Depth of Field and "Focus by Wire" - a Thought

[xpatUSA]

I do more close-up photography than I do rolling landscapes. Thus depth of field, hereinafter called DOF, is of concern - but the calculation, or even just guessing is a pain and often results in re-takes. And, if the bird has flown, not quite so easy to re-shoot

Witness this shot, a failure as regards DOF:



The needed DOF for the nearer cicada husk is between 25-30mm. Going back indoors and using a DOF calculator told me I was too close (surprise) for the f-number I had selected. I could have gone back and forth a few more times but, although taking the time is no problem for a geezer, it was 98F and I do lack patience.

Enter the mighty Panasonic micro 4/3" camera with it's live view and focus-by-wire lenses. By going to manual focus, a 10X live image of the subject is shown on the LCD or the EVF - whichever you're using. It has just occurred to me that all I needed to do was move the camera and then check the sharpness of all parts of the cicada husk in manual focus mode. 'Cos it's live view, ennit? A little awkward on a tripod but not impossible.

It's all done right there in real time, with a high probability of success. Haven't tried it yet.

Is this something that "everybody knows", making this post a waste of your valuable time? Any comments welcome!

BTW, I don't intend to get into focus stacking, mostly the shots are for cataloging and/or web viewing.

[rpcrowe]

Hi Ted, I am not familiar with "focus by wire" terminology....

[chauncey]

Your question is somewhat confusing...you have the ability to shoot in LV, you have 10x zoom
to check your focus and, you have the DOF numbers from on-line calculators.
Sounds like a simple matter to dial in the correct f/stop, adjust exposure using the LV histogram and fire away.

[Shadowman]

I do more close-up photography than I do rolling landscapes. Thus depth of field, hereinafter called DOF, is of concern - but the calculation, or even just guessing is a pain and often results in re-takes. And, if the bird has flown, not quite so easy to re-shoot

Witness this shot, a failure as regards DOF:



The needed DOF for the nearer cicada husk is between 25-30mm. Going back indoors and using a DOF calculator told me I was too close (surprise) for the f-number I had selected. I could have gone back and forth a few more times but, although taking the time is no problem for a geezer, it was 98F and I do lack patience.

Enter the mighty Panasonic micro 4/3" camera with it's live view and focus-by-wire lenses. By going to manual focus, a 10X live image of the subject is shown on the LCD or the EVF - whichever you're using. It has just occurred to me that all I needed to do was move the camera and then check the sharpness of all parts of the cicada husk in manual focus mode. 'Cos it's live view, ennit? A little awkward on a tripod but not impossible.

It's all done right there in real time, with a high probability of success. Haven't tried it yet.

Is this something that "everybody knows", making this post a waste of your valuable time? Any comments welcome!

BTW, I don't intend to get into focus stacking, mostly the shots are for cataloging and/or web viewing.

Not everyone's favorite piece of equipment but you could use a focusing rail when on a tripod.

[TonyW]

I understand that "focus by wire" is in analogy with the term "drive by wire" in the context of cars.

I don't think Ted's technique would work if I understand correctly. When you look at the image through the viewfinder, you get the image as seen with a wide open lens. The lens closes down to the set f number only when the picture is being taken. (Correct me if I am wrong here.) What you see is not what you get. My camera has a button to press to see the image with the lens closed down but I find it hard to use without moving the camera and the resultant image in the viewfinder is not clear.

If you remember a few key results for your lens derived from the DOF calculator, then you can use rules such as the DOF being proportional to the f number and being proportional to the the distance from the object (as measured on the object) and do some mental calculations. However, if I try to do that the bee will have flown long before I have finished. My technique is usually to set the camera so that the f number is as high as possible for the conditions and hope.

I also have several questions about the calculation of DOF and other relationships. Firstly, I don't know where the focus distance is measured from. Is it the front of the lens? With a simple single element lens, the principles of physics and geometry are relatively simple. However, with a multi-element lens like all modern ones, it is not clear to me how these calculations should be done. For example, with my 100mm macro lens, the minimum focus distance is quoted as 0.3m and the magnification as 1:1. These numbers are not what you would get with a simple lens.

[Stagecoach]

Hi Ted,

Not sure what your terminology 'focus-by-wire' lenses refers to exactly.

Personally I have found in practice that although we have aids on the camera to help show DOF, DOF calculators easily available, measuring devices, lens specifications and possibly experience you can't beat taking a few shots at the same time at different apertures up to the max you can achieve with your lighting, downloading them and selecting which is acceptable for your purposes.

Remembering at ratios near 1:1 you are going to get roughly even DOF either side of your focus point.

[dje]

When you look at the image through the viewfinder, you get the image as seen with a wide open lens. The lens closes down to the set f number only when the picture is being taken. (Correct me if I am wrong here.) What you see is not what you get. My camera has a button to press to see the image with the lens closed down but I find it hard to use without moving the camera and the resultant image in the viewfinder is not clear.

Tony that's certainly the case with my Canon DSLR, not sure what happens with mirror-less cameras.

Ted a related issue is that what you see in LV mode is a simulated exposure based on the shutter speed and aperture settings.

Dave

[xpatUSA]

Some good points: and the one about the live view being with the lens at full aperture kills it right there

Focus by wire means that the focus ring is not mechanically coupled to the focusing mechanism. In manual focus mode, the ring instead sends pulses to the camera electronics which then turns the lens focusing motor. Takes a bit of getting used to.

Thanks all,

[Mike Buckley]

When using my Nikon D7000 and D5100 SLRs in Live View, the depth of field definitely changes in the display when I change the aperture. I don't know how that happens but it happens. I use that capability regularly before releasing the shutter to check depth of field of clear glass that I photograph in my makeshift studio. Perhaps the Live View displays an electronic simulation of the depth of field?

[pat3pee]

A coincidence as yesterday I wanted to try out the Live View focusing and found a cicada sitting still for a few seconds. Although handheld and not a real macro shot I was quite pleased with the DOF.
Cheers/Patrik

[RustBeltRaw]

Many DSLRs have a depth of field preview button which will stop down the lens from wide open (if that's what you have set) to give you a realistic preview (in live view or optical viewfinder) of your DoF and exposure level. Combining that with the live view's 10X zoom, and you can probably dial in focus very accurately. Maigc Lantern also allows an small, overlaid, high-zoom window so you can keep an eye on the focus point's sharpness while maintaining composition. Handy for moving while shooting video.

[xpatUSA]

Many DSLRs have a depth of field preview button which will stop down the lens from wide open (if that's what you have set) to give you a realistic preview (in live view or optical viewfinder) of your DoF and exposure level. Combining that with the live view's 10X zoom, and you can probably dial in focus very accurately

Thanks for the tip, Lex.

I found the preview button on my Panasonic GH1 and gave it a try. Unfortunately, the 10X view disappears as soon as the preview button is pressed - which is a great pity.

What is Magic Lantern? Is it firmware available for all cameras?

[Brev00]

The problem may be pov (point of view) as well as dof. Your sensor in the sample photo is parallel to the bug on the left. But, the two bugs are clearly not on a flat surface and the bug on the right is not parallel to your sensor. This will always introduce problems requiring more dof than one may not even have available at such a close distance. No solution may work perfectly, but I would move my camera so that I split the difference between the two bugs so that each is equally at a slight angle to the sensor. I would move the aperture to whatever you have decided is as stopped down as you can stand, close your eyes, and shoot. Alternatively, I would probably just take a picture of the bug on the left and call it a day. A third option I have tried in the past (once) with some success was moving back, shooting at a wider angle, and cropping later. I am not sure if seeing the action in live view is going to change much except the smaller sensor of the 4/3 camera might give you instantly some more dof.

[xpatUSA]

No solution may work perfectly, but I would move my camera so that I split the difference between the two bugs so that each is equally at a slight angle to the sensor. I would move the aperture to whatever you have decided is as stopped down as you can stand, close your eyes, and shoot. Alternatively, I would probably just take a picture of the bug on the left and call it a day. A third option I have tried in the past (once) with some success was moving back, shooting at a wider angle, and cropping later. I am not sure if seeing the action in live view is going to change much except the smaller sensor of the 4/3 camera might give you instantly some more dof.

Thanks for further advice on the image. The POV would be a little difficult as there is a wall in the way and the truck is immobile. However, will be removing the wheels to change the tires soon so that will be a better opportunity.

Next move I'll make is, as suggested, to move back from the subject and then crop.

I had developed the habit of avoiding cropping with my 3.4MP Sigmas because those big 9.12 um pixels let me use f/16 or even f/22 with very little diffraction effect. But 12MP crammed into a MFT sensor is quite a different matter!

[RustBeltRaw]

What is Magic Lantern? Is it firmware available for all cameras?

Magic Lantern is a Canon firmware hack compatible with the 5D mkII, 60D, 600D, 550D, 500D, 50D, and 1100D cameras. Unfortunately the Nikon equivalent is still in the works, and I don't know of equivalents for any other brands. It's mainly designed to streamline video production and audio recording, but has plenty of handy photography tools (intervalometer, automatic HDR, highlight warnings, focus peaking, etc.).

[xpatUSA]

Magic Lantern is a Canon firmware hack compatible with the 5D mkII, 60D, 600D, 550D, 500D, 50D, and 1100D cameras. It's mainly designed to streamline video production and audio recording, but has plenty of handy photography tools (intervalometer, automatic HDR, highlight warnings, focus peaking, etc.).

Thanks,

I've been reading more about close-focus DOF and the formula most-quoted is DOF = 2Nc(1+m)/m^2

If one fixes c to match the sensor, or uses the oft-quoted 1/1500, then it should be easy enough to create and carry a 3-column nomograph that relates DOF, N and m. "m", in the world of photography, is the size in the image plane divided by the subject size - easy enough to estimate, I reckon ***. The cool thing about a nomograph is that one can get any one variable from the other two . . and it doesn't need batteries

*** [edit] easy enough if you have 100% live view and do actually know your sensor size.

Ted

[jcuknz]

Next time I photograph something small similar to your cicada husks I will try and remember the recent thread where Richard [Fiji] proved to us that cropping from further back gave one more DoF. To date I have always gone in as close as possible and accepted that only part of my subject will be sharp .. it maybe a hard habit to break but I will try ... I promise myself
More viable a technique now we have 16Mp and more cameras I think.

I have alwasy assumed that any viewing of enlarged images is done wide open to emphasise what is the focus point one has obtained and I suspect the x10 view is not like focusing with a zoom lens* and then coming back but rather a digital enlargement of the centre of the frame ...and * as Manfred mentioned awhile back there is a special name for lenses which can work this way, which of course 'bird brain' has forgotten, and the average DSLR lens doesn't work this way ... it is a characteristic of the movie lens which is where I first struck it and learnt to use the technique and didn't realise it doesn't always work in still cameras/lenses. Since I rely on AF in these days of digital it has not caught me out.

[GrahamS]

I also have several questions about the calculation of DOF and other relationships. Firstly, I don't know where the focus distance is measured from. Is it the front of the lens? With a simple single element lens, the principles of physics and geometry are relatively simple. However, with a multi-element lens like all modern ones, it is not clear to me how these calculations should be done. For example, with my 100mm macro lens, the minimum focus distance is quoted as 0.3m and the magnification as 1:1. These numbers are not what you would get with a simple lens.

Tony - Some manufacturers of macro (actually "close focus") lenses specify the minimum focus distance as being XXcm "from the front element" which gives you an idea of how close you can actually get from your subject. The actual focal length of ANY lens is measured from the "focal plane" which is the plane where the lens brings the image into focus when focussed at infinity, i.e. the film or sensor surface, to the nodal point of the lens group, the position of which varies according to the design of the lens. The nodal point of a simple single element lens is taken to be the centre of the lens in cross section. The focal plane is marked on most good cameras as a amall circle with a horizontal line through it on the top plate of the camera - that line is the focal plane position.

The figures 1:1 express the ratio between the size of the subject and the size of the image at the focal plane. 1:1 means that the size of the image on the sensor is life size or the same size as the subject. A lens that will produce a life size image is a true "macro" lens as opposed to a "close focus" lens which may only produce a 1:2 image.

[xpatUSA]

Next time I photograph something small similar to your cicada husks I will try and remember the recent thread where Richard [Fiji] proved to us that cropping from further back gave one more DoF. To date I have always gone in as close as possible and accepted that only part of my subject will be sharp .. it maybe a hard habit to break but I will try ... I promise myself
More viable a technique now we have 16Mp and more cameras I think.

Yes, my DMC-GH1 is a 12MP and the sensor height is 13mm with 3000 sensor px. I too, was always eager to fill the frame. Today, the Panasonic Leica Macro-Elmarit 45mm showed up, so I rushed outside for some test snaps, including the cicadas. I took a snap at a greater distance than prior (2nd shot):




The rubber surface is a bald tire, BTW.

At these distances, the formula DOF = 2Nc(1+m)/m^2 applies well enough, remembering that the circle of confusion c is anything between 2xpixel pitch and a rote value of 0.015 for MFT lenses, depending on your view of "in focus". The larger cicada husk is about 28mm nose-to-tail and occupies 1.69mm in the image plane. That makes m=0.06 and the DOF theoretically some 70mm at c = 0.015 or, alternatively 40mm at c = 2 x 0.0043mm pixel pitch.

The second shot was hand-held with OIS on, the first on a tripod with OIS off. The difference in DOF is quite evident, I reckon!

[TonyW]

Graham, Thanks for responding to my comment. You have introduced a new topic about which I need educating. The definition of nodal point seems to be subject of some discussion and I still don't understand what it is. One article I looked at seemed to say that there are two nodal points. I understand the No Parallax Point (NPP), which apparently sometimes is confused with the nodal point. I believe that it's the position of the NPP which is relevant to the magnification. While I understand the calculation of depth of field for single element lenses, it is still not clear to me for multi-element lenses. In any case, I would have thought that nodal point(s) and NPP move as the lens is focused to different distances.