85mm f1.2 on Medium Format

[MarkoRepse]

Apologies to anyone if this has already been discussed

Pham is correct about the effective fstop. Since I cannot focus at infinity it is as if I am using a thin extension tube on its native camera (but with a bigger sensor). Moving the lens further away reduces the effective aperture. However, to get to infinity requires moving the lens closer by a few millimeters at most. Unfortunately this is not possible. But in practice this amount is negligible and I very much doubt anyone would notice any difference. In any case, I am superbly happy with the results. I do not see how theoretical perfection would improve my photography.

Marko,
One question I have is why you had to hack the lens and the SLR camera to get them to working? Why did you not consider the other possibilities that does not include a mirror? Now you know that the mirror isn't all that useful.

Vivek, maybe I overstated the difficulty of focusing. It is indeed difficult to focus at 6m, however the same can be said for any lens. I will probably get a magnifier as it should help. Closer in and with good light its not a problem. I actually shot the whole 'city' series handheld. Practically impossible with the Hartblei.

Sergei, if I understand you correctly, unfortunately I do not think that would work. Because if I move the lens by 2-3mm further away than it is currently, the max focusing distance drops to maybe 1m. I had it like this for a while in an earlier stage and it was not very useful. If you succeed with your petzvals, do let us know

 

[Shashin]

Apologies to anyone if this has already been discussed

Pham is correct about the effective fstop. Since I cannot focus at infinity it is as if I am using a thin extension tube on its native camera (but with a bigger sensor). Moving the lens further away reduces the effective aperture. However, to get to infinity requires moving the lens closer by a few millimeters at most. Unfortunately this is not possible. But in practice this amount is negligible and I very much doubt anyone would notice any difference. In any case, I am superbly happy with the results. I do not see how theoretical perfection would improve my photography.

Actually, focusing at 6m would be the same whether the lens could reach infinity or not, so it is really not saying much as most lenses can focus at 6m and would have the same relative change. The increase of the effective aperture at the distances you work at are insignificant at the object distance you are working at--that is true for all lenses. Do you actually compensate for exposure when focusing to say 3m?

But it is a wonderful camera for what it does.

 

[SergeiR]

Sergei, if I understand you correctly, unfortunately I do not think that would work. Because if I move the lens by 2-3mm further away than it is currently, the max focusing distance drops to maybe 1m. I had it like this for a while in an earlier stage and it was not very useful. If you succeed with your petzvals, do let us know

Yeah. its kinda tricky. As with petzvals - its just easier for me to get it all on 4x5 No real reason to play with 645 body (although - i do have one to spare )

 

[PSon]

It would not make any difference, the 70mm entrance pupil will always result in a shallower DoF compared to the others--image circle has nothing to do with it. Larger sensors increase DoF with a given focal length and so that is not going to help with the other lenses. You would need to put the other lenses on a much smaller format in order to to beat the 85mm. And being able to focus further away will not decrease DoF.

No, the numbers do not lie in this case, the 85mm f/1.2 is going to have shallower DoF as the entrance pupil is nearly twice the diameter. This is not a surprising result. Especially since this lens is on a MF camera. The other lenses will simply not compete here.

There are several things need to be defined here:
1) Depth of field is not totally dependent solely on entrance pupil. Take a couple examples (more than two), a) it depends how well the lens is corrected for colors and b) the in general max-center resolution fall off from the center of the image circle. These are two examples of what other elements could effect the depth of field. Thus the entrance pupil alone cannot determine the total outcome of depth of field.

2) Since the ultimate goal is to achieve shallow depth of field, the focal length can be varied to the availability of lens with speed and size of image circle. For example the Kodak Aero 178mm f2.5 lens on the Crown or Speed Graphic 4x5 will result in thin depth of field.
http://marktucker.wordpress.com/2011/01/21/homemade/

3) By putting the lens onto smaller format does not necessary translate to thinner depth of field since the crop factor (using primarily the sharpest part of the lens image circle) to increase the focal length versus the native fall off of a full image circle of the lens. This is why folks prefer bigger full frame looks. An example is the Zeiss Sonnar 1.8/135 on the Sony A77 is about 202mm f/1.8. However the depth of field is greater than a Canon 1.8/200 EF lens on a full frame 5D. Numbers has to be adjust accordingly in the function.

 

[Graham Mitchell]

I would love something like this. I wonder about the focusing though as I know how hard it is just focusing the 110mm f2. Would be ideal if this could be done on a camera with reliable focus confirmation.

 

[Shashin]

There are several things need to be defined here:
1) Depth of field is not totally dependent solely on entrance pupil. Take two examples (more than two) it depends how well the lens is corrected for colors and the in general max-center resolution fall off from the center of the image circle. These are two examples of what other elements could effect the depth of field. Thus the entrance pupil alone cannot determine the total outcome of depth of field.

Lens aberrations have nothing to do with DoF. Basically, DoF can be defined with two criteria, the angular size of the entrance pupil and the permissible circle of confusion. The first is how large the angular size of the entrance pupil would be from the subject which defines how fast things fall out of focus. The permissible circle of confusion defines sharpness which is solely a factor of format size.

You can also define DoF based on object distance, aperture, and focal length, but those are really two sides of the same coin. You always need circle of confusion.

2) Since the ultimate goal is to achieve shallow depth of field, the focal length can be varied to the availability of lens with speed and size of image circle. For example the Kodak Aero 178mm f2.5 lens on the Crown or Speed Graphic 4x5 will result in thin depth of field.
Homemade Cameras, (carrying the torch). « Mark Tucker : Journal

No one is doubting that lenses can have a narrow DoF. In fact, the AeroEktar has the same size entrance pupil as the 85mm f/1.2. The DoF will not be identical however because of the difference in format and therefore the permissible circle of confusion.

The image circle per se has nothing to do with DoF. A 178mm f/2.5 lens with a larger image circle would have the same DoF on a 4x5 camera as the Kodak lens. And the Ektar on a 35mm camera would have even less DoF. The factor is the permissible circle of confusion for the formats.

3) By putting the lens onto smaller format does not necessary translate to thinner depth of field since the crop factor (using primarily the sharpest part of the lens image circle) to increase the focal length versus the native fall off of a full image circle of the lens. This is why folks prefer bigger full frame looks. An example is the Zeiss Sonnar 1.8/135 on the Sony A77 is about 202mm f/1.8. However the depth of field is greater than a Canon 1.8/200 EF lens on a full frame 5D. Numbers has to be adjust accordingly in the function.

What I am saying is if you put a specific lens with a specific focal length on two cameras with different format sizes, the camera with the smaller format will produce less DoF with that lens at a given aperture.

Your example is showing changes to two parameters at the same time. All you have said is a 135mm f/1.8 lens (with a 75mm entrance pupil) on a APS camera has a greater DoF that a 200mm f1.8 lens (with a 111mm entrance pupil) on a 35mm camera. As would would expect from a lens with a smaller entrance pupil and less magnification. So while the DoF changes proportionally with the format crop factor (permissible circle of confusion), the DoF also changes with the reciprocal of the focal length squared. So in this case, the longer focal length lens on the 35mm has less DoF--it also has a greater entrance pupil.

 

[Shashin]

Pham, here is an online pdf by Zeiss about Bokeh, but also covers DoF very well.

http://www.zeiss.com/c12567a8003b8b6f/embedtitelintern/cln_35_bokeh_en/$file/cln35_bokeh_en.pdf

 

[SergeiR]

*shudders* can we please not go into that whole discussion on how and what affects DOF

It was such a nice DIY thread..

 

[Vivek]

Sergei, One of these days I am going to put my 98mm f/1 Wild Heerbrugg on MFDB to make photos.

 

[SergeiR]

Got to admit - never heard of it before.. (thus - googled) Wild Heerbrugg Falconar 1.4/98mm ?
That would be interesting to see, as it should be just past "normal" i.e short tele- might be perfect for portraits if it covers enough

 

[Shashin]

I think the Falconer lens is corrected for particular parts of he spectrum and may not be great with white light--lots of CA. I think on the aerial camera, the lens was shimmed for certain wavelengths like red, yellow, blue, etc. But I have never seen one in the flesh.

 

[David Klepacki]

Lens aberrations have nothing to do with DoF. Basically, DoF can be defined with two criteria, the angular size of the entrance pupil and the permissible circle of confusion. The first is how large the angular size of the entrance pupil would be from the subject which defines how fast things fall out of focus. The permissible circle of confusion defines sharpness which is solely a factor of format size.

You can also define DoF based on object distance, aperture, and focal length, but those are really two sides of the same coin. You always need circle of confusion.

What Son is pointing out here is that the permissible circle of confusion is indeed influenced by actual lens designs that take certain corrections into account in order to achieve higher contrast and resolution. If this were not the case, then there would be no difference between a Zeiss lens versus a Leica lens versus a Canon lens or any other of the same focal length even if they had similar entrance pupils. However, we all know that this is far from true. Sharper lenses allow smaller circles of confusion and we can see this difference.

Basically, the MTF/OTF of a lens does affect the DOF. For example, a 85mm lens with almost perfect MTF characteristics (100% transmission) cannot have the same DOF as another 85mm lens with extremely poor MTF, even with a common entrance pupil. In fact, in the extreme limit of a hypothetical lens that is so terrible that it has 0% MTF, such a lens cannot possibly have any DOF.

Son merely points out here that field curvature and this variation of MTF from center frame to edge will also vary among lens designs of a given focal length. A lens with very quick fall-off from center will exhibit shallower depth of field compared to a lens of the same focal length (and entrance pupil) that is "flat" and does not have such fall-off.

 

[Vivek]

Got to admit - never heard of it before.. (thus - googled) Wild Heerbrugg Falconar 1.4/98mm ?
That would be interesting to see, as it should be just past "normal" i.e short tele- might be perfect for portraits if it covers enough

Not that one. A 9.8cm f/1 Reconar (really f/1.0!).

See: http://forum.getdpi.com/forum/sony/35238-really-fast-lens-nex-7-a.html

 

[Shashin]

David, which of these lenses under discussion would have such horrible MTF where DoF is impacted:

Mamiya 80 f/1.9 or Hassie/Rollei 110/2 or Rollei Schneider 80/2.

I still say that none of those lenses will produce less DoF than the 85 f/1.2 simply because of the difference in entrance pupil. That is the discussion.

Also, the ability to produce smaller circles of confusion does not change DoF--the permissible circle of confusion is just the value used to calculate DoF and can be set as you like. Lens aberrations are neither here nor there in this discussion.

 

[Stefan Steib]

I will be on the Canon CPS Roadshow 2012 starting next tuesday in Berlin.
I will have an Aptus II 12 from Yair for the show to put on the HCam and I will take a look about the Canon 1,2 85mm, I think I already did this before, but the Image circle was not big enough. What I have used a lot was my Contax Planar 1,4/85mm (also the new IR version) and this one filled the large chip easily.
And- we have a Handgrip for the HCam-B1, it´s made by Novoflex and especially the B1i or B1v are certainly usable with an additional finder.

Regards
Stefan

 

[David Klepacki]

Shashin,

You are not making any logical sense here. Previously, in your post #46 of this thread you say,

Lens aberrations have nothing to do with DoF. Basically, DoF can be defined with two criteria, the angular size of the entrance pupil and the permissible circle of confusion.

Then, in your post #54 of this thread you say,

Also, the ability to produce smaller circles of confusion does not change DoF--the permissible circle of confusion is just the value used to calculate DoF and can be set as you like. Lens aberrations are neither here nor there in this discussion.

This is a logical contradiction of your previous post.

I (and most photographers) maintain that circles of confusion have EVERYTHING to do with DoF.

 

[Shashin]

Shashin,

You are not making any logical sense here. Previously, in your post #46 of this thread you say,



Then, in your post #54 of this thread you say,



This is a logical contradiction of your previous post.

I (and most photographers) maintain that circles of confusion have EVERYTHING to do with DoF.

David, I don't know why saying aberrations are not a factor in DoF and then saying that have nothing to do with DoF is a contradiction.

Also please read what I write rather than imagining what I write. I state rather clearly that DoF is dependent on circles of confusion.

Before I leave this conversation, the term "permissible circle of confusion" indicates the maximum spot size that is considered to be the limit of sharpness. This is a subjective numbers and manufacturers can use any permissible cycle of confusion they believe would give them the best description of DoF. And individual photographer can do the same. But when we look and an image and perceive DoF, we will set a natural limit based on vision. What does it matter how small the permissible circle of confusion is if it is smaller under the actual permissible circle of confusion? (Rhetorical question, BTW.)

I posted a link to a Zeiss article. Please read that. It will help you understand this problem. I have already broken Sergie's command to stop talking about DoF. His wish is my command.

 

[David Klepacki]

David, I don't know why saying aberrations are not a factor in DoF and then saying that have nothing to do with DoF is a contradiction.

Also please read what I write rather than imagining what I write. I state rather clearly that DoF is dependent on circles of confusion.

In your previous post you also say that "circles of confusion does not change DoF". So, it makes no logical sense to say that and also maintain that "DoF is dependent on circles of confusion". This is a contradiction. Either circles of confusion change DoF or they don't.

Lens aberrations do contribute to blur, which directly relates to circle of confusion. Again, a very soft lens will limit its depth of field compared to a very sharp lens. And, you cannot subjectively change the circle of confusion to make a soft lens any sharper. This is ultimately limited to how well the aberrations are controlled by the lens itself.

 

[Shashin]

In your previous post you also say that "circles of confusion does not change DoF". So, it makes no logical sense to say that and also maintain that "DoF is dependent on circles of confusion". This is a contradiction. Either circles of confusion change DoF or they don't.

I am using the comparative "smaller." If the permissible CoC is X, than a CoC of X/2 is not going to change DoF.

 

[MarkoRepse]

I will be on the Canon CPS Roadshow 2012 starting next tuesday in Berlin.
I will have an Aptus II 12 from Yair for the show to put on the HCam and I will take a look about the Canon 1,2 85mm, I think I already did this before, but the Image circle was not big enough. What I have used a lot was my Contax Planar 1,4/85mm (also the new IR version) and this one filled the large chip easily.
And- we have a Handgrip for the HCam-B1, it´s made by Novoflex and especially the B1i or B1v are certainly usable with an additional finder.

Regards
Stefan

Stefan, do you have any samples of the 85/1.4 shot at 1.4 you could share? I'm curious how it would compare to the 1.2 version. And the Canon as well when you have them.

As for the DOF discussion, I prefer not to debate this, but my view is that the confusion comes from two ways you can look at DOF. DOF is caused by the bending of light. The first I will call absolute DOF, which is the result of just the bending of light and is a theoretical concept. In practice this determines how much the OOF areas blur. Then there is which I will call rendered DOF, and here we need to take into consideration also the system resolution, lens resolution, etc., as it essentially tells us how big the CoC is. Note that absolute DOF has a CoC of zero. So a better lens/sensor will let us "zoom in" more. This means that two different lenses with the same focal length and aperture will have identical absolute DOF, their rendered DOF can be different. I hope this has been clear enough.

To finish on a more interesting note, I also added two photos from the 'city' series

Square

Street