Leica Q -general-

[lct]

Advertisement (gone after registration)

In that case the formulas for calculating depth of field would be wrong. In fact our very concept of depth of field would be wrong. But given that it has been around for many decades, I don’t think it is. Which implies that yes, digital zoom does change DoF.

 

I'm no techie at all but DoF formulas are based on actual focal lengths aren't they? Just look at the background (no tripod here sorry). 

 

Elmarit 90/2.8, f/2.8:

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!

 

Elmarit 28/2.8 asph, f/2.8, cropped to 90mm FoV:

[MarkP]

Exactly,

 

but obviously when tis a  'digital' crop the 28mm photo (or equivalent within the camera using the same FL lens the DOF is still the same because the only thing altered is the amount of the image that is finally displayed.

[Advertisement]

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

[mjh]

I'm no techie at all but DoF formulas are based on actual focal lengths aren't they?

They are. But they are also based on the circle of confusion which is expressed as some fraction of the image diagonal – 1/1500 or whatever matches your idea of “sufficiently sharp (if barely)”. Cropping reduces the effective image diagonal and thus the circle of confusion, resulting in a change in DoF.

[jaapv]

No. DOF is not determined by focal length, but by magnification. The only reason a longer lens appears to have a more shallow DOF is because it magnifies more. If you equalize the magnification by cropping DOF will be virtually (there are some minor aspects, but they do not really come in) the same. Your background on your 28 lens is not sharp. Only the focal plane is sharp. You do not see the unsharpness because of the lack of resolution of your eye. Magnify it and it becomes apparent.

Please read this article by H.H.Nasse from Zeiss. It explains things better than I can.
http://www.zeiss.com/content/dam/Photography/new/pdf/en/cln_archiv/cln35_en_web_special_bokeh.pdf

Quote for an explanation why DOF is not a function of focal length:
 

The geometric explanation for the rule that the depth of field is not dependent on the focal length for a given size of the object field: with the same f-number, the size of the entrance pupils is proportional to the focal length and focusing distance. The light cones, and therefore also the circles of confusion, are always the same.

 

and:

 

 

The depth of field (almost) does not depend on the focal length at all but rather on the imaging scale, and we can understand that as follows:
A focal length that is twice as long creates an image of the same size from an approximately doubled distance, and with the same f-number its entrance pupil diameter is twice as large. Because of the increased focusing distance the object side cone of light is nevertheless the same. As a result, the “object side circles of confusion” are also the same.

 

 

 

However, as focal length is an expression of the magnification of a lens, it is perfectly usable for calculating DOF. There are formulas that do not use the focal length, btw.

[IkarusJohn]

They are. But they are also based on the circle of confusion which is expressed as some fraction of the image diagonal – 1/1500 or whatever matches your idea of “sufficiently sharp (if barely)”. Cropping reduces the effective image diagonal and thus the circle of confusion, resulting in a change in DoF.

 

That's true, Michael, but only half the explanation.  Circle of confusion is specific to the lens, not the sensor - changing the sensor will only have an effect if the same field of view is maintained.

 

Let me put this another way - I think we all accept that if you crop an image to give the same field of view (as LCT shows above) that does not change the depth of field of the original image.  That is just silly, and easily disproved.

 

However, if you maintain the same field of view, but use cameras of different sensor sizes, then you either need to move (changing the relative distances of the subjects from the sensor) or you need to change lenses, fundamentally changing the optical characteristics of the image.

 

For example, in the scenario above, if you have a full frame M camera with a 50 Summilux and you take a given image at f/1.4 focused on a subject at 3 metres, you will get a depth of field of 280mm (130mm front & 150mm rear).  In order to maintain the field of view with micro 4/3, you will need to change to a 25mm focal length; still at f/1.4 and focussed at 3 metres, you will get a depth of field of 600mm (330mm rear & 270mm front).  Now, if you switch to a 6x9 field camera, you will need a 125mm lens for the equivalent 50mm field of view, which results in a depth of field of 100mm (50mm behind and 50mm in front; that is rounded).

 

The result is there for all to see, exactly as expected.  Alternatively, you could have a 50mm lens on all three cameras - this would give the MFT camera a field of view equivalent to a 100mm lens in 35mm terms, and the field camera an equivalent field of view of a 20mm lens in 35mm terms.  I can't readily do the calculations for how far you'd need to move the MFT camera back, but once you achieved the same field of view, the relative separation of the subjects to the sensor and the natural foreshortening of what has become a tele lens would result in almost everything in the image being in focus.  

 

Conversely, the field camera would need to move so close to achieve the same field of view that the change in relative distances from the sensor would give a very shallow depth of field - I can calculate these if someone gives me the distances you'd need to travel to achieve the same field of view - the effects are not linear, obviously.

 

What we can say, comfortably, is that if you take an image on a given sensor and crop it in LightRoom it will have absolutely no effect on depth of field, unless you then enlarge it ...

[jaapv]

You will have to do, so because you'll end up with a smaller print otherwise.

 

And then we enter the realm of enlarging by interpolation, which changes the whole game.

Which LCT did in his examples. He - or rather LR for him- cropped and resized, which falsifies the result.

[IkarusJohn]

Advertisement (gone after registration)

Jaap's quote is useful, if read with care:

 

The geometric explanation for the rule that the depth of field is not dependent on the focal length for a given size of the object field: with the same f-number, the size of the entrance pupil is proportional to the focal length and focusing distance. The light cones, and therefore also the circles of confusion, are always the same.

 

Remember, focal length is the distance between the focal point (i.e., the focal plane, which is hopefully where your sensor is) and the pinhole point at which the image passes through your lens (i.e., the theoretical central point of the lens where all parts of the image converge).

 

What is happening here is that too many variables are being brought into play - if the object distance remains the same, but the field of view changes by changing the focal length of the lens (and the image is then cropped to produce the same field of view in the image), then of course the depth of field changes - LCT has shown that very simply above.

Edited June 7, 2015 by IkarusJohn

[IkarusJohn]

Which you will have to do, because you'll end up with a smaller print otherwise.

 

And then we enter the realm of enlarging by interpolation, which changes the whole game.

 

I don't think that really helps, which is why I discounted it.

[lct]

Focal length and CoC are both factors of DoF and cropping will change the CoC's value to a certain extent but not to the point where a sharper background can become blurred by miracle sort of. Even cropped to 90mm FoV, a 28 will remain a wide lens basically such that it will be more difficult to isolate subject matters from background and foreground with it than it can be with a true 90mm lens like in my snaps above. 

[mjh]

You will have to do, so because you'll end up with a smaller print otherwise.

And if it was smaller we would get closer and the result would be the same. The point is that the way we are viewing images has nothing to do with sensor size so ultimately a smaller sensor requires a larger magnification.

[mjh]

Focal length and CoC are both factors of DoF and cropping will change the CoC's value to a certain extent but not to the point where a sharper background can become blurred by miracle sort of.

The blur will be the same but a shrinking circle of confusion means that some amount of blur that used to be acceptable will cease to be.

[Jeff S]

 

I'm sure someone far smarter than me will explain...

You mean someone 'smarter than I', where 'I' is the the subject of the elliptical clause …smarter 'than I am….'.    Elliptical clause has nothing to do with circle of confusion.

 

Jeff

[lct]

 

Which LCT did in his examples. He - or rather LR for him- cropped and resized, which falsifies the result.

 

Just cropping with my old C1 v4. One may well resize as he or she likes, the blurred doors in the background will remain blurred and the sharper ones will remain sharper.

[regular]

Oop, I though it was the 'new Q' thread... but it looks like the remedial class for people who forgot to do their homework of optics

[Guest]

lct's closer to truth imo:

film/chip size matters.

 

One can enlarge a small one from aerial/Technical Pan or from the latest sub-24X36 chip taken with the latest high-resolution lens

to wall size

and there will be more DOF than in a a photo enlarged to the same size from a medium-large format film or the biggest commercial chips.

 

I think this theoretical question was answered in Wetzlar in practice almost a century ago, with the help of Agfa and Perutz.

[mod2001]

I don't know exactly, but I guess if you use the digital zoom to 35mm, the crop should have about 18MP, which is still enough flesh isn't it?

 

Not exactly, you have to divide the focal length in square (28mm/35mm)² and to multiply with the resolution of the sensor, in this case then 0,8²x24 = 0,64x24 = 15,36MP, crop to 50mm will be then 7,52MP

 

So let's hope Leica goes the GR road with the Q, really like the crop abilities of the Ricoh (35mm crop has there 10,32MP)

 

Yogi

Edited June 7, 2015 by mod2001

[viramati]

What would be interesting is if they were to implement the 35 crop area just with crop frame-lines sort of like a Leica M so that you get to see what is outside of the frame-lines as it comes into view. Would sort of keep the traditional RF shooting experience. I say this seeing that the specs seem to show a vey high resolution EVF which I also presume will be quite large.

[eckart]

Not exactly, you have to divide the focal length in square (28mm/35mm)² and to multiply with the resolution of the sensor, in this case then 0,8²x24 = 0,64x24 = 15,36MP, crop to 50mm will be then 7,52MP

 

So let's hope Leica goes the GR road with the Q, really like the crop abilities of the Ricoh (35mm crop has there 10,32MP)

 

Yogi

 

Thanks and well, you might be right, I really don't know.

I just did it after the "M8-formular", in which the 28mm changes into a 35 mm by a factor of 1:1,33, that means 24 MP/1.33~ 18MP

But that's how we call it in German "over the thumb"

Beside all theoretics we will see the real thing next Thursday, I am looking forward to it

[grillo]

Thanks and well, you might be right, I really don't know.

I just did it after the "M8-formular", in which the 28mm changes into a 35 mm by a factor of 1:1,33, that means 24 MP/1.33~ 18MP

But that's how we call it in German "over the thumb"

Beside all theoretics we will see the real thing next Thursday, I am looking forward to it

But you forgot that M8 is 10 MP and M9 is 18 MP, with the same pixel pitch, so the conversion factor is 1:1.8

[Advertisement]

Hallo,

Look here: Leica Q

[pop]

But you forgot that M8 is 10 MP and M9 is 18 MP, with the same pixel pitch, so the conversion factor is 1:1.8

Beware: one (the pixel count) measures an area while the other (the crop factor) refers to a length: the number of pixels vs. the length of the diagonal of the frame. Increasing the diagonal by 40% increases the area by 100%.