Hyper Focal Distance

[Destiny]

Hmm. At the moment my only worry is that some of the pictures will end up being printed as 18x12" minimum. Obviously, for GPx panoramas I go bigger.
KenS, I would really appreciate the links or any info on how to work this out to get the best results.

[Dick Roadnight]

rags DOF calculator

Is the best I have found... they include a link to DOF on Wikipedia which quotes three different formulae for DOF, for normal, close-up and Macro photography. There is so much nonsense written about DOF, but much of it is OK for standard amateur or DSLR expectations.

With rags DOF calculator you can specify any sensor size, but I think that if you specify 645 it assumes you are using film, and you need to specify the circle of confusion.

I seems that for 645 6 micros sensors 10, 20 or 30 micron CoF is OK for critical focus staking to hand-held portraiture... For a head shot, 20 or 30 microns gives enough DOF at 645/110mm/f16, and this gives good results in practice.

With most "normal" cameras, I think the Anti-Aliasing filter limits the CoC to about three times the pixel pitch, so 20 microns would be OK for a 6 micron FF 24 * 36mm sensor.

Most of what has been written about DOF does not seem to take into account that some of use large, high-res cameras because we want to be able to print large, but with res than can be appreciated from 250mm.

There are similar formulae in Merklinger's books, but they are confusing.

[RFPhotography]

Most of what has been written about DOF does not seem to take into account that some of use large, high-res cameras because we want to be able to print large, but with res than can be appreciated from 250mm.

And there, in all of this, is the crux of the issue.  All of this hand wringing about DOF, whether hyperfocal focusing is the right approach, whether focusing at infinity is the right approach, CoC size, really means nothing in 99.9999999999999999999999999999999999999999..............................% of cases.  We're talking about micrometers of difference at the sensor/film plane and in a print.  The only time this really matters, within reason, is when idiots like alleged photographers stick their noses right on a print or pixel peep at 300% on a computer screen.  At normal viewing distances these things are indistinguishable.  Yes, a very trained eye might be able to tell the difference (big stress on might) a certain, small percentage of the time but the vast majority of people won't be able to.  The vast majority of seasoned art buyers won't be able to.  This kind of discussion really is the purview of the internet photo forum and has little to no relevance or applicability in the real world. 

For those who don't do metric, 250 mm is 25 cm which is 10".  No one, except pixel peepers, is going to spend much, if any, time looking at a 20x30 print from 10".  Combine that with the distortions that putting the print behind glass or acrylic bring in and the even worse distortions that a good deal of the anti-glare glass or acrylic bring in, couple that with the softness created by a number of the printing media available on the market today (e.g., velvet papers, watercolour papers, heavily textured rag papers, canvas) and the relevance of this type of discussion becomes increasingly less at an increasingly rapid rate. 

[Dick Roadnight]

The only time this really matters, within reason, is when idiots like alleged photographers stick their noses right on a print or pixel peep at 300% on a computer screen.  At normal viewing distances these things are indistinguishable.

Some 24 * 48" picture would be observed from 48" or more... but, in a village scape people might look at individual houses or people, and in a group photo you would be looking at individuals... and it is nice to be able to produce a half reasonable A4 (10 * 8 ) picture of anyone in the group.

Many weddings are the last occasion at which an elderly relative is photographed dressed-up.

[RFPhotography]

and it is nice to be able to produce a half reasonable A4 (10 * 8 ) picture of anyone in the group.

That's got more to do with format and how much you can crop and still have enough film or pixels than it does with DOF.

[Dick Roadnight]

and it is nice to be able to produce a half reasonable A4 (10 * 8 ) picture of anyone in the group.

That's got more to do with format and how much you can crop and still have enough film or pixels than it does with DOF.

It is  not all about pixel count... there is little benefit in having enough pixels if you have too little DOF, and some heads in the group are OOF.

[RFPhotography]

I said 'more about', not 'all about'. 

And, quite frankly, if someone can't figure out how to get DOF right with a group photo without resorting to plugging micrometer measurements into a Cray supercomputer then they should probably step out from behind the camera. 

[Dick Roadnight]

I said 'more about', not 'all about'. 

And, quite frankly, if someone can't figure out how to get DOF right with a group photo without resorting to plugging micrometer measurements into a Cray supercomputer then they should probably step out from behind the camera. 

A view camera would work better, and let you get all the heads in the plane of sharpest focus, and allow for sharp focus on any decorative stonework above and behind the group.

[nma]

I wrote earlier in this thread about a practical solution to this problem. It landed with a thud; no one took notice.  If I were shooting the group photo with my Canon 5dii I would select an appropriate lens, maybe the  24-105, 0r 24-70, mount it on my tripod and set the camera to live view. Take special note, I AM NOT TALKING ABOUT AUTO FOCUS WITH LIVE VIEW. Then one makes a trial focus, say on a face in the middle of the group. The focus is effected manually with live set to 10x, effectively set at the single pixel level. Then, with the live view mode set to to exposure simulation, you push the dof preview button, the camera adjusts the screen brightness (you will be able to see the image clearly), and you can explore the dof, moving the focal area with the joy stick. You may wish to select a different point of focus to optimize for your composition, or set a smaller aperture, etc.  As I said before, this is the modern equivalent of the view camera's ground glass and focusing cloth, only the image is right side up.  The charm of this procedure is that the autofocus mechanism of the camera is eliminated; you are looking at a rendering of the image directly from the sensor; you have, in addition, a real-time histogram. You will be amazed how much sharper your images can be with this technique. Lenses you thought were soft may be resurrected when focused using live view.

[RFPhotography]

How long does all that roaming around at 10x magnification take?  Focus here, press this button, shift around over there, re-adjust, reset, work that setting.  Might be OK for a landscape or other static scene (e.g., architecture) where no one is going to get impatient but if someone were farting around with the camera, futzing with this and that for more than about 5 seconds and I were in the group, I'd say screw it and walk away.  Even with architecture, if I'm at a client location they want me in and out as quickly as possible.  Wasting their time by screwing around with the camera isn't going to make them overly happy. 

Dick, you're right.  It would.   

[Dick Roadnight]

A view camera would work better, and let you get all the heads in the plane of sharpest focus, and allow for sharp focus on any decorative stonework above and behind the group.

Dick, you're right.  It would.   

Thanks, Bob...

I am seriously thinking about this... I could set the camera up in advance, and use a pole or two to indicate the position of the POS, so arranging the heads in the POS would not be difficult or time-consuming... need to try it on a camera club or some of my relatives first.

[Bart_van_der_Wolf]

I am seriously thinking about this... I could set the camera up in advance, and use a pole or two to indicate the position of the POS, so arranging the heads in the POS would not be difficult or time-consuming... need to try it on a camera club or some of my relatives first.

Hi Dick,

That would work, although one could pre-compute the tilt angle needed for certain angles (e.g. stair steps often have a common slope, or a ladder has a known height) and then all that's needed is focus. Of course, more difficult setups (e.g. cresent shaped groups ) require more difficult calculations ... Setting up things prior to the group's arrival is the easiest approach in difficult settings anyway.

Cheers,
Bart

[Dick Roadnight]

Hi Dick,

That would work, although one could pre-compute the tilt angle needed for certain angles (e.g. stair steps often have a common slope, or a ladder has a known height) and then all that's needed is focus. Of course, more difficult setups (e.g. cresent shaped groups ) require more difficult calculations ... Setting up things prior to the group's arrival is the easiest approach in difficult settings anyway.

Cheers,
Bart

Yes, thanks, Bart.

If part of the object of the operation is to get something above and behind the group in focus, you measure the distance from the camera to it, and it's height above the ground... pick a line on the ground where the POS passes through and do a little Merklinger geometry to calculate J, the distance below the lens of the POS, and you can then calculate the tilt angle with trig - no Cray required, I would do it on the laptop or with an excel spreadsheet on my mobile phone, and I think it would produce a significantly better result, especially if the detail above and behind could put the picture in context with the venue.

If I was working rectilinearly I would arrange the group straight, if I wanted a crescent shaped group I might consider pan-and-stitch, but that might be a PITA with wedding guests.

[ErikKaffehr]

Hi,

My experience is that with full format on hi res DSLR depth of field is an issue even at relatively large distances. If you need to enlarge (like making A4s of individuals in a group) you would need all the sharpness you can have at the pixel level. There is nothing illegitimate looking at fine detail in a large enlargement.

I got confronted with this when I went from 135 to MF in the film days. My prints were simply not sharp. I didn't go to MF to have 1/15 mm of COC but to achieve better sharpness than with 135. So I learned not to trust DOF scales.

Another way to see it is that you invest in a certain number of megapixels, especially if you buy a high end MFDB. So you pay something like extra 20000 USD for a 60MP back over a 30MP back (or so). Do you really want to throw away all those pixels?! You may not need them, of course, and there are a lot of other ways of wasting perfectly fine pixels:

- No tripod
- No MLU (Mirror Lock Up)
- Sloppy focus
- Stopping down beyond optimum
- Uncalibrated autofocus

Best regards
Erik

[Dick Roadnight]

So you pay something like extra 20000 USD for a 60MP back over a 30MP back (or so). Do you really want to throw away all those pixels?! You may not need them, of course, and there are a lot of other ways of wasting perfectly fine pixels:

- No tripod
- No MLU (Mirror Lock Up)
- Sloppy focus
- Stopping down beyond optimum
- Uncalibrated autofocus

Best regards
Erik

We have gone back nearly 100 years to the "hand or stand" cameras that used the new fast 5 * 4" sheet film that was so fast that you could use the camera hand-held, if you were lucky.

It is difficult trying to get hand-held ambient MFD head-shots to look better than what you could do with 35mm, but I do not have a 35mm FF DSLR, so it does not stop me trying... to get adequate DOF you need f16, so you are starting to get diffraction, then you need to crank up the ISO as a compromise to reduce the effect of camera shake...

[ErikKaffehr]

Hi,

My point is just that extracting the maximum resolution from any equipment demands careful work. I also assume that those who are willing to invest a serious amount of money in high res equipment, like the P65+, would expect the system to resolve optimally. It seems that resolution comes at a premium price. There are plenty of lower resolution digital backs, a few more affordable. But anyone investing in the P65 over a lesser MFDB is probably striving for resolution.

MFDB´s are claimed to have many advantages. The top of the line backs have higher resolution and larger sensor surface. So you are paying a lot for those advantages, but at least the resolution is easy to loose. Just stop down to f/22 and much is gone!

Best regards
Erik

[Dick Roadnight]

Hi,

My point is just that extracting the maximum resolution from any equipment demands careful work. I also assume that those who are willing to invest a serious amount of money in high res equipment, like the P65+, would expect the system to resolve optimally.

Please note Like the H4D-60
You would like to imagine that people buying advanced amateur or professional equipment would have some idea how to get the quality they have paid for... but, in the real world, even some pros might imagine (initially) that a £30,000 camera will take a better picture than a £7,000 camera, regardless of the light or anything else.

MFDB´s are claimed to have many advantages. The top of the line backs have higher resolution and larger sensor surface. So you are paying a lot for those advantages, but at least the resolution is easy to loose. Just stop down to f/22 and much is gone!

Best regards
Erik

The larger sensor area reduces DOF, so it is not necessarily an advantage in all situations, so you use a smaller aperture to get back the DOF and loose the res through diffraction, and (if shooting hand-held) through camera shake,,, and do you lose less if you up the ISO to enable you to use a higher speed with a small aperture?

Even the old, low-res Hasselblad MFD's can use all the advantages of Phocus, and will make FF 35mm DSLR's look sick, if used well on a tripod or with flash.

The usual wisdom is that anyone with an expensive camera is an idiot:

... for having spent a great deal of money on kit that will not produce (proportionately) better results

... for having kit that they are not clever enough to use to it's full potential

Some people just make excuses because they are jealous... or they think there is no justice in the world as they think they could make better use of the kit. ...but, if you gave them the opportunity to use 60Mpx kit, you might not get it back in serviceable condition.

[Mike Sellers]

I have been using a program on my pocket pc for focus based on Merklingers work available here  http://www.dl-c.com/PPC/#DOF 1.0

[Dick Roadnight]

I have been using a program on my pocket pc for focus based on Merklingers work available here  http://www.dl-c.com/PPC/#DOF 1.0

I think the Merklinger formulae are the same as those on Wikipedia, and I have programmed one or two f them into excel on my Mac, and transferred it to my mobile phone...

...have you only included the "landscape" formulae. or the macro formulae as well?

...do you have an excel version?

[Mike Sellers]

Dick,
This program was made available by Jonathan Sachs at Digital Light and Color. I didn`t write it-just used it with excellent results.