A difficult question - shooting square

[pico]

Advertisement (gone after registration)

 I never understand this obsession wth square, either.

 

We can crop the square anywhere.

[Advertisement]

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

[mikemgb]

Why buy Leica if you want it to be Rolleiflex? I never understand this obsession wth square, either. All your have to do with square, is take it in the middle and decide later how to crop it on print. Portrait or Landscape .

 

My first two cameras were both square format, I just can't help liking it. I do quite often do as you suggest, crop to square later.

[wolfloid]

Why buy Leica if you want it to be Rolleiflex? I never understand this obsession wth square, either. All your have to do with square, is take it in the middle and decide later how to crop it on print. Portrait or Landscape .

I think you are making some assumptions here.

 

a) I don't think it is an 'obsession'.

 

Of course I can crop after the fact, I often do. The question is how to get the same field of view and perspective and also be able to achieve the sort of focus separation I have become used having.

 

c) I like the rendering of the Rolleiflex, particularly in its rendering of dof, and in its tonality, and would like to be able to approximate that. Approximate, not make it the same.

 

Maybe if this doesn't interest you, you should just ignore the post.

Edited June 4, 2017 by wolfloid

[pico]

Of course I can crop after the fact, I often do. The question is how to get the same field of view and perspective and also be able to achieve the sort of focus separation I have become used having.

 

c) I like the rendering of the Rolleiflex, particularly in its rendering of dof, and in its tonality, and would like to be able to approximate that. Approximate, not make it the same.

 

Stop down one more for the Leica than you must on the Rolleiflex. Approximate is correct: you will not get the tonal gradation with the Leica that you appreciate with the Rollei. Printing smaller helps.

 

I am particularly fond of small prints.

.

Edited June 4, 2017 by pico

[jaapv]

Hello Luigi,

 

What we are talking about is NOT in the "eye of the beholder".

 

What we are looking at is something concrete & measurable.

 

IF we are considering depth of field as defined by "circles of confusion", then:

 

what I wrote above is accurate. Because:

 

The original question was:

 

If a person stays at 1 spot & takes 2 pictures of the same quetzal with 2 lenses of 2 different focal lengths which are both set at the same aperture:

 

And then the image taken with the shorter lens is enlarged so that the 2 images are congruent when 1 image is put on top of the other:

 

Then the depth of field will be the same in both images.

 

This is math.

 

NOT the eye of the beholder.

 

Because: When the smaller image with more depth of field is enlarged to the size of the larger image:

 

What were previously point images or images of 1/30mm or less become larger discs & are no longer seen as being in focus.

 

All aspects of all things above being equal: The only difference between the 2 images will be that the photo taken with the longer lens will be of a better quality because it will contain more pieces of information per square millimeter in the finished print.

 

Best Regards,

 

Michael

The point is, Michael, that your calculations may be precise,  and I agree with your argument above, but they are based on the quicksand of assumptions as far as DOF in general is concerned.

1. The accepted value of COC was set in the 1930ies and was calculated at the time from thick, low-resolving film, the normal print size of 6x9 cm and the average resolving power of the human eye.

Nowadays we have precise sensors and large prints. The eye of the photographer is still as variable as it used to be. A realistic COC today would be about a quarter of the official size.

2. All calculations fail to take the subject matter into account. A finely structured high-contrast subject will have a far more narrow DOF than a misty landscape.

 

The only realistic definition of DOF is: the zone of acceptable sharpness for the individual photographer for a specific image. Calculations can help, but will never give a precise result.

[pico]

The only realistic definition of DOF is: the zone of acceptable sharpness for the individual photographer for a specific image. Calculations can help, but will never give a precise result.

 

That is a reasonable point particularly for us Olde Phartes who lived through transitions. A point:  in our youth we were largely concerned with what prints presented, what they looked like. Therein is our habitual perception to seek quality, be it abstract or very rarely constrained to irrelevant scientific metrics. Abstract won. Today we have a new constituency who live by a computer monitor and they have modest scientific metrics/arithmetic entirely separate from the experience of prints. They do not 'get it'.

Edited June 5, 2017 by pico

[harmen]

Advertisement (gone after registration)

I had a go at calculating the size that a single dot becomes when it is not in focus. When you take for example a dot at 1m distance, while focusing at infinity, it becomes a circle. That's because it focuses behind the film rather than exactly on it, as light coming from infinity would.

 

Comparing a 35mm focal length and a 80mm focal length, the size of this circle is significantly different. Assuming that the image of the film behind the 35mm lens gets enlarged more than the one behind the 80mm lens, still leaves the one from the 80mm about twice as blurry.

 

For 35mm on 24x24 to be equally blurry as 80mm f/2.8 on 56x56, the 35mm lens needs to be f/1.25

 

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!

[jaapv]

Thank you for giving a mathematical basis to the difference in bokeh between lenses of different focal length, which of course gives a different impression of DOF.

The location of the exit pupil of any given lens will give a different result as well.

 

https://www.l-camera-forum.com/topic/273058-a-difficult-question-shooting-square/?p=3288091

[Exodies]

Why buy Leica if you want it to be Rolleiflex? I never understand this obsession wth square, either. All your have to do with square, is take it in the middle and decide later how to crop it on print. Portrait or Landscape .

  

My first two cameras were both square format, I just can't help liking it. I do quite often do as you suggest, crop to square later.

Ko.Fe. is talking about cropping squares to rectangles. Others who quoted him (not just mikemgb) talk about cropping rectangles to squares. Other aspects of this tread are similarly murky - relating viewing distance to print size for example.

I'm not happy about comparing dof in terms of stops. That assumes a lens behaves the same at different apertures. Diffraction, focus shift ensure this isn't the case.

This is a discrete rather than analogue question because even with film the coc is fixed in the same way that digitisation fixes things - as you enlarge an image what is below coc or above coc changes.

[01af]

For 35 mm on 24 × 24 to be equally blurry as 80 mm f/2.8 on 56 × 56, the 35 mm lens needs to be f/1.25.

That's correct—approximately—but for practical intents and purposes, a 35 mm 1:1.4 will be close enough.

 

Moreover, always remember to use an aperture two or 2.5 stops wider than you would on the Rolleiflex if you want to get the same depth-of-field. So when you'd shot at, say, f/8 with the Rolleiflex, use f/3.5 or f/4 on the 35-mm-format camera.

 

 

The location of the exit pupil of any given lens will give a different result as well.

With the lens focused at infinity, the exit pupil's location doesn't matter. At distances shorter than infinity, the exit pupil's location does make in increasing difference indeed—but at rangefinder's distances (i. e. 0.7 m and longer) the difference will be minuscule so it can be neglected. Only at close-up and macro ranges, a non-unit pupil magnification must be taken into consideration.

 

 

I'm not happy about comparing depth-of-field in terms of stops. That assumes a lens behaves the same at different apertures.

That's true, Also, it assumes different lenses behave the same at equivalent apertures—usually they don't. Still, calculating equivalent f-stops is the best you can do (aside from doing practical field tests with the actual lenses involved). At least, it will get you in the ballpark.

[Michael Geschlecht]

Hello Everybody,

 

There appears to be some degree of "apples" as opposed to "bathtubs" types of comparisons going on here:

 

Instead of the "bathtubs" as opposed to other "bathtubs" situation that was being discussed at the beginning of this Thread.

 

The second set of relationships (Bathtubs as opposed to other Bathtubs.) are the parameters that I have been writing about here:

If a person, standing in a given spot: Has a lens WITH A GIVEN ANGLE OF COVERAGE, on camera "A", that is being used to focus on & capture a specific subject (ie: A quetzal.), within a given scene. And that lens & camera have an image capture surface measuring 56mm X 56mm: Focal length approximately 80mm:

 

And that person, standing in that same spot: Has a second lens WITH THAT SAME ANGLE OF COVERAGE, on camera "B", that is being used to focus on & capture that same quetzal, within that same scene. And that lens & camera have an image capture surface of 24mm X 24mm: Focal length approximately 35mm:

 

AND:

If the 2 separate captured images were printed on 2 separate surfaces which are the same dimensions in millimeters:

 

And: If the final image of the quetzal would be the same height in millimeters in each finished photo:

 

What F stop would have to be set on each lens so that the DEPTH OF FIELD ON EITHER SIDE of the quetzal (Which is the plane of focus for both photos.):

 

Would be the same?

 

The answer is : The SAME lens opening (F stop.).

 

Altho, if that "same" F stop was the largest F stop that might be somewhat problematical under certain circumstance because of different manufacturers standards concerning their largest apertures.

 

Better to use a medium aperture like F8.

 

Best Regards,

 

Michael

[01af]

What f-stop would have to be set on each lens so that the DEPTH OF FIELD ON EITHER SIDE of the [...] plane of focus for both photos. Would be the same? The answer is: The SAME lens opening (f-stop).

 

 

 

Yes, the absolute lens aperture must be the same—which for different focal lengths means different f-stops (= relative apertures).

[Michael Geschlecht]

Hello 01AF,

 

Thank you for your reply.

 

In order for the image captured on the smaller sized image capture surface to be recorded as the same size in millimeters on the same size in millimeters final photo:

Then, that image needs to be enlarged 56/24 times more = 2.33 times more approximately:

To be the same size:

 

This means an approximately 2.33 times increase in what had been the size of the original acceptable limiting circle of confusion:

 

This means that a 10X enlargement of the quetzal from the 56mm X 56mm capture surface would need an approximately 23.3X enlargement from the 24mm X 24mm image capture surface to produce the same sized quetzal in both photos.

 

Which means that: What had appeared previously as an in focus point measuring 1/30mm in diameter would now be seen as an out of focus disc measuring approximately 1/13mm: This would no longer appear as a point: But, would instead, be perceived as a disc or blur:

Thereby reducing the final print's depth of field in the image made from the 24mm X 24mm image capture surface from what it would have been if that image capture surface had only been enlarged 10X.

 

Best Regards,

 

Michael

Edited June 5, 2017 by Michael Geschlecht

[01af]

And your point is ...?

[jaapv]

That you are right

[Michael Geschlecht]

Hello 01AF,

 

The 'point" is, that:

It would appear to me that if you took a photo of a quetzal, 3.5 meters away from you, with an approximately 80mm lens, set at F8, which was on a camera capturing that image on a 56mm X 56mm image capture surface:

And then took a second photo of the same quetzal, at the same distance away, from the same point, using an approximately 35mm lens, set at F8, which was on a camera capturing that image on a 24mm X 24mm image capture surface:

 

And the went to print photos from the respective image capture surfaces:

On the same sized piece of printing surface:

So that the 2 different quetzals were the same size in millimeters on the respective finished photos:

Then, the 2 respective finished photos would have the same depth of field:

Which is the answer to the question asked at the beginning of this Thread.

 

Best Regards,

 

I always appreciate your insight & knowledge.

 

Michael 

Edited June 5, 2017 by Michael Geschlecht

[luigi bertolotti]

I had a go at calculating the size that a single dot becomes when it is not in focus. When you take for example a dot at 1m distance, while focusing at infinity, it becomes a circle. That's because it focuses behind the film rather than exactly on it, as light coming from infinity would.

 

Comparing a 35mm focal length and a 80mm focal length, the size of this circle is significantly different. Assuming that the image of the film behind the 35mm lens gets enlarged more than the one behind the 80mm lens, still leaves the one from the 80mm about twice as blurry.

 

For 35mm on 24x24 to be equally blurry as 80mm f/2.8 on 56x56, the 35mm lens needs to be f/1.25

 

IMG_2431.JPG

Thanks... I took time to interpret your writing... but your final result seems to confirm my idea : the added blur due to greater enlargement  (2,33 ratio i.e the 1/30 vs. 1/13  quoted by Michael) does not "overtakes" completely the difference due to longer focal... --> print from Rolleiflex will be perceived with "less DOF"... though keeping my idea that to say "twice as blurry" is a phrase with limited significance.

 

 (01AF... Yes, the absolute lens aperture must be the same—which for different focal lengths means different f-stops (= relative apertures).... you're surely right, but please, let's not introduce another variant in terms... I think that anyone, in this discussion, has referred to a 80mm lens with its f/stop index set to f 2,8 or 8 and to a 35mm lens with its f/stop index set to f 2,8 or 8...  )

Edited June 5, 2017 by luigi bertolotti

[adan]

I've got to throw a different fly into this ointment.

 

Even if you can get approximately equal lack of DoF (nice soft blurs in the background) with the 35 f/1.4 ASPH to equal the Rollei 80, on their appropriate films or sensors - you will not get as much sharpness in the sharp areas from the 35 f/1.4 @ f/1.4.

 

So it still won't look like your Rollei images, because what you have in focus won't have medium-format sharpness.

 

The 35 f/1.4 ASPH is a great (even world-leading) lens for its specs - aperture and field of view. But at f/1.4 it is not all that sharp. Just as good as or better than other lenses of the same spec. And its greatest strength is removing aberrations in the corners (which you'll be throwing away when you crop square).

 

Believe me - I tried exactly the same thing (shoot 35 ASPH @ f/1.4, crop square - to look like a Rollei or Hasselblad 80mm). Doesn't even come close to a Rollei 80mm Planar/Xenon @ f/2.8 on its larger format. Unless you print really small (6x6 inches/15x15cm) to hide the fuzzyness of the "sharpest" areas.

 

In absolute resolution, the 35 ASPH @ f/1.4 is about 25% sharper than the Rollei Planar @ f/2.8 (see mtf links below). But once you blow up your 24mm x 24mm capture to get the same print size as from the 56mm x 56mm capture of the Rollei, that is a 230% difference in magnification, which completely overwhelms a 25% improvement in original sharpness. More total mm blows away more lppm.

 

Put another way, the Rollei 80mm @ f/2.8 can achieve about 30 lppm @ 50% contrast - across 56mm. The Summilux @ f/1.4 can achieve about 40 lppm @ 50% contrast - across 24mm. The Rollei thus gives you 3360 lines of resolution across its whole image, while the Leica lens will give you only 1920 lines across its whole (cropped square) image.

 

Stopped down is a different story - but then you no longer have the nice blurry backgrounds of 6x6.

 

http://www.marcocavina.com/articoli_fotografici/Rolleiflex_vs_Hasselblad_2/35a.gif

 

https://cdn.photographylife.com/wp-content/uploads/2013/05/Leica-Summilux-M-35mm-f1.4-ASPH-MTF.jpg

Edited June 5, 2017 by adan

[Advertisement]

Hallo,

Look here: Leica M

[luigi bertolotti]

I've got to throw a different fly into this ointment.

 

Believe me - I tried exactly the same thing (shoot 35 ASPH @ f/1.4, crop square - to look like a Rollei or Hasselblad 80mm). Doesn't even come close to a Rollei 80mm Planar/Xenon @ f/2.8 on its larger format. Unless you print really small (6x6 inches/15x15cm) to hide the fuzzyness of the "sharpest" areas.

 

 

Interesting, Adan.. as still owner of an excellent (and unused for years...  ) Rollei 3,5f Planar TLR... I wonder, is your comparision made with Summilux 35 on film or on FF Digital ?  Printed in which format ?

[Michael Geschlecht]

Hello Everybody,

 

Part of the reason for what Andy is saying is that when you measure the actual information captured by the 80mm lens during the original picture taking & compare it to the amount of information captured by the 35mm lens, considering all other aspects the same, as per what was written above in many Posts:

You find that after you print the quetzal from the exposure done with the 80mm lens the same size in millimeters as the quetzal taken with the 35mm lens:

That: The photo of the quetzal done with the 80mm lens will be a better photo than the photo of the same quetzal done with the 35mm lens.

 

Because each square millimeter of that quetzal photographed with the 80mm lens will contain approximately 5 times the information that is in the equivalent photo done with the 35mm lens.

 

Best Regards,

 

Michael 

Edited June 5, 2017 by Michael Geschlecht