Depth of Field of the new Noctilux F/.95

[imported_donatellobirsa]

Advertisement (gone after registration)

Has the Depth of Field (DoF) of the new Noctilux F/.95 been ricalculated by Leica?

 

In the photo of the The new LNA (leica noct asph) in Puts's analysis of the new lenses,

 

the hyperfocal distance looks to be different .

Infact HD at infinity At F/16 is 25 feet (ca. 7.5 meters)

instead of 5 meters (old Noctilux F/1 and all 50 mm).

 

DoF recalculated by Leica,

 

or simply wrong? :-)

[lct]

Hyperfocal distance does not change with the subject distance but depends on the circle of confusion (CoC).

Perhaps 2 different CoCs have been used like 0.03mm for 35mm film and 0.023mm for the M8.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[imported_donatellobirsa]

Hyperfocal distance does not change with the subject distance but depends on the circle of confusion (CoC).

Perhaps 2 different CoCs have been used like 0.03mm for 35mm film and 0.023mm for the M8.

 

I agree.

But it is the first time I’have seen this change in the CoCs used , in the world.

I was just looking for some information, if you know.

I think all of us know what is CoC, me too, even if my english is terribile, because I never write in english.

 

But in the new 21 Lux the CoC is the usual one.

 

So I think my observation is pertinent .

I was just wondering what’s going on.

 

Ciao,

Donatello

[lct]

...it is the first time I’have seen this change in the CoCs used...

Hi Donatello, i don't know what happened there but you know that the circle of confusion depends itself upon the crop factor which is 1.33 for the M8 for instance. So given a usual CoC value of 0.03mm for 35mm film, one get 0.03 : 1.33 = 0.023mm for the M8. FWIW

[imported_donatellobirsa]

Hi Donatello, i don't know what happened there but you know that the circle of confusion depends itself upon the crop factor which is 1.33 for the M8 for instance. So given a usual CoC value of 0.03mm for 35mm film, one get 0.03 : 1.33 = 0.023mm for the M8. FWIW

 

Could be, but they did not say anything.

and why this change in DoF references just for the Noctilux?

The new Luxes use the old doF scales, apparently.

I'm curious to see the tables of the DoF of the new Noctilux.

 

P.S.

I Frequently read your posts lct,

and i also love the old 50 Lux (above all the preasph),

and the old Noctilux for the mood,

I use practically only films (provia 100 F and Velvia 50 - Old (in freezer, and new),

and I am tempted by the new "24 Lux, above all...

[gravastar]

This could get very confusing .

 

The depth of field to a first order only depends upon the chosen Circle of Confusion, overall Reproduction Ratio and Aperture. Calculators that ask you to input focal length and subject distance but not final image size are using those two parameters to calculate reproduction ratio based on projection onto a sensor or film frame. They have no way of taking into account final print size or viewing distance (which require an appropriate choice of CoC). Unless Leica tell us we don't know what assumptions they have used to calculate the depth of field.

 

 

Bob.

[zapp]

Advertisement (gone after registration)

Has the Depth of Field (DoF) of the new Noctilux F/.95 been ricalculated by Leica?

 

In the photo of the The new LNA (leica noct asph) in Puts's analysis of the new lenses,

 

the hyperfocal distance looks to be different .

Infact HD at infinity At F/16 is 25 feet (ca. 7.5 meters)

instead of 5 meters (old Noctilux F/1 and all 50 mm).

 

DoF recalculated by Leica,

 

or simply wrong? :-)

 

Although there are a lot of errors in Leica data sheets, it makes sense to reduce the hyperfocal distance in the calculation. The circle of confusion is the factor in the calculation that changed. This is not just due to a crop factor, but to the higher resolution of digital cameras. Today we have 6-9 micron wide discrete pixels while the resolution of film has a resolution of aproximately 12-15 micron (the grains maybe smaller, but the noise is much larger on film). So with digital sensors slight unsharpness is more disturbing than for film, meaning that you should reduce the circle of confusion compared to 30 micron that were often used for film.

2-3 x the pixel pitch would be a realistic value for the CoC - yes, I know that a CoC of 15 micron is tiny, but that's the way it is.

By the way at f1/16 the small pixel size of modern sensors imposes a diffraction problem - bending of light at the aperture which causes unsharpness. For a pixel pitch of 6-7 mircon refraction will limit the resolution when the aperture is closed to f/8 or further. Byebye sharp 16.

Final remark. Digital sensors require lenses with much better resolution than film did, and this perfomance must be reached at intermediate apertures. Having both, which is not easy to find at the moment - especially outside of some German lens catalogs - digital sensors can resolve much more detail than film did for a given format.

[lars_bergquist]

The Leica circle of confusion standard of 1/30th of a mm made sense during the 1920's, when it was introduced. It is generally agreed that if in a print viewed at the normal smallest reading/viewing distance of 30cm/12" the c.o.c. is no larger than 1/10th of a mm, then the print looks acceptably sharp. At that time the most common amateur format was 6x9cm, and most negs were actually printed contact, i.e. same size. That was the competition. A 1/30th c.o.c. in the negative then permitted a 3x enlargement ratio, giving a print of 7x11cm size. All well and good.

 

But madmen like Dr. Wolff did start to do larger enlargements of 10x or more already during the 1930's, driving the development of better emulsions and developers. The 'official' c.o.c. was obsolete. I have repeatedly wondered aloud why Leica don't revise their standards, which are insufficient even for a 10x15cm print. The reason is probably that they don't dare—people would believe that their lenses had suddenly gone bad, because they delivered less d.o.f.! Are Leica users that stupid? I think the majority of them are.

 

Actually, depth of field itself is obsolete. There can be no hard and fast criteria. Modern lenses too don't give their best when stopped down much below 5.6 because of diffraction. So even if I do a landscape, I stop down only to 5.6 or thereabouts and put the plane of maximum definition on a feature somewhere in the far foreground. The far distance will look a bit soft. What about it? Total sharpness all over, even if it could be achieved, would look utterly boring. Forget about it. Point focus and enjoy the sharpness where it is. Much of the sense of depth in the picture comes from the contrast between sharp and soft.

 

Remember, there is no zone where everything is tack sharp while outside it everything is mushy soft. There is always a plane of maximum sharpness, and it is a plane without depth! On both sides of it, definition falls off increasingly, until it may pass some arbitrary limit. Or not ... you decide the limits.

 

The old man from the Age of Meniscus Lenses

[imported_donatellobirsa]

Thanks You!

A lot of interesting information,

but a kind of “Elvira Coot’s advices”, so to say,

I do not know the right translation for

“i consigli di nonna papera”

referring to something selfevident.

 

I think all of us know (I really hope so) that:

 

“The Leica circle of confusion standard of 1/30th of a mm made sense during the 1920's”

 

“The 'official' c.o.c. was obsolete”

 

“I have repeatedly wondered aloud why Leica don't revise their standards”

 

And so on…

 

 

I usually use my Luxes and Noctilux and never stop down more than F/2.8.

Nevertheless I could notice that Modified references for the DoF just for the Noctilux.

The new Luxes (21 and 24) use the old doF scales, apparently.

The 21 at F2.8 has an hyperfocal distance (HD ) of ca. 5 meters (look at the image of the lens focused at infinity in thePut’s site

. Photokina 2008

 

So, if a change there’s been, it is just for the Noctilux and I don’t know anything about it.

Someone das?

 

Personally I could be happy for this chage, or indifferent

But I like to know something. Leica. Knows…

 

Evidently I’m the only "problematic" people here.

 

P.S. I agree that actuall lenses on the crop digits are not good at F/16.

But no one told me that the Noctilux has not to be used on films, has it? :-)

 

If You like and understand italian language,

visit us at:

L E I C A P A S S I O N F O R U M :: Index,

and here we told about the same subject:

Redirect

 

Have a nice week end,

 

Ciao,

Donatello

[lct]

We've discussed this on the previous forum already, but reducing CoC values below usual ones would have the consequence that many Leica lenses could not be focused accurately on the M8 any more.

If one reduce the 35mm film CoC value to 0.020mm instead of 0.030mm for instance, the CoC value of the M8 would be 0.020 : 1.33 = 0.015mm.

With such a tiny value, the effective base length of the M8's rangefinder would have to be longer than 57mm to focus a 90/2.8 lens at full aperture where that of the M8 is only 47.09mm.

 

To check the figures above:

- The effective base length of the Leica M8 is 69.25 mm mechanical base length * 0.68 image magnification = 47.09 mm.

- The required base length is given by the formula b' = (e * f^2) / (k * z) where b' is the effective base length, e the visual acuity (0.0003 at approx. 1 arcmin), f the focal length, k the aperture and z the circle of confusion.

- See Josef Stuper, Die photographische Kamera, Editor: Lindemanns, ISBN-10: 3895061921, ISBN-13: 978-3895061929

[lars_bergquist]

We've discussed this on the previous forum already, but reducing CoC values below usual ones would have the consequence that many Leica lenses could not be focused accurately on the M8 any more. ...

Now this is nonsense. Let's say that I am focusing a 35mm lens at 5 meters. The d.o.f. scale on the lens, computed for a c.o.c. of 1/30th of a mm, claims that everything from c. 2.5m to infinity will be nice and sharp at f:8. I take the picture and the print does show that this is definitely not the case. All right, there is a plane of sharpness at 5 meters, but how is the situation in front of and behind that plane?

 

What DOES look sharp does depend on print size and viewing distance, but if this is an 18x24cm or 8x10" or A4 print held at 30 cm or so, I will likely find that acceptable visual sharpness goes from about 3.2 to 10m. If the d.o.f. scale is recalculated for a maximum allowable c.o.c. of 1/60th of a mm, then this is also what the scale would have told me.

In fact, what would have happened would have been that the scale markings for f:4 would have been renamed 'f:8' and that's all. The markings for f:4 of course would have moved to where f:2 was before ...

 

Now please do tell me if and how that would have made my focusing at 5 meters less precise? I'm listening. In fact, most pros using film cameras have long ago gone over to using the markings for the next more open stop, in this example for f:5.6. Was that detrimental to their focusing? Now with the M8 sensor we move one stop further. That's all.

 

The old man from the Age of Scale Focusing

[lct]

Now this is nonsense. Let's say that I am focusing a 35mm lens at 5 meters. The d.o.f. scale on the lens, computed for a c.o.c. of 1/30th of a mm, claims that everything from c. 2.5m to infinity will be nice and sharp at f:8. I take the picture and the print does show that this is definitely not the case...

Appreciate the 'nonsense' thanks but it is accurate if you use your M6 and it is not accurate with your M8 for the simple reason that the CoC is not the same any more.

[lars_bergquist]

A P.S. What I wanted to say in the second part of my first posting—that d.o.f. thinking is obsolete—is that with modern print/projection sizes and with sharper films and sensors, zones of acceptable visual sharpness have shrunk to the degree that they have become meaningless, except with very wide lenses.

 

Besides, pictures where the main subject is glued directly onto a tack sharp background are boring. Is the background equally important as the background? Yes of course if the picture is "my girlfriend in front of the Eiffel Tower/the Royal Pavilion/the San Pietro in Vaticano etc." Is that the kind of pictures you take? The placement of the plane of sharpness is one of our main tools in directing the viewers attention, just as composition is. "But I'm more interested in the wastebasket there to the left in the background." —"That's another picture."

 

And pictures without a main subject are even more boring.

 

The old man from the Age of the 6x9cm Contact Print

[lars_bergquist]

Appreciate the 'nonsense' thanks but it is accurate if you use your M6 and it is not accurate with your M8 for the simple reason that the CoC is not the same any more.

The circle of confusion is the little disc of light (surrounded by a diffraction pattern) that is the image in the image plane of a point on the subject plane. If this little image (in a print at normal reading distance) is more than 1/10th of a mm, then the point image is not sharp. The size of the point depends on the quality (resolution) of the lens, and on how well it is focused. The allowable c.o.c. in the negative/on the sensor depends on the degree of enlargement. Enlarge ten times and the allowable c.o.c is 1/10 / 10 = 1/100 of a mm.

 

The film or the sensor can degrade that value. But it can't improve on it.

 

The same old man

[tom0511]

All I can say that when I once compared old Nocti, 50 Summilux and 50 Cron at same f-stop that DOF seemed different. It also might have to do with correction etc., but thats what I saw.

I also have the theory that the "visible" DOF has to do with sharpness of a lens and AA-filter. A very sharp lens with a thin AA filter delievers a "sharper" focal plain, so the "sharpness-difference" to the other areas looks more pronounced compared to a softer lens and a weaker AA-filter.

If I use the 50lux asph and the M8 I get a different results vs using the D3 at 70mm.

 

I have to say that I find the difference between sharp and unsharp area less pleasing in digital images compared to film (I would think caused by the thickness of the film emulsion).

 

What I wanted to say is that I believe the visual effect of DOF can not be easyly described by 1 or to formulas - there are many factors which influence what we finally can see in the image. Just my opinion.

Cheers, Tom

[SJP]

As far as I know the lens barrel markings are obtained purely by computation, using the historical value of 31 micron for the CoC. Leica does not measure the depth of field.

For example the 28/2-ASPH lens markings are identical to the 28 mm setting of a MATE, and the same applies to a 35mm summaron vs. MATE at 35. So 50 years of lens development did not change the DOF scale.

 

It therefore seems that Leica has decided to change their definition of CoC to a more stringent value of 20 micron on the new Noctilux.

 

I checked using the spreadsheet, fill in 50 mm, f /11, coc 20, So+Si = 10000, then the DOF is from 9.98 m to infinity, exactly as on the new Noctilux barrel.

[lars_bergquist]

All I can say that when I once compared old Nocti, 50 Summilux and 50 Cron at same f-stop that DOF seemed different. It also might have to do with correction etc., but thats what I saw.

I also have the theory that the "visible" DOF has to do with sharpness of a lens and AA-filter. A very sharp lens with a thin AA filter delievers a "sharper" focal plain, so the "sharpness-difference" to the other areas looks more pronounced compared to a softer lens and a weaker AA-filter. ...

Your observation is correct. Take a pinhole camera e.g. The definition is bad, but it is uniformly bad from near to far, so the d.o.f. can be said to be infinite! The smaller the difference is between the definition in the plane of best focus and planes on either side of it, the wider is the zone where the sharpness is not appreciably worse than at best focus. The old Noctilux has a somewhat 'wooly' definition wide open, and this serves to partially hide the fall-off outside the plane of best focus.

 

Some people have complained that the modern Leica lenses are just too sharp ("there is a cloud to each silver lining"). This great sharpness in the plane of best focus accentuates the fall-off outside it, and some people with little grasp of concepts think that this makes for 'bad bokeh'. But the out-of-focus parts of the picture have not become more out of focus, or necessarily more unpleasant. In fact, modern aspherical lenses do often have an appreciably better bokeh, because spherical aberration is under better control. If the complainers think that better lenses are less good, there are two remedies: Use 'classical' less well corrected lenses, or buy a mild soft filter ...

 

The same old man

[Guest volkerm]

Hi Lars,

 

I don't agree with your conclusion on bokeh. The quality of bokeh (smoothness) is determined by the distribution of light within the out-of-focus circles (blur disk). For a nice smooth bokeh, we want the brightness with the oof circles to drop towards the edge of the blur disk.

 

Some manufacturer offer lenses where you can control the amount of residual spherical aberration (e.g. Nikon DC Nikkor). From my experience with these lenses, a neutral setting with minimum spherical aberration will create a "neutral" bokeh where the light distribution within the blur disk is homogeneous. In other words: that does not provide the "nice and smooth" bokeh which some (including myself) are looking for.

[lars_bergquist]

But it is exactly the amount of spherical aberration present that causes this distribution of light within the blurry circles produced by e.g. out-of-focus point highlights—which is an important component of what we call bokeh. Beyond the plane of best focus, undercorrected spherical makes for circles with a brighter edge than center, while overcorrected creates circles with a concentration of light in center. Neutral spherical makes for evenly bright circles. And on the hither side, the effects are in fact the opposite!

 

None of this has of course per se anything to do with the SIZE of the theoretical circle of confusion, because such a circle is an image of a dimensionless point, and no real highlight is actually a geometrical point. A large c.o.c. will mainly lower resolution, and in more drastic cases, edge sharpness. And the large c.o.c. can be caused by several other optical aberrations besides spherical, including diffraction. An effect that you could actually use in the darkroom to combat grain.

 

The old man from the Age of the El-Nikkor

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[Guest volkerm]

while overcorrected creates circles with a concentration of light in center.

 

That is what I was talking about - "nice smooth bokeh" and "perfect correction" are different optimization goals. A perfectly corrected lens will not have a nice smooth bokeh, and vice versa.