patta]

Nice, 2011 is the earliest use of M12 lenses I've seen
did you invented yourself? or can we find someone earlier?
For the same use I'm looking for pinhole-type objectives, with forward pupil, maybe wider and with better optics. Marco Cavina described the Zeiss "Sonderobjektiv", a pinhole lens for 35mm, but that's unobtainable. I'm trying to use telescope eyepieces in that place.
The M12 pinhole with relay has been been used "professionally" for example to shoot the documentary about ants https://www.bbc.co.uk/programmes/p05rgrcb

I think that pinholes have more working distance, while the small fisheye should be better corrected

[enricosavazzi]

Nice, 2011 is the earliest use of M12 lenses I've seen
did you invented yourself? or can we find someone earlier?
For the same use I'm looking for pinhole-type objectives, with forward pupil, maybe wider and with better optics. Marco Cavina described the Zeiss "Sonderobjektiv", a pinhole lens for 35mm, but that's unobtainable. I'm trying to use telescope eyepieces in that place.
The M12 pinhole with relay has been been used "professionally" for example to shoot the documentary about ants https://www.bbc.co.uk/programmes/p05rgrcb

I think that pinholes have more working distance, while the small fisheye should be better corrected

I don't know anything about M12 lenses, but compound optical systems in wideangle macro have been used (or at least experimented with) for decades. The principle of "relay optics" is as old as compound telescopes and microscopes.

[patta]

Yes the "older setup" described by Wizen is identical to Nicky Bay's one, and is pretty common.
The new setup, he didn't put images so just my guess is that in practice is a macro with a small wide angle attachment put in front, a little bit before its focus (the door peephole lens.) It should look like more or less as the Endoscope objective, see slide 14 at Gross lectures; the objective is the group at the left (forget the big relay at right) and our Wide angle attachment (peephole, in Wizen setup) is the last negative lens on the left; https://www.iap.uni-jena.de/iapmedia/DC ... 002675.pdf

I haven't tried that setup yet. Maybe soon; wide angle attachment can be a door peephole, a phone "wide angle clip lens", a telephoto attachment mounted inverted , or a small negative meniscus, like see also Max Simpson https://www.maxamillion.online/blog/wid ... hotography
The relay as by Nicky Bay works well, look at his photos ! Apart from the poor Tamron macro that is a bit overstretched with all the spacers, and its delicate focusing front supporting the Raynox, rings and front fisheye (with plastic filter thread, I'm doing same, unsure how long will last ) . I checked his website, now he's using a Laowa 100mm 2x instead of the Tamron. Big issues still, working distance (fisheye-to-subject) are minimal to none and resolution not better than 10 microns (I couldn't find insect eyes well resolved in such photos).

I'm working hard on a similar setups to solve those issues & freely share them to you all: instead of the CCTV lens we can put a telescope eyepiece or a fast prime; less wide than a fisheye (40-100 degrees) but good working distance and no hard magnification limit, thanks the "virtual aperture" in front of the objective, as in the Pinhole setup described by Savazzi; eyepieces and primes may have better optics (and higher prices) than the pinholes. If anyone cares..
Another true ~180° fisheye setup, working up to 10x, f/2 object-side (though with short working distance), I'll keep it secret unless somebody's interested.

The best wide-angle relay macro I've found published (drawing) was from the 1971, designed by... Al Nagler!
In his Probe setup, front objective is basically an eyepiece, plus a wide-angle attachment meniscus at the tip, like the Laowa24mm probe: http://cs.astronomy.com/asy/b/astronomy ... e-vue.aspx
Today's borescopes for industrial inspection may be our lens too, but the good ones cost! I haven't bought one.

Edit: I've found just now the company INON selling both of the above wide-macro attachments (relay fisheye and peephole) ready made for underwater use; prices seem bit higher than the DIY CCTV, but reasonable, much better than the Innovision or T-Rex probes. http://www.inon.jp/products/housing/ino ... efs60.html

[dolmadis]

Another true ~180° fisheye setup, working up to 10x, f/2 object-side (though with short working distance), I'll keep it secret unless somebody's interested.

Of course we are interested !!!

And thank you for such research.

John

[patta]

Another true ~180° fisheye setup, working up to 10x, f/2 object-side (though with short working distance), I'll keep it secret unless somebody's interested.

Of course we are interested !!!

And thank you for such research.

John

Happy to hear that!
You may then take one of your microscope objectives (I know you have some, if not too many ), better one with high NA; high.
Look at it from the back, at some distance. A ready made, micro-fisheye objective will miraculously materialize in your hand!
Now it is just matter of getting this image onto a sensor, focus closer and stack.
If you happen to have a microscope with Bertrand lens or "collimation telescope" that's one easy way to go.

[CharlesT]

Hmmm,
I have just stuck a 40x (4mm N.A 0.65) lens on the 'front' of a reversed Canon 35mm efs macro and its only giving me around a 60-90 degree field of view:

Is this too low N.A lens to use?
I removed the stop at the back of the lens, but maybe the objective's tube itself could also trimming the field of view?
[The image has been un-fisheyed with a Paint.net plugin].

[patta]

Hmmm,
I have just stuck a 40x (4mm N.A 0.65) lens on the 'front' of a reversed Canon 35mm efs macro and its only giving me around a 60-90 degree field of view:

Is this too low N.A lens to use?
I removed the stop at the back of the lens, but maybe the objective's tube itself could also trimming the field of view?
[The image has been un-fisheyed with a Paint.net plugin].

Nice, yes it is how it work; and that is already "Wide Field"! You can try a 100x oil immersion too (dry).
The NA determines the Field of View; in this setup the objective works with NA and field swapped. The macro relay is focused on the aperture of the objective; the black sharp disk that limits your photo is the aperture baffle, and it is in the middle of the objective lenses, can't be changed. The dust there also, can't be removed... I tried too to remove the back baffle, but no improvement. The tube or the relay macro are not trimming (I believe).
FOV = 2* arcsen (NA)
NA 0.25 -> FOV ~30°
NA 0.65 -> FOV ~80°
NA 0.999 -> FOV ~179°
NA>1 -> FOV ~180°