Home

Please support mflenses.com if you need any graphic related work order it from us, click on above banner to order!

SearchSearch MemberlistMemberlist RegisterRegister ProfileProfile Log in to check your private messagesLog in to check your private messages Log inLog in

Internal bokeh modifier
View previous topic :: View next topic  


PostPosted: Sat Nov 01, 2008 11:46 pm    Post subject: Internal bokeh modifier Reply with quote

I started this new thread as the C/Y 50 thread wasn't really appropriate for this anymore... Smile Still awaiting input on how to dismantle the Contax, I thought to try the bokeh-modifier on a lens that I can dis- and re-assemble in the dark - the Nikkor 85/2.0...
The idea is to install a baffle that limits light entering the camera from the outer rims of the lens, as close to the back nodal point as possible - kind of like the Minolta STF, but with a baffle in stead of the apodisation element - which has to be designed into the optical formula. I made a test baffle out of crude cardboard, this will be replaced later by a chemically stabilized synthetic board.

The first test baffle looks like this. Standard 1mm cardboard blackened by a marker-pen:


Mounted in the lens, it looks like this:



I had already prepared some tests, I covered the head of a speedlight with some Al-foil and poked a very small hole in the foil to simulate a controlled specular highlight. The slightly cryptical texts are the aperture and back-focus ratios. 1:1 means that the focusplane was halfway to the speedlight, 2:1 means focus was at 2/3 of the distance. Total distance was roughly 3m. The results looked like this:


Not to good... This will clearly show in a normal picture. Well, I had to do some "from the hip" shooting before modifying the concept.

Note the intermodulation effect on the defocused text:


This is downright awful! Intermodulation all over the place...


But less detailed backgrounds show what I'm looking for! Nice, reasonably smooth:


Here you can see the effect on speculars, the starshape is clearly visible:


Well... back to the drawingboard... What if I use twice as many points in the starshape? Speculars will still show some of the shape, but the risk of getting intermodulation effects would be drastically reduced! Said and done. Here the baffle is munted again, this time with 24 points in the star.


Now, THAT'S more like it! Smile


In some cases, the star can still be visible, but the effect is several times less intrusive than what just the doubling of the amount of "points" in the star would suggest... Smile Just look at that buttery blend...




I tried deliberately to provoke some stars, but it almost seems easier to do this in the foreground defocus now... This was the worst I could do. Speedlight bounced at crumpled piece of metal foil (the foil used in earlier test... I shouldn't have crumpled it. It would have been nice to redo the controlled test again...)


It seems the effect was the intended one... This will definitely stay in the lens. I'll replace the cardboard with a machined plastic board when I try the C/Y 50... I think I'll increase the point-count to 32 also, this will be easier when I don't have to do the baffle by hand with a scalpel... Smile
But now it's dark outside, so some real photoshoots will have to wait until tomorrow. I'll try to finish the Mamiya 135/2.8 conversion too, then I'll have lots of new toys to play with... Smile Just keep your fingers crossed, hoping for some nice weather...!
BTW, looking at exposures, and comparing to f/4.0, it seems that I have lost about half a stop of light comparing to the unbaffled lens (at F/2.0 only of course). I don't really mind this, the 85F2 was never sharp enough for my taste at F2 anyway... It seems to be WAY sharper now, at wide open (which now is ~F/2.4??) than it did before. I never used it below F/2.8, so I just might actually have GAINED half a stop of light with this lens.... Very Happy


Last edited by the_Suede on Sun Nov 02, 2008 12:34 am; edited 3 times in total


PostPosted: Sun Nov 02, 2008 1:36 am    Post subject: Reply with quote

Many thanks for this account. I enjoyed it very much and I shall certainly come back to it. Please keep us posted.


PostPosted: Sun Nov 02, 2008 1:21 pm    Post subject: Reply with quote

WOW. Fantastic. Let me know more about this.

Rino.


PostPosted: Sun Nov 02, 2008 1:48 pm    Post subject: Reply with quote

Excellent fun experiment!
Thank you for taking the time to share all the steps with us.
I could never do this but, it is interesting to see that it is possible and the outcome is as anticipated.


PostPosted: Sun Nov 02, 2008 4:28 pm    Post subject: Reply with quote

your experiment are interesting! thanks for sharing your progress


PostPosted: Tue Nov 04, 2008 1:40 am    Post subject: Reply with quote

First - thank you all for your interest Smile Here's an update.
Yesterday we had a thick grey overcast, and I didn't even try to go outside to test some more distant subjects... Today I was lucky enough to have some late afternoon sun (that lasted about as long as the duration of a ProPhoto 1100Ws monolight driven by 4 NiMH DD cells...) to test the lens some more. The object of today's exercise was to test more distant subjects, and I think I got some more valuable input. All shots are, of course, wide open. Crops are at ~50-66% magnification from original size, I cropped to get the interesting parts to fit within the 900 pixel limit without loosing to much of the original resolution.

First - No problems. Even, smooth, and IMO good sense of depth:


Second - A little to "busy" for my personal taste...


Third - Busy, again. B/W contrast shows a lot more longitudinal CA than the unmodified lens. Sad This is evident in several more shots, but this one example is enough to see the effect.


I think that the principal point inside the lens shifts towards the front of the lens as you focus closer to infinity (the 85/2.0 AIS is a unit focusing lens). This would explain the increase in LCA, and also some of the increase in edgy-ness in the back-focus blur. The baffle is no longer at it's optimum position. One solution would be to move the star-baffle about three mm further forward along the main axis in the lens - the lens unit moves ~6mm when racking focus from infinity to .85m. Unfortunately that would place it at 0.5mm in front of the aperture blades, and this is almost impossible to realize from a mechanical point of view. Sad

To ascertain that this isn't all about me hallucinating, I did a few close-range shots too, and they look exactly like before (the speed-lit inside shots). Smooth, less LCA, and generally all-in-all much to good for lens that cost me 45€ and an hour of work.

You'll have to excuse the old lady's presentability in this shot, she's all roughed up from rolling around in heaps of leaves and chasing intruding kitties off the yard... Smile

The distance problem IS a real problem, I liked this mostly lens for it's medium- to farfocus properties before the modification, and it's sad to loose an old favorite in that department. Up close, though, it's a lot better now than before. Guess I'll have to buy another one, and leave that in it's original shape... Smile


Last edited by the_Suede on Sun Nov 09, 2008 1:10 am; edited 1 time in total


PostPosted: Fri Nov 07, 2008 7:44 pm    Post subject: Reply with quote

You are THE man !!

Thanks for having tested the "star shape" befors I did it, at least it show that the goal can be reached as Minolta did ont the STF lens in a more elegant way, but so difficult to do for hobbyists like us.

Now we have to find some less disturbing shape - I was thinking of small dots : more near the center and more spaced near the border, all the same diameter, in order to do the gaussian function. Must be done with Laser or High Pressure Water cuter Wink


PostPosted: Sat Nov 08, 2008 5:10 pm    Post subject: Reply with quote

Very cool experiment, thanks for sharing!


PostPosted: Sat Nov 08, 2008 7:16 pm    Post subject: Reply with quote

The "dots" idea seems very plausible.... I'll inquire as to how small holes the HPW cutter can manage. The beam is .2mm, but I don't know about the minimum step in the NC controller.

One thing that worries me about the "dot" solution is the diffraction effect - I want as little "spreading" of the light beam as possible... I would want many small holes to counter the shape-problem, but the smaller the holes get, the higher the diffraction effect (light beam spreading) from each hole will be. To much diffraction will only give me a lower local contrast, not a smoother OOF blur. I'll give it some thought and model the effect before trying. Smile Thanks for the suggestion.

On another subject: My experience with the longitudinal CA (in this case green-fringing in the back bokeh) is only valid for this specific lens, mind you. It all depends on the placement of the baffle and how and where the lensformula does it's CA correction. In the 85/2.0 that seems to be mainly in the back group... Sad


Last edited by the_Suede on Sun Nov 09, 2008 1:10 am; edited 1 time in total


PostPosted: Sat Nov 08, 2008 7:43 pm    Post subject: Fuji used round holes Reply with quote

Fuji used round holes for the plate in their 85/2 soft-focus lens, Check it out if you can I think there's pics/info around on the web. Basically same idea, with a series of round holes. works well I guess, it's an oldish idea, Imacon used it in the past.

-Ed


PostPosted: Sat Nov 08, 2008 8:54 pm    Post subject: Re: Fuji used round holes Reply with quote

EdSawyer wrote:
Fuji used round holes for the plate in their 85/2 soft-focus lens, Check it out if you can I think there's pics/info around on the web. Basically same idea, with a series of round holes. works well I guess, it's an oldish idea, Imacon used it in the past.

-Ed

Sounds the same as tested here but actually it isn't: those soft focus lenses are lenses with bad spherical correction and the holes select and mix some part of the poor "quality" (miscorrectedà) edge image part with the better quality central part. The effect achieved by such lenses is in no way matching the SFT (Soft Focus Transition) Minolta one's wich is a very well corrected one - IMHO and for what I've tried to understood.

About the small holes & distraction, yes I have it also in mind.

Imagon holes plates :



PostPosted: Sun Nov 09, 2008 4:48 am    Post subject: Reply with quote

I will not get into discussions regarding better/worse in lenses I have never used... Smile Some of the best photo's I've ever seen aren't really that sharp, it's usually "something" else that draws my eye and my admiration. But:
My ideal lens is a super-achromat with very low SA and coma - combined with a bokeh modifier. I'm a sucker for absolute sharpness, so aberrations that are visible on the plane of focus is an absolute no-no (in my own pictures!). This is why I like the Contax 50/1.7 so much more than the (in many cases more usable due to it's softer OOF) Nikkor 50/1.4 or the Summicron 50 that I also own and use - The sharpness on the plane of focus makes you go Shocked - omg! - everytime you've nailed a shot with the Planar.

Unfortunately perfectly corrected SA often gives "perfect" OOF discs both in front and in back of the focal plane. I use the word "perfect" as it seems to be the universally accepted expression to describe a totally uniform flat brightness OOF disc (with a very slightly blurred outline).
This definition of "perfect" does not concur with MY definition of good bokeh, I like it to be a little softer - maybe half-way between the "perfect" and a gaussian blur? You can find this behaviour in some lenses, but always at the cost of absolute sharpness and local contrast.

I am aware that absolute sharpness is a very small part of what makes a good picture, but I can't help being a perfectionist techno-geek, can I? Wink

Tomorrow I think I'll give the 85/2.0 another shot. I'll remove the front group and try the "soot" modification on the rim of the lens-element closest to the aperture.

I wish someone would buy me a Leica APO-Summi-90... I'd have no moral qualms about butchering it and STF-modifying it, as long as I don't have to PAY for it... Smile


PostPosted: Mon Nov 10, 2008 12:41 pm    Post subject: Reply with quote

This is a great thread, very nice work.

What kind of dog is that? She is adorable.

Smile


PostPosted: Mon Nov 10, 2008 1:12 pm    Post subject: Reply with quote

That's my girlfriend's dog, and she's a 13-year old Tibethan Spaniel... (the dog, not the girlfriend!!!) They were used as watch-dogs in old Tibet by both monestaries and normal village-people... Story is that they were the eyes and ears, the Tibethan Terrier was the legs, and the Tibethan Mastiff the brawn in guarding and defending the villages / monestaries high up in the mountains of Tibet... Once upon a time we actually used dogs as working animals in stead of just keeping them as friends and company Smile


PostPosted: Tue Nov 11, 2008 12:57 pm    Post subject: Re: Fuji used round holes Reply with quote

Flor27 wrote:
EdSawyer wrote:
Fuji used round holes for the plate in their 85/2 soft-focus lens, Check it out if you can I think there's pics/info around on the web. Basically same idea, with a series of round holes. works well I guess, it's an oldish idea, Imacon used it in the past.

-Ed

Sounds the same as tested here but actually it isn't: those soft focus lenses are lenses with bad spherical correction and the holes select and mix some part of the poor "quality" (miscorrectedà) edge image part with the better quality central part. The effect achieved by such lenses is in no way matching the SFT (Soft Focus Transition) Minolta one's wich is a very well corrected one - IMHO and for what I've tried to understood.

About the small holes & distraction, yes I have it also in mind.

Imagon holes plates :




If I'm not wrong, the imagon had the optic elements between the holes plate and the film; it had not optic at the front, after the "quality" is very limited.

The fujinon SF 4/85 had a design similar to a sonnar of 4 elements, with the holes near to the aperture blades. At F/ 8 and 11 it was very sharp!!

Your job is fantastic. Congratulations!!

Rino.