Retrofocus on 75 Summilux

[fgianni]

The crop may be 115, but the DOF is more like 90 mm. That is the math of different formats.....

That is exactly what I said a couple of posts ago, only that my maths leads me to 91mm for DOF instead of 90.

 

[jaapv]

Sorry, busy busy busy, didn't see your post 🙁

 

[jaap]

My guess is that it's an issue of focus shift due to the limits of the lens design. On the new 75 Summicron, Leica introduced a floating element for the first time on a telephoto lens in the M-system. If a floating element is needed on a lens at f/2, I would think that it would need it all the more on one at f/1.4.

The Summilux design is almost 25 years old. It's interesting that Leica introduced a Summicron at this focal length rather than update the Summilux. I don't know, but it makes me wonder if the combination of speed & focal length aren't just too much to overcome to make a lens that can focus accurately close up & wide open.

Huck

Yes indeed that's another factor focus shift. That is the focus shifts slightly depending the apeture. Especially with high speed lenses and longer focal lenght.
The new Zeiss IKon lenses are well corrected for this error. Especially important with RF lenses!

 

[jaap]

The crop may be 115, but the DOF is more like 90 mm. That is the math of different formats.....

Yes ,indeed thats right!

 

[fgianni]

Ok I think we agree that the issue is most likely due a limitation of the lens design, since the lens is usable anyway it's a keeper.

Thanks Jaap and Jaap, please say hello for me to your friends Jaap Jaap and Jaap 😀 😀 😀 :angel:

 

[vizioneer]

Hehe...unfortunately there is no perfect lens out there (fast tele). Unless Leica comes out with a new 75 lux APO-Asph FLE! Don't see that happening anytime soon.

 

[StuartR]

You are asking for heroics out of a small baselength rangefinder. My personal suspicion is that the other lenses you tried were not as critical in terms of focus as the 75 lux (yes, even the noctilux and nokton you tried). I am not saying Leica never fails or anything, but it is much harder to knock a lens out of collimation than it is to do it to a rangefinder. I expect that the 75 is within its tolerances, but the short baselength cannot focus it with complete reliability. Remember, it is only physical baselength that determines true accuracy, magnification helps you get the best out of the baselength you have, but it cannot create accuracy that the baselength can't provide. In doing his massive test on the RD-1, Sean Reid came to the conclusion that it was not possible to reliably focus the 75 lux wide open at close focus...so that is one expert's conclusion.

In any case, I would not necessarily recommend sending in the camera or the lens for calibration. Just don't use the 75 lux at close focus and wide open for a critical shot. Stop down to f/2 or so or back up a little bit. Worse comes to worse, it is digital so you can look at the screen and see if you nailed it, and if you didn't you can shoot it again. Either way I don't think it is a big deal, you should be fine with the 75 lux in normal use.

 

[fgianni]

Hehe...unfortunately there is no perfect lens out there (fast tele). Unless Leica comes out with a new 75 lux APO-Asph FLE! Don't see that happening anytime soon.

When you get yours let us know how it does behave at the shortest focusing distance, you might be lucky and get one spot-on at all distances.

 

[StuartR]

Yeah, just to say a bit further, I don't think it is a problem with the lens design or the len's tolerances. The tolerances of the Leica lenses are higher than the Bessa bodies. That does not mean Leica is super A#1 and Voigtlander is crap, it is just the way it is. You are paying vastly more for Leica to manufacture to higher tolerances and do more extensive QC. The RD-1 is a great camera, but its tolerances are looser and the rangefinder is a much shorter baselength than the Leicas, Konicas and Zeiss Ikons. Leica and Zeiss would not make 75/1.4's, 90/2's and 85/2's if their bodies could not reliably focus them. But at the same time, Voigtlander does not make any similar lenses precisely because THEIR bodies can't focus them reliably. There is no good or bad here, just manufacturing tolerances and optical capabilities. Anyway, that's my two cents. My personal experience with the 75 lux is that I have not had any problems focusing it reliably, even close up and wide open. I have used it on a .6 hexar RF, .72, .85 and .91 Leicas...

 

[StuartR]

And the reason that I think the other lenses did not test the VF in the same way is that the Noctilux may be an f/1 lens, but it only focuses to 1m compared to the 75 which is both longer and focuses to .7m, even if it is not an f/1 lens. So while you may not have run up against the limits of the rf mechanism before, I think you are now...

 

[fgianni]

I have used it on a .6 hexar RF, .72, .85 and .91 Leicas...

Well that does not disprove what I said, a 35mm film camera has a DOF that is 22% more than the DOF of the RD-1 with the same focal lenght (not same FOV).
Given that the focused point on the RD-1 is at the edge of the DOF (sometimes just inside, sometimes just out), it will most likely fall inside the DOF for a 35mm full frame camera, so you simply may not be able to see the issue even if it might be there.
I think that the 75lux is probably OK on any Bessa, but not on the reduced-frame RD-1

 

[StuartR]

But what makes you feel that the lens is not delivering something that is just barely outside its tolerance if at all, when the rangefinder is being used WAY past its tolerance? What I mean is that the 75 lux is aligned very very carefully at the factory to be accurate. It has a long focus throw and a stiff focus ring so that it can be moved carefully. On the other hand you are trying to focus it with a rangefinder that has a baselength of only 37mm. The lens was designed to be used on a camera with a baselength of 69.25 mm of baselength...nearly double the accuracy. The depth of field on the RD-1 is an interesting phenomenon, but it does not change the rangefinder's capabilities. If the RD-1 was designed to focus the 75 lux, it would have had a rangefinder baselength like the ZI or the M series. In any case, it just seems to me that mechanical precision in the lens would be an order of magnitude greater than focusing precision of the optical RF, which is not only more mechanically complex, prone to mechanical error (being knocked out of alignment), and being used past its intended accuracy threshhold, but also dependent on the operator's eyesight, where the lens is aligned by a machine. It just seems to me that the viewfinder has to be the weakest link here, not the lens.

 

[fgianni]

Stuart

I have to admit that you may be right, however If the resolution of the RF was being exceeded, I would expect to have sometimes back-focus, and sometimes front-focus, while I get constantly back-focus between 1/2 and 1 inch.
If I focus and then move my head backwards about 1 inch I can see the RF not being anymore spot-on, but every shot is in good focus.
Also remember tha focusinc accuracy depends on the base lenght, but also n the magnification, and the RD-1 viewfinder has a magnification of 1:1 which helps.
Again, we are here probably in the realm of speculation, and may be that the effect I see is due to a combination of the two: reduced accuracy of the RF and reduced DOF of the sensor.

 

[StuartR]

I hope I did not come off as being too fanatical on this, it is just that it seems pretty cut and dry to me. If you are constantly getting a backfocus of 1/2 to 1 inch at .7 m at 1.4 it is more likely that this same back focus is present in all other lenses as well, just that it is being covered by depth of field.

I was curious, so I went to the depth of field charts. Interesting stuff...well, not really, but it proves a point.

Depth of field for the 50/1.4 ASPH (closest to Nokton) at f/1.4 and .7m =0.015m or .6 inch
For the 50mm f/1 at 1m (its closest) there is .023m or .9 inches of depth of field.
For the 75 lux at f/1.4 and .8 m (it focuses closer but the table did not have info) there is only .01m or .39 inches of DOF.

So the noctilux actually has the most depth of field of the group, with the 75 lux having less than half an inch of DOF, probably more like a 1/3rd of an inch at closest focus. So, it makes sense to me that the DOF of the other lenses might cover up the error of the RF, where the 75 lux has about a 1/3rd of the depth of field of the Noctilux and about half the DOF of the 50/1.4. It makes more sense to me that your RF is very very slightly off and it is only showing up with the 75. But whatever, at this point you are pushing the gear way more than it wants to be pushed, and something is microscopically out somewhere along the line. It just seems to me that the most likely candidate is the RF for the reasons I stated above.

But I am burnt out! Post some 75 lux shots, it is a beauty:

 

[Hektor]

Fgianni Hi, I've been away for a few days so I'm catching up with this thread.

I doubt that your 75 is significantly in error but we could compare it with mine if you wish.

The 75 has the same DoF as the 90F2 so maybe you could try one of those.

The Noctilux at F1.0 has 50% more DoF than the 75F1.4 and the 90F2 all wide open, all focussed at the same distance.

It may be useful for you to know that the 135F2.8 has half the Dof of the 75 and is therfore even more critical of rangefinder calibration. I suggest you try one !

I sold my 135F2.8 because, having tried it, I thought it was a silly lens to put on a rangefinder, and was much easier to use on a reflex camera. A 135 reflex is also about 10% of the cost second-hand, - a much better deal all round.

I would speculate that your RD-1 requires a minute adjustment of the rangefinder, which would not affect the accuracy with other lenses, but practical tests are the only way to be certain.

StuartR is wrong about a Noctilux having more DoF than a 50F1.4 asph, with both lenses wide open.

 

[StuartR]

Hektor -- please read my post carefully. I said that with all the lenses at their closest focus the Noctilux has the most depth of field. Obviously if they are all at 1m then the 50/1 will have less than 50/1.4, but at .7m, the 50/1.4 has less DOF at f/1.4 than the Noctilux has at 1m at f/1.

Feel free to do the math yourself. I don't trust myself, so I used a calculator and a unit conversion utility.
The depth of field charts are here:
http://www.leica-camera.com/imperia/md/content/pdf/objektive/datenbltterm-objektive/38.pdf
http://www.leica-camera.com/imperia/md/content/pdf/objektive/datenbltterm-objektive/40.pdf

 

[fgianni]

StuartR is wrong about a Noctilux having more DoF than a 50F1.4 asph, with both lenses wide open.

I would not be so sure about that, the noctilux has a closest focus distance of 1m, while for the 50mmf1.4 is 0.7m, maybe we should ask him about his sources and the formula.

 

[Hektor]

Here's a comparison in case it's of assistance to you,

All wide open, all focussed on 10 feet, rounded down to the nearest inch.

135 f2.8 dof 3"
90 f2 dof 5"
75 f1.4 dof 5"
50 f1 dof 8"
50 f2 dof 17"
35 f2 dof 36"

Have fun.

p.s. If you insist on taking pics at F1.4 and 2 feet distance, then use your Canon !

 

[fgianni]

The Nokton hasn't got a massively greater DOF than the 75 Lux, and the focus on the Nokton is spot on, that's why I am a bit reluctant in blaming the camera.
Anyway with 0.39 inches DOF it is always going to be a bit of a hit and miss, that's why I plan to bracket the focus with the Summilux, that way I get consistently good focus.

 

[fgianni]

Here's a comparison in case it's of assistance to you,

All wide open, all focussed on 10 feet, rounded down to the nearest inch.

135 f2.8 dof 3"
90 f2 dof 5"
75 f1.4 dof 5"
50 f1 dof 8"
50 f2 dof 17"
35 f2 dof 36"

Have fun.

My 20D autofocus must be brilliant! It consistently gets spot on shots with my 150mm at f2.8, and at 15 in distance, I wonder what's the DOF in that case, probably less than 0.4 inches, but I don't have the formula.