Question about hyperfocal and R-D1S sensor

[ambientmick]

Having just got the R-D1S last week from Robert White (great service btw), I am learning about the limitations of RF photography as well as the freedom it provides compared to DSLRs. The manual controls are brilliant as opposed to a compromise as with most other cameras and I really appreciate the small size (including lenses) and relative quietness of the shutter. I must admit I was very surprised at the high quality of construction and ease of use when I received it. I would urge anybody who is thinking of getting the R-D1/S to try/buy one before there are none left. BTW I got the 40mm Nokton SC version which produces stunning B&W files.

Now to my question. I got the R-D1S primarily for street photography and other people based stuff and have read a little about zone focussing. Seems easy enough. Just set a hyperfocal distance with a reasonably small aperture (depending on the light) and get well focused pictures within a distance range read from the lens. The problem is that the sensor in the R-D1S is smaller than 'standard' 35mm so surely the hyperfocal scale on lenses is inaccurate? I don't have enough experience to know what adjustments to make (if any) and I would like to know if there is a 'rule-of-thumb' that would work for all lenses including my Nokton.

Thanks for reading.

 

[jlw]

First, keep in mind that DOF is only an optical illusion anyway (caused by the eye's inability to distinguish between a sharp image and a very slightly unsharp.) So, trying to get super-accurate about calculating it is a bit of an exercise in futility!

So if you'll be sure to remember that this is only a rough guide:

DOF is based on only two things: the subtended angle of the viewer's eye when viewing the image (this takes into account both enlargement/projection size and viewing distance) and the numerical aperture of the lens used to make the picture (actual diameter of the "hole" in the lens.)

The difference between the readings on the DOF scale of your lens and what you'll actually get is caused only by the difference in final magnification -- the R-D 1 images will need more magnification because the original image (on the sensor) is smaller. It's the same effect you'd get if you shot a 35mm picture than then enlarged only part of it to make your print; you'd need to magnify the image more to get the composition you wanted, so you'd have less total DOF.

The easiest way to do this (and doing it the easy way is OK, since DOF is only an optical illusion) is to apply the R-D 1's 1.53x "crop factor" to the lens focal length, and do the calculations as if you were using a 60mm lens on a 35mm camera (since the lens marks were calculated for a 35mm camera.)

Using a bunch of fairly hairy formulas I copied out of a Rudolf Kingslake book on camera optics, I once made up a DOF calculator in the form of a spreadsheet. I fired this up, input a 40mm lens set at f/8, at a magnification of 9x and a viewing distance of 10 inches -- very critical viewing conditions. The spreadsheet gave me the total DOF available at various distances. Then I tried poking in various other aperture values until I got roughly the same total DOF for a 60mm focal length.

I found that you'd need to stop down a 60mm lens roughly 2-1/3 stops to get the same DOF as a 40mm lens.

Since digital images don't have the same maximum level of detail as the best film images (which tends to give the illusion of greater DOF) and since I was using viewing conditions more critical than you'd probably normally use, I'd say that you could ignore the 1/3-stop part.

That means (finally, a conclusion!) that you should get pretty good DOF results by using the marks for an aperture two stops smaller your lens. For example, if you're shooting at f/16, use the DOF marks for f/8; if you're shooting at f/5.6, use the DOF marks for f/2.8; and so on.

As I said, after trying this for a while, don't be surprised if you find you actually get a bit more apparent DOF than this method suggests; if that's the case, feel free to tweak the method. For example, you may find you're satisfied with the DOF you get using the marks for one stop smaller.

I'll try attaching my spreadsheet as a .zip file in case anyone wants to play with it. I saved it from NeoOffice in Excel format, so don't be surprised if the formatting gets a bit weird. And don't bother replying to say "You should have used a different formula for the near-distance limit to take into account the effective focal length increase" or whatever -- remember, DOF is only an illusion, and even Kingslake said that trying to calculate it too precisely is a waste of time!

 

[richard_l]

There are (at least) two ways.

1. Divide the f-number by the crop factor. Hyperfocal distance would be hyperfocal distance for the new f-number. Example: If you are using f/8, 8/1.5 = 5.3, so set the hyperfocal distance for f/5.6 (or nearest marked f-number to 5.3). I.e., set the infinity mark to 5.6.

2. Multiply the focal length of the lens by the square root of the crop factor. If the crop factor is 1.5, that would be about 1.2. Hyperfocal distances would be the hyperfocal distances for the new focal length. Example: For a 35mm lens. 35 x 1.2 = 42, so use hylerfocal distances for a 40mm lens.

Richard

P.S. This has been argued about many times. My numbers are correct, and I even posted the mathematical derivation a year or so ago; so if anyone wants to argue, argue with someone besides me.

 

[ambientmick]

First, keep in mind that DOF is only an optical illusion anyway (caused by the eye's inability to distinguish between a sharp image and a very slightly unsharp.) So, trying to get super-accurate about calculating it is a bit of an exercise in futility!

So if you'll be sure to remember that this is only a rough guide:

DOF is based on only two things: the subtended angle of the viewer's eye when viewing the image (this takes into account both enlargement/projection size and viewing distance) and the numerical aperture of the lens used to make the picture (actual diameter of the "hole" in the lens.)

The difference between the readings on the DOF scale of your lens and what you'll actually get is caused only by the difference in final magnification -- the R-D 1 images will need more magnification because the original image (on the sensor) is smaller. It's the same effect you'd get if you shot a 35mm picture than then enlarged only part of it to make your print; you'd need to magnify the image more to get the composition you wanted, so you'd have less total DOF.

The easiest way to do this (and doing it the easy way is OK, since DOF is only an optical illusion) is to apply the R-D 1's 1.53x "crop factor" to the lens focal length, and do the calculations as if you were using a 60mm lens on a 35mm camera (since the lens marks were calculated for a 35mm camera.)

Using a bunch of fairly hairy formulas I copied out of a Rudolf Kingslake book on camera optics, I once made up a DOF calculator in the form of a spreadsheet. I fired this up, input a 40mm lens set at f/8, at a magnification of 9x and a viewing distance of 10 inches -- very critical viewing conditions. The spreadsheet gave me the total DOF available at various distances. Then I tried poking in various other aperture values until I got roughly the same total DOF for a 60mm focal length.

I found that you'd need to stop down a 60mm lens roughly 2-1/3 stops to get the same DOF as a 40mm lens.

Since digital images don't have the same maximum level of detail as the best film images (which tends to give the illusion of greater DOF) and since I was using viewing conditions more critical than you'd probably normally use, I'd say that you could ignore the 1/3-stop part.

That means (finally, a conclusion!) that you should get pretty good DOF results by using the marks for an aperture two stops smaller your lens. For example, if you're shooting at f/16, use the DOF marks for f/8; if you're shooting at f/5.6, use the DOF marks for f/2.8; and so on.

As I said, after trying this for a while, don't be surprised if you find you actually get a bit more apparent DOF than this method suggests; if that's the case, feel free to tweak the method. For example, you may find you're satisfied with the DOF you get using the marks for one stop smaller.

I'll try attaching my spreadsheet as a .zip file in case anyone wants to play with it. I saved it from NeoOffice in Excel format, so don't be surprised if the formatting gets a bit weird. And don't bother replying to say "You should have used a different formula for the near-distance limit to take into account the effective focal length increase" or whatever -- remember, DOF is only an illusion, and even Kingslake said that trying to calculate it too precisely is a waste of time!

Thanks for the calculator. I'll have to study it for a while.

 

[LCT]

One can use as well the hyperfocal formula

where f is the focal length, N the f number and c the circle of confusion (CoC).
For a usual CoC value of 0.02mm, this leads me to use the DoF markings of the nearest faster f stop of my M lenses i.e. f/5.6 when i choose f/8 for instance.
Works fine for me.
Best,
LCT

 

[JeffGreene]

DOF Calculator

DOF Calculator

Here's a nice little depth of field calculator written by Jonathan Sachs, the author of Picture Window Pro. It's a very handy little utility. It is no longer available on his site, but he does have the Pocket PC versions available there. Just google "Digital Light and Color" to get to them. It's freeware, so enjoy!

You'll need to take the 1.53 factor into consideration along with JLW's comments.

 

[LCT]

This DoF calculator is my favourite.
Based on 0.02mm CoC for R-D1 and R-D1s.
Works with Macs as well.
http://www.dofmaster.com/dofjs.html

 

[gyuribacsi]

Hallo! I agree with you all. But how would you manage the problem if the lighting of the scene is very poor, your lens is quite open and you are at 1600 ASA to get 1/15sec. In such situations you can`t focus because it ist too dark, you can`t read your DOF-caculator because your subject is mooving. So you can only estimate the distance an shoot and hope the result will be good or leave the place and wait for a sunny day.

Cheers George

 

[Joe Mondello]

Alright, I am a dummy at this math stuff, so in easy to understand English can anyone explain to me why DOF would change at all? The sensor is just cropping the circle. For a given Field Of View, I can see that the lenses would behave very differently, but in terms of simple focus and DOF of an image coming off the lens, what difference does cropping make -- if any?

I am -- genuinely -- confoozled at this point!

 

[LCT]

Cropping reduces the circle of confusion (CoC) on which DoF is based.
For instance the usual CoC value of 135 film is 0.03mm and that of APS-C digicams is 0.03:1.5 = 0.02mm.

 

[RichC]

Alright, I am a dummy at this math stuff, so in easy to understand English can anyone explain to me why DOF would change at all? The sensor is just cropping the circle.

If you printed the R-D1 image in exactly the same way as, say, a 35mm film image taken with the same focal length lens and at the same position, i.e. your R-D1 print is two-thirds the size of the 35mm print, then you're right: no difference in depth of field. Of course, you have a physically smaller print from the R-D1 compared with the 35mm film image - i.e. a crop of the latter two-thirds smaller.

However, the reason for the change in DOF is that R-D1 images will usually be enlarged when printed, as opposed to printed two-thirds the size of a 35mm print. This means that the out-of-focus areas will be magnified, so what seemed in focus on the smaller print is no longer sharp.

In short, depth of field is dependent not only on the lens and subject distance but also on how much the image is magnified when printed.

Personally, I'm happy with assuming a one-stop reduction in DOF for my R-D1, e.g. when using my lens to scale to focus, I set the distance scale to, say, f/5.6 when 35mm film convention would suggest f/8.

 

[jobo]

Alright, I am a dummy at this math stuff, so in easy to understand English can anyone explain to me why DOF would change at all? The sensor is just cropping the circle. For a given Field Of View, I can see that the lenses would behave very differently, but in terms of simple focus and DOF of an image coming off the lens, what difference does cropping make -- if any?

Only that a (e.g.) 8x10 print from a cropped image will have less DOF than a 8x10 print from an uncropped one (all other things being equal).

2c, /Jobo