DOF for ASPH

[Mylek]

Advertisement (gone after registration)

Hello,

 

i read on a few posts comments between the DOF that was different between old lenses (preASPH) and the new ASPH at the same aperture particularly on the 28mm.

What are your experience with these findngs?

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[jaapv]

Do you mean DOF of Bokeh?

 

DOF is, all else being the same, mathematically identical between asph and non-asph lenses. However, the newest generation M lenses is so well corrected, high-resolving and precise that the sharpness falloff that produces the illusion of DOF in our perception is better defined, so for practical purposes it is a bit more narrow. Having said that, I have learnt not to rely on DOF for focussing especially with modern lenses on a sensor. There is but one sharp area in a photograph and that is the plane of focus.

[lars_bergquist]

If you compare two lenses with the same focal length and at the same aperture, one 'spherical' and one aspherical, they will have the identical depth of field. The formula for computing it is the same in both cases.

 

Now 'depth of field' refers to a space or zone within which sharpness is acceptable. It is emphatically NOT a zone within which everything is tack sharp and everything just out of it is abruptly out of focus. A lens is always sharpest (i.e. has the best resolution and contrast) in one plane, the plane of best focus. Fore and aft of this plane, sharpness drops off gradually. Calculations of d.o.f. are always based on some kind of assumption about how much of this fall-off you can tolerate. Classical d.o.f tables and scales since the 1920's have been based on the notion that a circle of confusion (the 'disc' that an unsharp point is reproduced as) of 1/30th of a mm is acceptable. Modern work is often and more reasonably based on an acceptable c.o.c. of 1/60th. But clearly, the actual c.o.c. that a good lens can produce at the plane of maximum sharpness is a lot sharper than that!

 

This is what has changed. The maximum sharpness/definition/acutance of modern lenses, not only those with aspherical surfaces, is usually clearly higher than with older designs, because technology has advanced. Think of the difference between a smoothly rolling ridge, and a knife-edge of bare rock. The crest is higher, but the lower slopes are unchanged. Consequently, inside the (unchanged) d.o.f. zone, the very 'crest' of sharpest detail may well be sharper to the eye than it was with a lens that was designed several decades ago. So while the places where the c.o.c. changes from 0.016 to 0.017mm or whatever – i.e. the limits of the d.o.f. – are the same as before, this 'acceptable unsharpness' is compared to a much greater maximum sharpness. And that can change the way we perceive the image. People who do not know any optics may well imagine that d.o.f. has changed – but that is by definition impossible. The maximum sharpness somewhere inside the d.o.f. zone simply does not enter into the definition and the calculations of the d.o.f., which consider only the outer limits of the zone.

 

Ever since the introduction of the first anastigmat lenses in the 1890's some people have complained that the new sharper lenses were 'harsh' or 'vulgar' or 'un-artistic' or whatever. So the fuzzies complain about the modern lenses, and the blame falls on the conspicuous letter combination 'ASPH'. But, it is not the aspherics, it is the advanced optical design. And I for one thinks these advances are really advances, not backward steps. Those who don't should take up croquis drawing or something (I have made some nice ones in my day, too – but that is a different medium of expression).

 

The old man with the camera and the croquis pad, sticking his neck out

[lars_bergquist]

I do basically agree with Jaap (who posted while I was writing my essay) but I am maybe less rigid. Zone focusing has its place when you wade into a celebration, demonstration or riot with a wide angle lens. There, other considerations than maximum acutance take over (as long as the acutance is not on something that hits you over the head).

 

Sharpness is the fetish of boring photographers.

 

The old man, sticking his wrinkled neck out even further

[pgk]

I rather like a comment made in a Zeiss publication (Depth of Field and Bokeh by H. H. Nasse, Carl Zeiss Camera Lens Division, March 2010): "Depth of field is based on the acceptable blurriness and is therefore essentially based on arbitrary specifications".

[lars_bergquist]

I rather like a comment made in a Zeiss publication (Depth of Field and Bokeh by H. H. Nasse, Carl Zeiss Camera Lens Division, March 2010): "Depth of field is based on the acceptable blurriness and is therefore essentially based on arbitrary specifications".

 

Perfectly true. And as long as our "arbitrary specifications" are the same for spherical and aspherical lenses, depth of field is the same. See my first posting for the numbers of the arbitrary specifications.

 

The old man from the Age of the Slipstick

[01af]

Advertisement (gone after registration)

Formulas for depth-of-field are simplified and don't take the effects of diffraction or lens faults into account. They are accurate only in a world where lenses are perfect and light is just rays, not waves. Unfortunately, that's not the world we live in.

 

As a matter of fact, many (if not all) Leica M Asph lenses indeed have less depth-of-field than their spherical counterparts, and consequently, more out-of-focus blur beyond the depth-of-field range. The difference is small but clearly noticable in direct comparison. Not sure if the same is true for aspherical lenses from other makers. I don't know the exact reason for this phenomenon but I don't think it's a direct consequence of the aspherical elements. Instead I guess it has to do with the kind or character of the residual lens errors and hence, is only an indirect consequence of the Asph design. Also purely spherical designs can exhibit different depths-of-field ... think of the Rodenstock Imagon lens, for example, which was famous for having more depth-of-field than other lenses of the same focal length.

[pgk]

I don't know the exact reason for this phenomenon

Try reading the Zeiss publication that I mentioned in my earlier post which has some explanations (I downloaded it from the Zeiss website).

[Erik Gunst Lund]

Please also note that it is a vey few, usually close up lenses, that have a DOF between two 'planes' as in 'flat' 'planes'

 

They are intended for copy work of 'flat' subjects like a book or drawing.

 

Most lenses have DOF between two spheres!

 

Same applies for Depth of Focus btw.

 

Some very sharp lenses have a very distinct 'look' to the unsharp image details in the zone where sharpness go to unsharp, for this I really only have good sample from the Nikon range of lenses...

[adan]

Technically, what a lens produces on the image plane is "Depth of focus." "Depth of Field" is dependent on enlargement of the picture afterwards.

 

It is true that depth of focus should be identical for any two ideal lenses of the same focal length. However, in the real world, spherical aberration can subtly increase the range over which a lens's image is "equally sharp."

 

See the following link to cross-sections of light rays passing through a lens (left to right). The center image is an "ideal" lens without aberrations. The top and bottom images are with spherical aberration added, which spreads out the zone of "almost sharp" - the brightest white band - such that one has slightly more leeway to capture "as sharp as possible" when slight misfocused.

 

http://upload.wikimedia.org/wikipedia/commons/thumb/1/10/Spherical-aberration-slice.jpg/398px-Spherical-aberration-slice.jpg

 

So I'd say, yes, the non-ASPH lenses with residual spherical aberration may very well give more leeway of focus, at the cost of less-perfect maximum sharpness. I'm not sure this applies as much to 28mm lenses, but more to the longer lenses (90 f/2, 75 f/1.4 and 50 f/1.4 non-ASPH compared to the ASPH equivalents).

 

BTW - this is pretty much agrees with what Lars said. Better absolute sharpness with ASPH, with a quicker drop-off to fuzzy; longer zone of "equal apparent sharpness" with pre-ASPH lenses.

 

 

IMPORTANT EDIT: And yes, it is not the ASPH glass in and of itself that makes the difference. But more modern designs, which happen to incorporate ASPH elements as a tool. I agree with 01af and Lars on that. But ASPH is such convenient shorthand....

___________________

 

A side note (since someone mentioned the "B" word) - this diagram also shows why a lens with spherical aberration can produce either very soft glowing bokeh (bottom image) or nasty bright-ring bokeh (top image). Note how the outer rays are diffused in the one (undercorrected SA), and concentrated into a bright line in the other (overcorrected SA).

 

The bottom would reflect the creamy bokeh of the 75 Summilux or 90 Summicron non-ASPH - the top reflects the real-world look of the non-ASPH 35 and 50 Summicrons.

 

The "perfect" lens (center) produces "neutral" moderately-blurred rays - which is my experience with the ASPH Leica lenses.

_____________

 

On yet another side note - this diagram also shows why lenses with spherical abberation can have "focus shift" as the lens is stopped down. Cut away the outer rays, as your aperture blades would do, and the point of best sharpness slides along that long white band from one end to the other.

 

Again, the more ideal lens is focusing all the rays in one place, so that stopping down does not change the point of best focus.

[Erik Gunst Lund]

....The "perfect" lens (center) produces "neutral" moderately-blurred rays - which is my experience with the ASPH Leica lenses...

 

Thanks Andy & et al! Really a great thread!

 

Andy, for future reference; I would really like to know which ASPH lenses you're referring to? Before buying more M lenses,,, I'm a sucker for good B.

[adan]

In my experience, the ASPH 35 f/1.4, 50 f1.4, and 75/90 f/2 lenses all have neutral bokeh - minimal bright rings. Generally, there is not enough blur with wider lenses to see much difference, although I've noted the 28 f/2 ASPH has smoother bokeh (such as it is) than, say, a 28 f/2.8 v.3, which produces bright rings.

 

The "best" bokeh still comes from a couple of the older non-ASPH designs: 75 f/1.4 and 90 f/2. At the cost of a bit less knife-edged precision in the focused areas.

[pico]

This has been the best thread on DOField (and in once case DOFocus) I've ever read, and I've read hundreds.

 

Perhaps repeating in fewer words - note the given Circle of Confusion (CoC) presumed in the DOF charts. For _practical purposes_ each capture format size has a preferred CoC, and a presumed viewing distance. I wonder (but have not looked) if M9 and M8 charts might differ in CoC. I do know that in the earlier literature of the Hasselblad SWC they warn that if one is printing a photograph to enlarge particularly large (say, 10X), that a more conservative CoC should be used (and they recommend the same a FF 35mm). I guess they presume a pessimistic viewing distance. As Jaap has taught me that with digital, it encourages may of us pixel peep more than we would with conventional film enlarging which leads to pessimistic CoC requisites. It's a real shift in expectations (for printing).

 

Regarding the comment by 01af about the Rodenstock Imagon lens - IMHO it's got no more DOF than any other lens of the same focal length. At least in my experience. (Stopping it down a lot w/o seive destroys its particular virtue of a soft lens, but bokeh is so wonky you have to shoot to understand its behavior in that regard.) However 01af, and certainly Lars, might be taking into account our brain's penchant for resolving an image based upon what is supposed to be sharp, and adjusting background proportionately. That, to me, is the case and one reason I am habituated to non-ASPH lenses. ASPH are just too 'sharp' - micro contrast is extreme. (Old lenses for old eyes?)

 

I look forward to any corrections

--

Pico - the blind one

[01af]

This is pretty much agrees with what Lars said. Better absolute sharpness with Asph, with a quicker drop-off to fuzzy; longer zone of "equal apparent sharpness" with pre-Asph lenses.

No, it does not agree with what Lars said.

 

Lars said with a lens that has better maximum sharpness, depth-of-field only appears to be narrower but technically remains the same always. But in fact, technical depth-of-field, as determined against a given circle of confusion, will be different. So will out-of-focus blur.

[Erik Gunst Lund]

In my experience, the ASPH 35 f/1.4, 50 f1.4, and 75/90 f/2 lenses all have neutral bokeh - minimal bright rings. Generally, there is not enough blur with wider lenses to see much difference, although I've noted the 28 f/2 ASPH has smoother bokeh (such as it is) than, say, a 28 f/2.8 v.3, which produces bright rings.

 

The "best" bokeh still comes from a couple of the older non-ASPH designs: 75 f/1.4 and 90 f/2. At the cost of a bit less knife-edged precision in the focused areas.

 

Thanks! I agree from own experience 35 1.4 ASPH and 70 2 ASPH and I can add the Leica M 28 2.8 ASPH Elmarit from my list; Very fine neutral Bokeh.

[adan]

No, it does not agree with what Lars said.

 

Well, OK - Lars said there appears to be more DoF field with older lenses (but really isn't) and I said there appears to be more DoF with older lenses (because there really is, thanks to SA). We agreed on the "appearance," if not the underlying reality and causality.

[pico]

Well, OK - Lars said there appears to be more DoF field with older lenses (but really isn't) and I said there appears to be more DoF with older lenses (because there really is, thanks to SA). We agreed on the "appearance," if not the underlying reality and causality.

 

You know you have really ruined the day for military resolution target fanatics. And this: "underlying reality and causality" really makes the day for philosophers and .... well, if I make a picture after drinking Guinness, should I require my viewers to have the same cause and underlying reality?

 

Ok, ok .... I'll experiment.

.

Off for a bicycle ride. It's a toasty 40F today.

[jaapv]

Well, OK - Lars said there appears to be more DoF field with older lenses (but really isn't) and I said there appears to be more DoF with older lenses (because there really is, thanks to SA). We agreed on the "appearance," if not the underlying reality and causality.

Well, actually DOF is all about appearance, as it is a subjective phenomen, dependent on the perception of the viewer.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[DOUG66]

I have just looked at a series 4 Summicron and an ASPh Summicron held side by side and up to a light source. Looking at the size of the exit pupil of both lenses it is obvious that the series 4 lens subtens a larger pupil than the ASPH lens. In fact the exit pupils are roughly the same if the series 4 lens is stopped down to 2.8 and the ASPH is left at f2.

Both lenses then subtend the same angle with the eye. Does this not mean that the ASPH lens has enhanced DOF ?

Doug.

[DOUG66]

Apologies I should have said I was refering to 35mm Summicrons.

Doug.