M Lens DOF question

[63strat]

Advertisement (gone after registration)

This is the review that threw me. This guy claims the 28/2 Summicron has less DOF than the 28/2.8 Elmarit. What do you guys think? True? Here's the review:

 

"I bought the f2 28mm Summicron Aspeherical to replace a 28mm 4th version Elmarit. The Elmarit was already excellent for my work, but, given the reviews of the 28mm f2 Summicron, I assumed I would see at least a degree of improvement in results.

 

The Summicron is, without doubt, a high resolution optic. However, I was unprepared for the marked reduction in available depth of field, even at smaller apertures, compared to that of the Elmarit. I happily use the latter at 2.8, and seldom miss a shot due to focusing error. I even use the 35 f2 Aspherical in the same way, and have never experienced this limitation with the lens.

 

With a rangefinder camera, I always work at wide apertures and very close in, using the 28mm as a 'standard' lens, more than as a wide-angle. With the Summicron, the reduced depth of field proves very difficult for this sort of working, even when stopping-down to f4. I have good technique, but this 28mm Summicron challenges it. I found myself using higher ISO settings to gain a stop or two in aperture. This is not what I was hoping for. I'd bought the 28mm Summicron for the extra stop - suddenly, I was working at least a stop down from my usual working aperture." PhotographyReview.com

[Advertisement]

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

[lct]

This is the review that threw me. This guy claims the 28/2 Summicron has less DOF than the 28/2.8 Elmarit. What do you guys think?...

Looks like this reviewer's Summicron and/or M8 need some adjustment. Whatever their wider aperture may be, two 28mm lenses have the same DoF at the same aperture provided their are used with the same film or sensor format. BTW i own both the 28/2 and 28/2.8 asph and never noticed different DoFs with my Epsons so far. I have no experience with the pre-asph Elmarit v. 4 though.

[Guest JasonG]

DOF depends on two parameters only: (1) the f-stop used, and (2) the reproduction ratio

 

Hi,

 

I'm pretty sure you'll find that the f-stop is related closely to the focal length, hence DoF involves focal length as a parameter.

 

Tra

 

Jason

[jaapv]

Mmmmmm... need to think whether that is right, Lars

 

Meanwhile click here Depth of field - Wikipedia, the free encyclopedia the ultimate expression is:

[ATTACH]117086[/ATTACH]

 

With N aperture, c circle of confusion, f focal length, s the subject distance

 

Based on that if you keep N, c and s constant the DOF goes down with increasing focal length. This is the 'normal interpretation, wide angle has a larger depth of field than a tele.

 

 

To confuse the matter I disagree ( a little) The formula is good for all practical use, but notice the wavy "equals =" That means it is an approximation leaving out one or more minor parapeters for acceptable interpretation.

And guess what one of these minor parameters is, contributing just a little bit, but still something? Focal length...To complicate matters more, DOF is dependent on lens design. All formulas are based on symmetrical (double gauss) lenses, but a retrofocus or tele lens will show a different DOF, albeit by a small margin.

 

For instance, this formula which I lifted from this Depth of Field and the Small-Sensor Digital Cameras - photo.net excellent article incorporates the focal length and subject distance to represent the magnification. If you work it through, you'll find that focal length is not completely cancelled out;

 

F is the focal length. D is the subject distance, c is the circle of confusion and fn is the f# (f-stop) of the lens.

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!

[SJP]

Jaap I think we agree on the approximate nature of the equations, although they are not too far off compared to the 'exact' result.

 

Comparing the 28 and 90 mm at f/2.8 on the M8 we have the following from the spreadsheet which does not suffer from any major approximations.

 

1 No cropping comparison at 2 metre focussing distance

28mm, CoC 23.5, f/2.8, DOF = 70 cm

90mm, CoC 23.5, f/2.8, DOF = 6 cm

DOF ratio 11.6

 

2 cropping the 28mm to 90mm equivalent frame (reduce CoC by factor 28/90):

28mm, CoC 7.3, f/2.8, DOF = 22 cm

90mm, CoC 23.5, f/2.8, DOF = 6 cm

DOF ratio 3.67

 

Comparing to the shark test shots, my guestimate of the Depth of Field (situation 2, with cropping) is way too lenient but the measured ratio 3/1 is more or less in agreement with the calculated ratio 22/6 (see images 2 and 3 previous page).

 

Comparing images 1 and 3 (no cropping, i.e. situation 1) convincingly shows the predicted large difference in DOF in that case.

[jaapv]

Yes, we do agree, Sander. But to return to the OP, Take two lenses of the same focal length at the same aperture and DOF is the same.

The variations may be:

1.st Lens construction - but that difference is unimportant for actual images

2.nd The "steepness" of the DOF and the harshness of the out of focus areas. This can make a vast difference in the perception of DOF.

3.rd This is related to #2, the contrast of the lens. A contrasty lens will seem to have less DOF

4 th Again related, the initial sharpness of the lens.

5 th Film or sensor.

 

All these factors combined will influence apparent DOF (at a guess) up to two stops, imo.

[SJP]

Advertisement (gone after registration)

'Johan' we agree on some things at least which is good.

 

Re. the OP although I can see your point on percieved DOF concerning contrast & what have you, I would be surpised if it would amount to 2 stops difference when comparing lenses of the same focal length at the same aperture.

 

If we take the Leica datasheets the DOF is the same for all 50mm lenses ever made (at the same aperture) & the same applies to all other focal lengths (compare e.g. 90 APO & TE). If the 2 stops difference would be true from now on we can safely ignore everything about the theory of how DOF comes about. In addition the markings on the lens barrel would be of no value at all not even as a guideline.

[lct]

...If the 2 stops difference would be true from now on we can safely ignore everything about the theory of how DOF comes about.

Theory can explain 1 stop difference when using APS cameras. It just comes from the CoC divided by the crop factor.

[roguewave]

63strat (as in Stratocaster?) The real issue is how the OOF areas make the focused plane areas (the IN-focus areas) look & feel to the viewer. One can often find many ways to give an illusion of more depth with a smaller DOF. Leica lenses have different signatures and they make the viewer experience the space in the image in different ways. What's most important, unless you you are using the lens as a "copier", is to register the feeling and specificity of the most salient features in the frame in a way that leaves no doubt about how you experience this space, not what it looks like.

[ho_co]

(or a '63 Stradivarius?)

 

Do you have a link to the statement? Not important but sometimes helpful.

 

Is it possible that the 'reviewer' had recently switched to digital?

 

As shown in one of the LuLa links I posted earlier, DoF doesn't _appear_ to work the same way with digital as we were used to with film, simply because film has thickness and the digital sensor doesn't.

 

Leica mentioned the same point in the LFI article in which they discussed the focus shift of the 35/1.4 ASPH.

 

The 'reviewer' may have been remembering "how things were with film" and not expecting change with digital.

 

This is just an amplification of point #5 in Jaap's post 26 above.

 

 

Nonetheless, by definition, formula and practice, as a number of people have said here, all 28mm lenses at a given aperture and a given distance on a given format will produce the same depth of field. Jaap has mentioned a range of issues that can influence depth of field, but as you've seen, they are disregarded in practice. As Stephen has said, Leica uses the same depth of field scale for all its 28mm lenses. But Leica also admits that the data used for calculating these values are out of date. And in the M8 instruction manual, Leica erroneously states that the depth of field engravings on their lenses hold both for film and the cropped sensor of the M8.

 

As this thread makes obvious, no matter how simple and direct the depth of field formulas, the thing in itself is much more complicated.

[SJP]

Theory can explain 1 stop difference when using APS cameras. It just comes from the CoC divided by the crop factor.

Sure no problem, same applies to FF vs. M8, MF, enlargements to A0 etc. What (simple) theory does not allow for is differences between lenses in DOF terms if they have the same focal length and are used on the same sensor/crop. So a Noctilux on an M8 at f2.8 is 'the same' as a 50mm elmarit at f/2.8, or a 50 summilux at f/2.8 in terms of DOF. The signature will be different obviously and out of focus rendering by the Noctilux will probably appeal more than that of the elmarit (or vice versa).

 

That we may be more critical of focus in the digital age is presumably due to the immediate ability to evaluate down to the pixel level. The "depth" of a film vs. that of a sensor is an interesting point that requires some more thought. A film emulsion I guess is about 20 micron thick, less than 5 is not opaque enough, more than 50 is a waste of silver (= money). Does anyone know the emulsion coating thickness?? A digital sensor is not strictly 2-D (infinitely thin) as it also will accept convergent and divergent beams entering just above or below the 'optimum', but this requires dissecting the sensor imaging geometry as well - with shifted microlenses etc. Sounds like fun

[jaapv]

The newer emulsions are about 30 micron thick.

And yes, the DOF scale on the lens is nowadays not more than a guideline.

[lct]

If there were significant differences between film and digital as far as DoF those differences would be visible with full frame cameras would not they. Do you see actually such differences folks? I don't see any personally so either in practice or in theory there is no such thing as a digital DoF to me.

[ho_co]

Do you see actually such differences folks? I don't see any personally so either in practice or in theory there is no such thing as a digital DoF to me.

lct, certainly you're not saying that the depth of field in the LuLa link is what you would expect from film?

 

Perhaps I'm extrapolating in relating the LFI article to depth of field, but if focus shifts, that also shifts depth of field, doesn't it? And Leica explained quite clearly how a lens that with film was noted for being extremely sharp was suddenly with digital recognized as having a severe focus shift.

 

When you say that you don't see it personally, remember that the fact that I couldn't identify the problem in a shot by Tim Ashley doesn't mean he couldn't.

 

You are quite correct when you say "either in practice or in theory there is no such thing," and that is the reason we're still talking about it: Although the equations and theory we grew up with don't show it, practice does.

[lct]

lct, certainly you're not saying that the depth of field in the LuLa link is what you would expect from film...

Sorry Howard i don't know who LuLa is i'm afraid aside from the president of Brazil. I simply say that i use the DoF markings of my Leica lenses the same way with my Canon 5D as with my Leica R4s. So i simply ask you and other friends here to forget for a minute the crop cameras problems and to check if the DoF scale of their lenses still work with FF digicams. Fair enough no?

[SJP]

Concerning the OP question I plan to do a comparison between a Summaron 35 and a tri-elmar @ 35mm both at f/4. Similarly summicron 28 and tri-elmar @ 28mm. Obviously I will be using the shark as my model.

 

Anyone else prepared to test a few lenses of the same focal length to sort fact from fiction? Testing at wider apertures would make sense (f/2 or so) but the MATE does not let me do that.

[jaapv]

Sorry Howard i don't know who LuLa is i'm afraid aside from the president of Brazil. I simply say that i use the DoF markings of my Leica lenses the same way with my Canon 5D as with my Leica R4s. So i simply ask you and other friends here to forget for a minute the crop cameras problems and to check if the DoF scale of their lenses still work with FF digicams. Fair enough no?

No they don't. As a rule of thumb calculate DOF one stop tighter on the M8.On FF half a stop tighter.

[lct]

..On FF half a stop tighter.

Half a stop tighter than film? May i ask you which cameras and lenses you're using this way?

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[ho_co]

Sorry Howard i don't know who LuLa is i'm afraid aside from the president of Brazil.

With respect, lct, you seem always to try to avoid the question and ask that others approach things your way. I asked whether you felt the results shown at a particular link were in line with what you expected from film. Your answer as I see it is, in so many words, "Gee, I dunno. Should I look?" No offense intended. It's just a difference in presentation style.

 

I simply say that i use the DoF markings of my Leica lenses the same way with my Canon 5D as with my Leica R4s.

So you should. Theory says that the results should be the same comparing two full-frame cameras. For general purposes there should be little enough difference between film and digital as to make no difference. This forum became aware of the importance of the distinction because of results from a particular lens. Maybe the changes in quality control that Leica put into effect after studying Tim Ashley's 35 Summilux-ASPH will have made the distinction moot. As far as that goes, we have no way of knowing whether the equipment that generated the results shown at Digital Focusing Part One was properly calibrated.

 

So i simply ask you and other friends here to forget for a minute the crop cameras problems and to check if the DoF scale of their lenses still work with FF digicams. Fair enough no?

Both missing the issue (as I see it) and already covered in the previous paragraph. "Modern" depth of field scales have for at least forty years been recognized as being traditional and inaccurate. If you or I find them accurate enough, so be it. They are certainly 'in the ballpark' and good enough that Leica hasn't modified them. Perhaps one of the reasons the newest AF lenses don't include DoF scales, IR focus indices or even indicators of hyperfocal distance is to avoid such discussions as this thread. Does your 5D instruction manual include the word "hyperfocal"? Mine for the Nikon D200 doesn't.

 

Don't misunderstand. I think this back-and-forth of suggestions and contradictions is enlightening. We're all coming to different conclusions despite the fact that we all come from the same background. The very least that will come out of such exchanges is that we will all have a better understanding of the issues. And since there's so much wiggle room, we'll probably all come away feeling that our initial positions have been vindicated.

[ho_co]

Concerning the OP question I plan to do a comparison between a Summaron 35 and a tri-elmar @ 35mm both at f/4. Similarly summicron 28 and tri-elmar @ 28mm. Obviously I will be using the shark as my model.

 

Anyone else prepared to test a few lenses of the same focal length to sort fact from fiction? Testing at wider apertures would make sense (f/2 or so) but the MATE does not let me do that.

Stephen, I'd love to join the brigade, but I don't have a stuffed shark.

 

Seriously, I'm a bit pressed for time at the moment. I may be able to compare two 50mm lenses at f/4 or two 21s at f/4. If someone else can do that, I'd like to see the results.

 

I'm glad to see your practical approach, and fascinated to see that its results apparently differ with the outcome of Michael Reichmann's comparison.

 

We'll never discover the truth about depth of field, but the wandering is certainly worthwhile!