Rear format reducing converter

[David Kilpatrick]

I am not very good at finding the right formulae to work out optical maths, but from a few experiments this morning I believe that in order to condense the image from a full-frame Alpha lens on to NEX APS-C, a converter tube roughly 11 to 12mm long (maybe as short as 10mm) containing an aspherical LD two-element positive lens group, 40mm in diameter, positioned to be as close as possible to the rear element of the Alpha lens, with a power of maybe 10 dioptres/f=100mm would perform the function.

I don't have a +10 CU lens here to test this out. I have used my 85mm Samyang manual Alpha lens to project its image circle, then placed a +4 CU lens (the most powerful I have) up to the rear element. This reduced its rather generous image circle from around 50mm to around 40mm, with a corresponding 20% or so reduction in the flange to focal plane distance. I may be able to construct an optical bench, which I do not have now (I once had one) to test this with precision.

One thing I'm sure of - it is possible. Whether a converter could be made using a simple, slim design with only two or so elements I am not so sure. New glass types, aspherical moulding and hybrid elements could surely keep aberrations to a minimum.

David

[Javelin]

I'm just going to restore the register distance of the original lenses (SR mount) 43.5mm Why would you want to have glass in the adaptor? purely for size?

[Javelin]

I see the answer in your other thread. you want to be able to convert the image circle of the lens down to APS-C to restor the lenses field of view and than can haopopen because there is no mirror. can that be done with one element? does it need to have coatings to eliminate reflections ?

[Greg Beetham]

If the flange to image plane distance was shortened by about 20% does that also mean perhaps that the 85mm lens became an approx 68mm lens with a larger maximum aperture? (with the new smaller image circle)
Greg

[peterottaway]

All this talk has me dusting off my collection of MD lenses not to mention the couple of OM wides that I has almost forgotten that I had.

None of the m 4/3 have interested me that much when looked at in the flesh and the x 2 factor just doesn't make them worthwhile.

[Javelin]

the idea of a restored FF field of view has me intrigued. I was just going to make mechanical adaptors. I have a contact at Shott I wonder if he could get the diopters made. that would be very expensive.

[David Kilpatrick]

In theory it can be done with a single meniscus element - a close-up lens stuck on the back element of the main lens. As I have indicated, it needs to be a powerful one.

I have done some experiments with a 35mm lens which fits into a teleconverter tube (an old 1960s invention - 35mm f/5.6 wide angle, extension tube and 2X tele converter all in one kit). Even with this I can get a reduced size image, but it is automatically limited to f/5.6 brightness because the 35mm element is very small. The converter element must be big enough to fill the entire converter diameter.

Greg - yes, the image scale of the 85mm was of course reduced, meaning is was being converted to something like you suggest. I need to dig out my copy of Sidney Ray's Optics textbook, and remind myself of the formula. I know it for two identical elements - if you take a 200mm focus meniscus, add another 200mm focus to form a pair, you get a 100mm symmetrical lens.

What I don't know is whether a reducing converter can be made which can have an afocal conversion factor. My proposed converter might be highly effective on a 50mm f/1.4 lens, but too strong for a 20mm f/2.8, and too weak for 200mm f/2.8. It might be matched to one lens only, unlike tele converters.

Javelin - it's easy enough to obtain dioptre lenses in any value from scientific (school and college) lab suppliers. Or any optician can get a circular dioptre lens made, in many types of glass, thicknesses and diameters.

The principal is the same as the relay lenses used in early digital SLRs from Minolta (RD-3000, RD-175), Fuji and Nikon. These relay lenses often had too much to do, far more than a mere 0.66X factor. They did seem to work with a good range of lenses. They had an unwanted effect of limiting the maximum aperture, and this I do not understand, unless it is related to the distance they were positioned away from the exit pupil of the master lens.

What I propose is really nothing more than an adaptor containing a weak relay lens assembly, positioned as close as possible to the exit pupil and with a large enough aperture to maintain or enhance the T-stop of the combination.

David

[Javelin]

well a lens the diameter of the non nex side bayonet can be made to fit in there i guess. but how close is close?

I could make a unit up in 3 peices. Nex side bayonet, a body shorter than required (for testing), a non NEX side bayoonet and finally another holder for a lens and work out the distances between them out by trial and error. Information on A mount and MD mount shows there is a possible 1/10th mm or more error in the numbers that are reported to be the flange distances. if i'm going to make this it had better be right and i'd like to nail that so the idea is to test it all and know for sure with real parts before I have 500 useless peices. My problem is optics.. and how to measure the resulting image circle. the point of sharp focus. and such. also how do I know what the tolerance is. I'm sure that there is a point where a few thou won't matter because pixels are so huge right?



.

[bakubo]

David, I am pretty sure you are already aware of this, but just in case you are not I will mention it. If you can come up with a good design but need machining and optical expertise and manufacturing ability to produce it there are people in China, I think, who can do this. I have seen photo accessories before that have been rather cheaply produced there and then sold. I don't know how to find the right people but if you did and could produce something at a reasonable selling cost then you may have found a product niche to sell to. The idea sounds cool.

[WaltKnapp]

Just a general thought, but if this sort of thing could be done on the NEX there is no reason it could not have been done on APS to bring full use of wide angle lenses. But it was not and by some pretty expert companies in optics. I expect it's much more complex than expected. Remember, for one thing that light is made of a range of wavelengths and simple lenses do tend not to respond the same to different wavelengths. By the time you get those corrections in there you are probably out of room in NEX.

Walt

[Greg Beetham]

I'm no optical expert, but I've got a suspicion that it might be ok with some primes 'as a concept' design but it might run into trouble with vari-focal zooms, my old MD70-210 f4 for example had a moving rear element group and it wasn't even vari-focal, and my AF100-300APO does have a moving rear group as well (it is vari-focal), I'm only guessing but I think you'd probably need to have a lens design with a rear element that remained in a fixed position, I think some AF primes have a moving rear group too, how many I don't know, (I just checked my MD135 f2 and it has a moving rear element group as well...moves as you turn the focus collar).
Greg

[bakubo]

I think some AF primes have a moving rear group too

I have 2 AF primes handy to check and both have rear moving elements, but neither one extends past the collar that surrounds them. Minolta 50mm f1.7 and Sigma 24mm f2.8.

[Greg Beetham]

Yeah Henry, no danger of contact if done correctly, but I've got a feeling that that moving rear element group spells doom for the experiment just the same.
Greg

[David Kilpatrick]

Just a general thought, but if this sort of thing could be done on the NEX there is no reason it could not have been done on APS to bring full use of wide angle lenses. But it was not and by some pretty expert companies in optics. I expect it's much more complex than expected. Remember, for one thing that light is made of a range of wavelengths and simple lenses do tend not to respond the same to different wavelengths. By the time you get those corrections in there you are probably out of room in NEX.

Walt

It can't be done on APS-C DSLRs because the mirror is where the converter elements need to be. With the NEX, the glass from a converter could extend as far into the body as needed - and the main camera lens can be positioned a full 25mm closer to the sensor plane, to permit the use of a positive dioptre lens group behind it.

It has never been possible because no camera has ever had an 18mm flange to sensor distance. Now that one does, it is possible. So is the use of S-mount cine lenses and Leica M or M39 lenses directly.

David

[Javelin]

making stuff in China isn't a big deal. getting what you asked for is. Your best to hire a rep there who is familiar with the type of manufacturing required to look after your interests if your not going to travel there yourself.