Do you agree with Leica?

[msendin]

Advertisement (gone after registration)

Text bellow is from the instructions booklet of the 50m lux asph, when talking on page 12 of the 1.33 factor of lenses used on the M8.

 

" Therefore, when used on the Leica M8, these lenses have angles of view corresponding to lenses with focal lenghts that are longer by a factor of 1.33. This has the respective effects on the perspective, but not on their depth of field, which, with the Leica M8, can also be read directly off the lense (see depth of field scale,p.14) "

 

As far as I understand DOF depends ONLY on "amplification", or enlargement, of the image in relation to the real size of the object.

A smaller sensor is a larger enlargement so DOF is smaller than the orientative indications in lenses made for full format.

As DOF scales are just an approximation I usually consider with the M8 the next wider stop of the DOF scale in the lenses..... for example that is how I work with the CV 15mm.

 

I may have missunderstood the Leica text...... what do you think about it?

 

Regards

[Shootist]

My take on this is the sensor captures less, smaller size, then what the lens can produce. So the DOF is about the same as it would be on a film camera.

The smaller sensor really doesn't increase the focal length of the lens, a 50mm lens will always be a 50mm lens. It only becomes apparent when the image is printed to the same size print as the same shot taken from the same location with a film camera. The film negative will include what the digital sensor cropped out and when both of these are printed to say a 8x10 the digital will look like it was taken with a longer lens. But it wasn't.

Other will say I'm wrong, OK. Whatever it is it works for me.

[Tom_W]

DOF is not altered by the crop factor (smaller size) of a sensor the DOF on your lens markings is correct.

 

T

[mjh]

I may have missunderstood the Leica text...... what do you think about it?

This has been debated several times, I believe. Leica’s statement is certainly at odds with the standard formulas for calculating depth of field. DOF depends on the circle of confusion, which is usually derived from the image diagonal. In the case of the M8, this is 32.4 mm, compared to the 43.3 mm of 35 mm film, and thus the CoC is smaller, resulting in an increased depth of field.

[Shootist]

This has been debated several times, I believe. Leica’s statement is certainly at odds with the standard formulas for calculating depth of field. DOF depends on the circle of confusion, which is usually derived from the image diagonal. In the case of the M8, this is 32.4 mm, compared to the 43.3 mm of 35 mm film, and thus the CoC is smaller, resulting in an increased depth of field.

mjh I tend to agree with you but other on this forum state otherwise.

They say something like "Since a 50mm lens becomes a 67mm lens on the M8 the DOF will be less because of the increase of focal length".

Looking at it this way they could be right but if you took a 35mm negative and cropped out the edges so that it only contained 75% of the original image (that is the 1.33 crop factor) would the DOF be different, NO.

 

I agree with Leica.

[marknorton]

There was a very interesting article on LFI last year, one of those you need to read a couple of times to understand.

 

I think you can only really compare DoF if both cameras are set up to record the same thing, and then enlarged by whatever factor their crop factor demands to the same size image.

 

I need to dig out that article again to refresh my memory...

[msendin]

Advertisement (gone after registration)

Yes Shootist but that way if you have a print the size of the 35mm. format and make another print 0,75 smaller (same size as the M8 sensor) your DOF will have increased.

I think DOF has ALWAYS to be related to the same enlargement or it is not a valid for anything. I just find Leica statement very confusing and if the english translation from german is accurate and my understanding of english is not confusing myself (which often does) I believe the Leica statement, as written, is even incorrect. But of course I do not dare to question Leica on a technical statement, so all I can say is that I do not understand the meaning of their sentence.

[mjh]

mjh I tend to agree with you but other on this forum state otherwise.

Sorry for the misleading post. I had stated that the CoC is smaller because of the smaller image size, which is correct, but of course a smaller CoC results in a narrower DOF – and that’s what I had meant to say, but for whatever reason, I wrote just the opposite.

 

They say something like "Since a 50mm lens becomes a 67mm lens on the M8 the DOF will be less because of the increase of focal length".

Now that line of reasoning is obviously erroneous as there is no increase in focal length. A given lens has a certain focal length, and nothing can change that. It is the real focal length that enters in to the formulas for hyperfocal distance and depth of field; no crop factor gets applied. But the conclusion is correct anyway as the smaller image size corresponds to a smaller CoC, resulting in a narrower DOF.

 

Now one might ask: If the Summilux 1:1.4 50 mm on the M8 has the same field of view as a hypothetical 67 mm lens on an M7, would its DOF be similar to that of an 1:1.4 67 mm on an M7? But it isn’t. The formula for he hyperfocal length on which the DOF calculations are based is H = f^2 / (N * C), where f is the focal length, N the f-number, and C the diameter of the CoC. The focal length is squared and is thus the stronger influence compared to the CoC. For that reason, the DOF you get with the Summilux 50 mm on the M8 is greater than the DOF one would get with the hypothetical 67 mm lens on the M7 (at the same F-number).

[jaapv]

Basically DOF has nothing to do with focal length. The only thing that counts is angular enlargement. The only reason a lens with a shorter focal length appears to have a deeper DOF is because it enlarges more. You can only judge DOF if you enlarge the same subject to the same size in the final print and view it at the same distance.

If you take everything into account a given lens on the M8 will have about 2/3 of a stop less DOF than on film. If you change focal length one step down to compensate your field of view for the sensor crop the whole equation changes. If you are interested you can work it out yourself on Hyperfocal Distance and Depth of Field Calculator - DOFMaster . Take into account that in general DOF appears to be about 1/3 of a stop less deep on a sensor because of the difference between a film and a sensor in recording the image.

[mike prevette]

Calculating the CoF using the old film equations doesn't tell the whole story here. A sensor is a grid of pixels that have a linear structure, and gaps between the sensitized areas. Film is a random pattern of sensitized particles covering the entire surface of the filmbase. My point is that for the same size film area, and the same amount of magnification, sensors will display different DoF characteristics. Subtle differences in falloff etc. Making digital MORE critical than film. I rarely use DoF tables on lenses, but if I do I conservatively go with a full stop difference over film. So If I'm shooting at f8, I look at the f5.6 markings.

 

_mike

[ho_co]

I think the seminal article on the topic was Olaf Stefanus's report on the presentation by Leica's head lens designer, Peter Karbe, which was run under the title "Form Follows Format" in LFI 3/2006, pp 40-47.

 

See also the article on the Mamiya ZD by Efrain Garcia and Rubén Osuna at Mamiya ZD A 35mm-like Medium Format Camera. They cover a lot of good digital basics.

 

Another interesting study on the matter is by Charles Sidney Johnson at Luminous landscape. Johnson says provocatively: "We can take an essentially identical photograph with any sensor size by scaling the focal length, the f-number, and the ISO sensitivity." Nathan Myhrvold joins the discussion with some equally interesting arguments.

 

On a related topic, Gary Ferguson raises the question whether we should even try to think in traditional terms when we are talking digital in his two articles beginning at Digital Focusing Part One.

 

To generate the "equivalence" values for both field of view and depth of field, simply apply the crop factor (4/3 in the case of the M8) to both focal length AND aperture:

A 50/2 wide open behaves as would a 67/2.7 on a 24x36 frame;

and when stopped down to f/5.6 it becomes the equivalent of a 67/7.5.

 

[As I understand it, this rule doesn't quite hold with very closely spaced pixels (as in the Digilux 2 and most digicams), because the circle of confusion must be larger than three pixels in order to be resolved at all.]

 

In short, Mario, the M8 manual seems to be in error in the cited point. And your technique of using the engraved marking for the next larger aperture is good practice.

 

--HC

[Stan Yoder]

Text bellow is from the instructions booklet of the 50m lux asph, when talking on page 12 of the 1.33 factor of lenses used on the M8.

 

" ....... This has the respective effects on the perspective,......."

 

As to perspective, subject distance is the only real determinant. If images of the same subject from the same distance are made with lenses of various focal lengths, and the prints are cropped to show the same field of view (that of the longest F.L. used), and all else is equal, the perspective will be identical.

 

So, if you shoot with an M8 and, say, an M7 from the same distance and with the same lens, and crop the M7 print to match the M8 print, apparent perspective will be the same.

 

Or am I full of it?

[adan]

No, Stan. You got it. "Perspective" depends solely on the "point of view" of the lens - the precise location in space the light rays enter the lens. So long as that doesn't change, perspective will be the same regardless of focal length, sensor/film size, enlargement etc.

 

DOF is dependent on enlargement, and unless one includes final print enlargement in the calculations, DOF is "undefined"*. A given photograph, with a given lens, at a given aperture, may appear totally sharp in a 4" x 6" enlargement, yet some things in the picture may become more noticably unsharp in a 20" x 30" enlargement.

 

*Choosing a CoC dimension is the usual way of accounting for final print size in DOF calculations.

 

Since an original image of dimensions 18mm x 27mm (M8) requires MORE enlargement for any final print size than a 24mm x 36mm original (film M), the general effect will be for LESS final DOF in a print with the same LENS (not same FOV, same LENS) at any given aperture and subject distance.

 

If one tries to accomodate the crop factor to get a wider field of view, either by using a shorter lens, or by moving back further from the subject, one has changed values in the DOF equation, and thus no valid comparison can be made. However, the apparent effect in those cases will be MORE apparent DOF.

 

Read those last 2 paragraphs several times, and one can see where confusion sometimes sets in.

 

Net result: I disagree with Leica on both counts:

 

1) Perspective of a lens does not change because of the M8's crop factor, only if the photographer and lens change position. Some witll say, "Yes, but to keep the same cropping with an M8, I have to move back" - and I will say "Yes, but it is YOUR decision to move, not the crop factor itself, that causes the change in perspective."

 

2) Final DOF as viewed in a print (the only place it matters) WILL be less with an M8 shot than with a 24 x 36 shot, all other things (aperture, shooting position) being equal, due to the extra enlargement needed - therefore the DOF scales are no longer trustworthy (if they ever were - a different debate). What the lens projects on the sensor or film does not change, but what we see as the final result once it is blown up to different amounts is - different.

 

Again, some will say "But to match the field of view of a 28mm lens with the M8, I need to use a 21, and so I get more DOF." Which is true, but once you are using two different lenses, with two different DOF scales, you are no longer talking about THE (single) DOF scale on the original 28.

[luigi bertolotti]

Text bellow is from the instructions booklet of the 50m lux asph, when talking on page 12 of the 1.33 factor of lenses used on the M8.

 

" Therefore, when used on the Leica M8, these lenses have angles of view corresponding to lenses with focal lenghts that are longer by a factor of 1.33. This has the respective effects on the perspective, but not on their depth of field, which, with the Leica M8, can also be read directly off the lense (see depth of field scale,p.14) "

 

As far as I understand DOF depends ONLY on "amplification", or enlargement, of the image in relation to the real size of the object.

A smaller sensor is a larger enlargement so DOF is smaller than the orientative indications in lenses made for full format.

As DOF scales are just an approximation I usually consider with the M8 the next wider stop of the DOF scale in the lenses..... for example that is how I work with the CV 15mm.

 

I may have missunderstood the Leica text...... what do you think about it?

 

Regards

 

These arguments, expecially the DOF issue, have been scrutinized and discussed a lot in the forum... you can find very well technical explanations navigating with keywords like "DOF" or so.

My opinion is that in this topic the manual you quote is misleading / unprecise on both of its statements :

 

- "...respective effect on perspective..." : perspective depends only on distance from the subject... OK, to frame in a date way a certain scene, with M8/50 mm you must position yourself to a higher distance than with a M6/50 mm... perspective changes... but the sentence in the manual is unclear in its form.

 

- "... not on their depth of field..." : even if NOT considering the complex issue of "how is DOF on digital sensors" the sentence is someway contradictory with the assumptions of the previous one : if you have "effect on perspective" it means you have positioned yourself to a different distance from the subject... and this ALTERS the absolute value of DOF.

By another point of view Leica statement is not erroneous: forget sensor and imagine your M8 is simply a 35mm film camera with a mask that exposes only a 18x27 area on each 24x36 neg; next to you there is a people with a M6 with the same lens as yours: to frame in the same way the same scene, you have to position yourself farther from the subject : supposed both are focusing on the same target, you have an HIGHER absolute value of DOF (for example, say, 30 cm in front of the target, the M6 shooter has 24 cm or so...). But, to achieve the same PRINT, you'll have to enlarge MORE your negative , the famous CoC is DIFFERENT, and this eliminates your "DOF advantage"..

[archi4]

The projection on the film plane from a given lens is always the same regardless of the size of the film plane. The size of the film or sensor just determines the crop or what part of the projected image is used. DOF and perspective remain the same (perspective only changes with point of view (where you and the lens are) You can verify this by cropping an M8 image taken with a given lens in post processing, or take the same lens onfilm, scan and then crop to the same size as the M8 image.

 

Sorry Andy, I wrote this before reading your post which explains everything even better!

[jaapv]

The projection on the film plane from a given lens is always the same regardless of the size of the film plane. The size of the film or sensor just determines the crop or what part of the projected image is used. DOF and perspective remain the same (perspective only changes with point of view (where you and the lens are) You can verify this by cropping an M8 image taken with a given lens in post processing, or take the same lens onfilm, scan and then crop to the same size as the M8 image.

 

Sorry Andy, I wrote this before reading your post which explains everything even better!

This is correct if you assume one makes a corresponding size print so a smaller one by 30% from the 1.3 cropped sensor. Obviously this is not the case, thus the image will be enlarged by 30%, decreasing DOF slightly over half a stop.

That is why the circle of confusion on 24x36 is standarized at 0.03 mm and on a 17x24 sensor to 0.023 mm. This standarisation is just a figure anyway, as it assumes a relatively small print.

[zapp]

You are so picky. If you read off the DOF from the lens you are looking at the blurr circle used for film which is 0.03 mm. Now look at the pixel pitch of the M8. If you allow 1/2 pixel of blurr to be sharp, you will realize that the whole engraved values are not suitable at all.

Forget the 1.3 to 1 difference of enlargement this is peanuts compared to a 10 to 1 difference of the blurr circle.

[jaapv]

You are correct in a certain way, that is the zone of maximum sharpness in the plane of focus. However, DOF is something else. That is the zone of adequate sharpness that the human eye can distinguish in the final print at normal viewing distance.

Film can resolve far higher than 0.03 mm as well, up to 40 lp/mm at 50% contrast.

[Shootist]

A very interesting discussion and well layed out. All in all I will still use the scale on the lens for my general reference for shots that I want low DOF or high DOF, as I think we all do.

Thank you all for sharing your knowledge on the subject. I feel smarter already.

[Advertisement]

Hallo,

Look here: Leica M

[jaapv]

Actually I don't use the scale at all. WhenI want narrow DOF (most of the time) I use the lens wide open, when I want selective focus over a wider range I set the lens to f 5.6 or f8.0, in the rare cases I want everything in focus I use a WA at f 16. I focus exactly on the subject and let the DOF fall where it may...