[sewebster]

not really, you maintain the true DOF is zero, which is different than perceived sharp.

I think we're just saying the same thing in different words. There is one plane of focus, and as you move away from that sharpness drops, but within a range this may not be detectable, hence the concept of DOF. Then of course there are various definitions of DOF based on reproduction ratio, visual acuity, viewing distance, sensor etc.

[sewebster]

I thought that I would toss out the question on this thread - given the discussion on the one of the most current lenses of the crop. The SMC Pentax-DA 35mm F2.4 AL. How would you go about applying any sort of hyperfocal adjustment using this lens for example? No distance scale, so do you just sort of wing it?

• SMC Pentax-DA 35mm F2.4 AL Lens Reviews - Pentax Lens Reviews & Pentax Lens Database

So would you flip it into manual, rotate it all the way to the right for infinity, and then back off a "smidge"? Or, would you rotate it stop to stop to see what the angular travel would be from minimum distance to infinity and then auto focus on something of interest, then back off 5 or 10% or the angular travel (less towards infinity and more towards the minimum focusing distance) and take a test shot?

If I wanted to use this lens "hyperfocally" I'd calculate the hyperfocal distance for each f-stop of interest and either remember that or take the numbers with me. Then when setting up the shot I'd focus on something that I thought was further away than the specified value for my selected fstop (further in case I misestimated).

Personally I tend not to use hyperfocal much, because if I screw up I lose infinity sharpness, and in a lot of cases that is way more annoying than getting a bit more foreground DOF. It certainly can be used well though....

[RioRico]

Very soon, I shall list for sale a circa-1963 Voigtlander Bessamatic Deluxe film SLR with a Color-Skopar X 50/2.8 lens. Lenses for this type of camera have a curious DOF mechanism. As aperture (which is mechanically linked to shutter- and film-speed settings) is adjusted on the camera body, little pointers on the lens move across the distance markings, to indicate the DOF at any aperture and focus distance! A quick glance down at the lens thus reveals everything!

Now, this is an over-50-year-old design of a mechanical system, that is rather more informative than modern digital systems. Curious.

[interested_observer]

The fstop to use to put everything behind the scene in focus is about the focal length (35 in this case) divided by the width of the scene in meters or yards.

For example a scene 5 yards (meters) wide would need
F= 35/5 = 7 so I'd use f:7 or f:8. For a scene 10 meters wide f:3.5 should be enough with your 35mm lens.

Try it.

Wonderful explanation -

focal length / scene width = aperture

35/5 = 7.0
35/6 = 5.8
35/7 = 5.0
35/8 = 4.4
35/9 = 3.9
35/10 = 3.5
35/20 = 1.8
35/30 = 1.2
35/40 = 0.9
35/50 = 0.7

The problem with this rule of thumb, is as you expand the scene width to infinity the aperture is driven to 0, when essentially it should logically run in the opposite direction.

However, the point I was trying to make is illustrated a bit better by sewebster .....

If I wanted to use this lens "hyperfocally" I'd calculate the hyperfocal distance for each f-stop of interest and either remember that or take the numbers with me. Then when setting up the shot I'd focus on something that I thought was further away than the specified value for my selected fstop (further in case I misestimated).

Personally I tend not to use hyperfocal much, because if I screw up I lose infinity sharpness, and in a lot of cases that is way more annoying than getting a bit more foreground DOF. It certainly can be used well though....

This is a scenario closer to where I was going. Without a distance scale on the lens or having the camera report it (which would just be more electronics to fail on the lens, with a more complex design to drive the price even higher), you are back to guesstimating everything with little or no input. Set f8, top the focus out at infinity and then back it off a smidge so as to pull more of the foreground in, and take the shot.

One easy and relatively inexpensive way to help here is to have DoF reported on the rear screen. This would be an easy software function to add with essentially near 0 cost. You have all the information - focal length, aperture and you can take a look at the object's distance you just focused on (while we still have our old fashion distance reported on the lens' barrel).

I am all for trying to keep the price of new lenses in the affordable range, however taking the basic information away from the user, does not necessarily help them to learn the "art" of photography, or develop as photographers. I am going to guess that 10 years from now, we will be talking about - way back when there were distance scales, and when the distance pin on the mount dies, you need to ........

Very soon, I shall list for sale a circa-1963 Voigtlander Bessamatic Deluxe film SLR with a Color-Skopar X 50/2.8 lens. Lenses for this type of camera have a curious DOF mechanism. As aperture (which is mechanically linked to shutter- and film-speed settings) is adjusted on the camera body, little pointers on the lens move across the distance markings, to indicate the DOF at any aperture and focus distance! A quick glance down at the lens thus reveals everything!

Now, this is an over-50-year-old design of a mechanical system, that is rather more informative than modern digital systems. Curious.

... and Rico - I like this concept. In terms of lenses, optics and mechanical internals are great. The only thing I really like about electronics in lenses, is the focal length and aperture reporting. Also, the mechanical screwdrive is wonderful. I still do not have a SDM lens.....

10 to 15 years from now the M42, K, M, A, FA and DA lenses will still be working. I think we will be living with the SDM lens failures and if only we could revert to screwdrive on these - particularly when the SDM motor goes Spaulding...

---

Last edited by interested_observer; 03-18-2011 at 07:10 PM.

[newarts]

The problem with this rule of thumb, is as you expand the scene width to infinity the aperture is driven to 0, when essentially it should logically run in the opposite direction.

It is telling you the smallest aperture to have focus all the way from what you are focusing on to infinity. When you are trying to focus on something already at infinity ANY aperture will do - that's why it tells you f:0 is ok.

The relationship is expressed that way because it answers the photographer's question - "What f stop should I use for hyperfocal conditions?"

The equation can be turned around to tell you what scene width would be hyperfocal for a particular lens and f-stop, but that isn't as useful.

It is not a rule of thumb but is mathematically exactly the same as the equations used by the on-line calculators; it uses a "Circle of Confusion" 1/1000th the width of the displayed image - about the same as a pixel's width for a laptop or LCD display.

If a person can divide in his/her head it is an easy equation to remember and use:

hyperfocal-fstop = focal-length.mm/scene.width.meters

PS. using width instead of distance is a great simplification for mental calculations - the equivalent calculation using distance involves the square of focal length - almost nobody is good at mental calculations involving squares!

[newarts]

One easy and relatively inexpensive way to help here is to have DoF reported on the rear screen. This would be an easy software function to add with essentially near 0 cost. You have all the information - focal length, aperture and you can take a look at the object's distance you just focused on (while we still have our old fashion distance reported on the lens' barrel).

For modern lenses the camera also knows the focusing distance. It would be trivial to automagically calculate and set the hyper-focal f-stop.

[newarts]

I think we're just saying the same thing in different words. There is one plane of focus, and as you move away from that sharpness drops, but within a range this may not be detectable, hence the concept of DOF. .....

sewebster is right for geometric optics and in his statement "within a range this may not be detectable". - for real physical optics the detailed situation is not so simple as the crossing of geometrically straight lines..

It turns out that there is a real out-of-focus range where it is hard to know if focus is precise. The "Airy Diffraction Disk" not only has a diameter it has a height (about four times its diameter I think - like a little sausage) - within that height focus looks about the same!

Here's a sideways view of actual "focal points" from Wikipedia (the middle image is for a perfect lens; the others have lens defects):

Airy disk - Wikipedia, the free encyclopedia

When there's a digital sensor involved the size of the individual sensor pixel also puts a limit on effective DOF...because the focal point can fall anywhere on the pixel and it will look the same to the pixel.

---

Last edited by newarts; 03-19-2011 at 05:26 AM.

[sewebster]

Thanks newarts. Good idea that for "real" optics the "absolute" DOF could actually be non-zero.

[rhodopsin]

Physicist or not, you are incorrect

Depth of field like acceptable sharpness from camera shake is a function of "acceptably sharp" specifically what your eye can resolve from a specific distance. While infinitely sharp focus only occures at one point, acceptably sharp has a defined r ange which is a function of enlargement at a set viewing distance

I would say my collegue was 'imprecise'. The focal plane has no thickness (and position depends on wavelength). DOF is a definition, as given above. Perceived DOF depends on magnification.

[Lowell Goudge]

I would say my collegue was 'imprecise'.

precision is all important in scientific descriptions., but anyway.....

[John Poirier]

I would say my collegue was 'imprecise'. The focal plane has no thickness (and position depends on wavelength). DOF is a definition, as given above. Perceived DOF depends on magnification.

I think the concept of perception is key to the discussion. As well as magnification, individual beliefs about what constitutes a sharp image are a factor. There is also the tradeoff between DOF and loss of sharpness due to diffraction. What it boils down to is that the simplest way to deal with the problem is to test your own lenses under field conditions and determine what settings meet your own requirements. Depending on what you have for gear, you can then develop your own fudge factor for DOF calculators or prime lens DOF scales. Or you can make notes for zoom lenses.

For example, when shooting 35mm film I tend to close down one stop more than the DOF scale indicates. I tend to make fairly large prints and DOF that looks good for an 8x10 doesn't necessarily cut it for a 12x18..

John

[sewebster]

precision is all important in scientific descriptions., but anyway.....

Fair enough. I shouldn't have mentioned the physicist thing anyway, I wasn't trying to claim I was beyond reproach or anything... I think it was just somewhat nice to actually discuss something I had some training in for once.

In my defence though, I was responding to a post by Jimfear where he discusses his concept of "absolute DOF" which, I guess, is flawed terminology. So I was trying to discuss things on that level (e.g. by saying that "absolute DOF" was zero or something, but this is not very meaningful as this is not a correct definition of DOF). I also included a sentence in my reply to not get hung up on my imprecise terminology because DOF is defined according to a certain set of assumptions. But it seems like this warning didn't work. Oh well