No oval bokeh in the 1.33x Bokeh is round.

What I can do now is to chose the best combination between the positive cylindrical diopter and the negative cylindrical squeeze to build the 1.33x adapter. Using the available glass this adapter will allow the use of lenses from 35mm to 50mm. In GH2 it will be equivalent to 70mm to 100mm in full frame terms. Maybe it can allow 85mm lenses also, need to test IQ.

The chromatic aberration is low, but it is there. I will try to cancel it using a flint glass as diopter, but I cannot improve the optical system to cancel all the corner softness because my knowledge and resources are not enough, so will try my best using the design I can do to minimize softness.

I quit developing the anamorphic adapter using acrylic because if I use a big front negative acrylic to do squeeze to allow using more wide camera lens, the corners will be more soft. The bigger the width used in the lens the bigger the softness is. I believe this kind of adapter needs aspherical lenses or a more complex design using more spherical elements and this is beyond my resources. The aspherical lenses cannot be grind, they are built molten / constructed fused.

I did the design to do a black plastic body and if I get a good IQ after the final test with the flint diopter using the best squeeze/diopter combination, I will hire my plastic partner to do it. Also I will do multicoating in the lenses.

I am not worried about this adapter become a product anymore. Now I am doing this just for fun. There are chances to get a better quality with the two new designs I said in previous post. So after these two final tests I will assemble one multicoated unit in black plastic body, and, if i realise it is good enough, I can build some units to sell (showing its features and also its limitations).

Hi @Ian_T let me clarify something. It is important to not confuse "out of focus" with "softness". In the 1.5x image, the bottle and the tree is in better focus. But if you zoom out the images in your browser and look at them from corner to corner you will realise the wall and the grass is sharper in the middle and start to become soft at the edges of frame. This is more pronounced in the 1.5x, and this behavior increases when the squeeze increases because the lens curves are stronger to get more squeeze. I could use weaker curves to do the same squeeze but the front lens distance would increase so much and vignette would be a problem. So the 1.25x will always be better than 1.33x and 1.33x will always be better than 1.5x considering the edges IQ in the designs I can build. Also if you zoom in and magnify the images you can see the chromatic aberration is bigger in the 1.5x squeeze also due to more strong curves in glass.

Sometimes the prototype assembling with adesive tape does not allow precise focus because it is difficult to put the lenses in very precise distance from each other, but it allows to understand what I can get from the final product.

What i am trying to do now with the two new designs is to improve the quality and minimize softness and chromatic aberrration in the 1.33x squeeze using different curves, flint glass and achromatic doublet. It seems these two new tests are the last ones to do, because anything further than this will be beyond my optical knowledge and manufacturing resources.

@Ralph_B What a nice website! There are moments in my life I feel sorry for having changed the option from course of physics at college to video production.

It's no worse looking than a Century Optics imo.

@apefos Just wanted to say that I am glad you are still working on it.

Set of cylindrical lenses to test the four new optical designs. two designs are 3 elements and two designs are 2 elements.

The guy who had some experience in film was saying that:

First Choice would be "rear mounted" Anamorphic elements. These are elements that can be added to the rear of the prime lens. The advantage is they don't vignette and you don't have to focus them separately. These were made for PL and Arri B mounts. They usually sell for about $2,500.00 but they will work well on zoom lenses.

Would it not be better/easier/cheaper to do a rear anamorphic adapter that would fit nikon or Canon lens with the flange distance we have with micro 4/3.

I did some calculations using the magnification optical formula and the room distance in the fd/m43 adapter is not enough to get the 1.33x squeeze.

If the rear anamorphic adapter inside the FD/m43 or inside the Nikon/M43 use all the room, it will be 15mm room distance in the FD/M43 and 19,5mm room in the Nikon/M43.

Using the magnification optical formula M = -f2 / f1 the lenses power would be so strong to get the 1.33x squeeze and I believe the IQ in corners would be very very bad, also the lenses curves would be very pronounced and maybe the room inside the adapters would be small to fit them.

It seems like to try to put New York inside Asheville.

The flint diopter helps minimize chromatic aberration in the two element design.

Waiting feedback!

@Vitaliy_Kiselev @redstan @johnnym @endotoxic @Ian_T @bannedindv @oscillian @thepalalias @anti12 @Tightroped @danyyyel @P4INKiller @Jasketti @JohnBarlow @JohnBarlow @RRRR @LPowell @Meierhans @fatpig @jamesjjs @au8ust @sicovdplas @mpgxsvcd

Please express your opinions!

@apefos I think it looks pretty good. It would be nice to see some other subjects like Ralph_B said, and to see what happens if you shoot with a more open aperture; portrait f.i. (blurry background).. shots w/o the adapter for reference would also be nice to see.

If you´ve got some other lenses to try out I would be very interested to see that as well.

There seems to be something funny going on @f16, where there is a slight streak of blur a little bit from the edges of the picture.