Depth of Field Test Results, Any thoughts on focusing techniques

greenspun.com : LUSENET : Large format photography : One Thread

The Set Up: Toyo 4x5 Monorail, Bogen Tripod and Head, Rodenstock 150mm APO Sironar N, Velvia 4x5 and 6x7 120 roll, Horseman Back. Three objects. One at 10ft 2nd 50ft 3rd 150ft approximation

Both standards are parallel and leveled, all items are on the same plane. I'll cut to the chase....

I focus far then near, the extension difference equal f22 + 1/2, 6 divisions on my Rodenstock calculator. I set it to f32 for good measure. The test consists of focusing:

1. 1/3 of the way in from the closest object (the object 10ft away) two divisions to be exact. 2. 1/2 the distance, 3 divisions. 3. 2/3 of the way in or 4 divisions.

The results: Were Very Repeatable, not much of a shock there....

When I focused 1/3 of the way in, the object 50 ft away was in focus but some what soft. The object approximately 150ft away was defenitely soft.

1/2 of the way in, the equivalent of splitting the depth of field in half, showed a slight softening on the item ten feet away but still acceptable, and a noticable improvement of the object at 50ft and 150ft-not much of a surprise there either since your shifting the focus towards the objects at 50 and 150ft away.

2/3 of the way in was the real surprise for me. The foreground was still clear enough to consider sharp and the objects at 50 & 150 really dialed in.

So where's this going? I've read the DOF charts by Schneider,been on countless web sites ( Robert Wheeeler, Photo.net, Sinar, Chris Perez's site- a real good one-I might add, really gets into the numbers and off the "name brand hype") and I really expected to see the testing favor the first 1/3 method since this is what I've always been told works. Of course you get the guys who swear the answers to all of your questions is on the GG,and I'll admit there is allot to be said for that method also but in all honesty folks can anyone really see a difference when stopped down to F22-32? I sure as heck can't!!! So for those of you who are traveling down the same path I was on, It's well worth your time and effort to experiment for yourself. And If I were asked today where I set my focus point I might very well say a 1/2 - 2/3's of the way in.

Any thoughts would be appreciated and considered, just don't have me go blind staring at the GG at F32.

Albert

-- Albert martinez (martinez@unitekmiyachi.com), August 10, 2000

Answers

I am not sure what you were attempting to accomplish here? But a few bits I can offer. What you see on the gg, assuming one can see at f stops like f32 is close to what the chrome may look like with no magnification. So if you are making enlargements, that is not effective. Plus, film records way more detail than a gg can, so this is not the most reliable method.

Secondly, the amount you plan to enlarge the chrome dictates the required circle of confusion. So simply work backwards... figure the lpmm you want on your final print size, divide by the enlargement factor and that is max. cc the chrome can handle to enlarge succesfully to meet your "on print" resolution requirments. Enter the cc into the DOF formula and you will find the exact point of focus required for each near and far. Focus to that point, and stop down to the fstop required. To simplify this, I make handy DOF charts I carry in the field that has all this worked out. Then I compose my scene, use a laser range finder to get the near and far distance, look down the chart for each lens, and find the focus point and f stop required. Works perfect, no guessing. There is no hard rule here, you must work backwards to find out your focus point and f stop.

Did not want to bore you with the DOF formulars, coonsidering all the web sites you visited, you should have them... best of luck..

-- Bill Glickman (bglick@pclv.com), August 10, 2000.


One comment, if you read Harold Merlingers books on focus he talks about the idea fo things being acceptably identifialbe from the focus point back to the camera when your f/stop is smaller that the objects. Your f/stop was about 4.7mm which is smaller than a blade of grass so even things that are technically out of focus will still have detail in this range. By focusing deeper into your shot you are favoring the farther objects and your f/stop was small enough to keep the closer objects identifiable if not sharp. "The Ins and Outs of Focus" has more on this approach to focussing.

-- Jeff White (zonie@computer-concepts.com), August 10, 2000.

Laser range finder, charts, Bill you make me ashame on how easy I take this issue! No wonder my slides are not always as sharp as they should be!

-- Paul Schilliger (pschilliger@smile.ch), August 10, 2000.

Marklinger has some stuff (I'm sorry - I don't remember the exact cite - I can look it up if you want)where he demonstrates that the 1/3rd in rule holds under only one condition - when the far limit of depth of field is precisely twice as far from the lens as is the near limit of depth of field. He's done lots of work on DOF issues and might be worth checking out if you want more details. Me, I trust the GG. I know film can resolve more but thats another way of saying that if its sharp on the GG, it will be sharp on the film. I'm sure there are differences between how stuff looks on the GG and on film, but I think its possible to develop an eye for these differences. That is, we can tell how it should look on the GG for it look sharp on film. Just one opinion, of course. If your leaning is towards more methodical working, I think Merklingers site or books might repay the visit. Cheers, DJ.

-- N Dhananjay (ndhanu@umich.edu), August 10, 2000.

Just a few extra comments.... Albert, I did not clearly explain one point, that helps defy the gg theory for absolute acheivement of your goal. I will try.... lets say you want a 10x enlargement with 5 lpmm on your final print..(a high standard indeed) so you will need 50 lpmm MINIMUM on the entire chrome...however, unless you are focusssed at infinity, this never happens. Rather the point of focus (PSF - plane of sharp focus) will have the highest lpmm of resolution...as you get further away from the PSF on each side, the resolution drops fast. So the goal of 50 lpmm on the chrome will be acheived only at the near and far points and it will increase going towards the PSF. This is the purpose of the DOF formula. (Of course we are approaching the max. film can resolve and lense quality, etc... but it was simply an example.... so), the reason the gg is unreliable is because it has been estimated that it can only resolve 10 lpmm... so the best you can check for is 10 lpmm not 50 lpmm which is the min. you want. That is why charts are so effective. However, I will be the first to admit, the gg plays a critical when using movements! When I use a view camera with no tilts, I simply look at my DOF chart, determine the focus distance, find something at that distance and focus on it, DONE. It is the simpliest and easiest method I ever had and produces pefection nearly every time. After composure, I spend less than 10 seconds determining the focus point from the chart and focussing on it!

As for Merklingers book, the Ins and Outs of focus.... I am a big fan of what he has done for understanding and visualizing tilts, he was billiant. But the Ins and outs, leaves me and many others with some doubt. For example, as mentioned above... Merklinger states that the diam. opening of the f stop dictates the smallest item that can be resolved. However, if you shoot a wall of grass at f 4.0, in theory none of the grass should be resolved since the diam of the apt. at f 4.0 is much wider than the grass... even slightly in front and behind the PSF. And will all know the grass would be in perfect focus.

Also, I have tested his theory on the focus point and how the ability to resolve things moving further from the PSF continues to open in a cone shape.... meaning going towards infinity the cone becomes huge, when focussed less than halfway to infinity, meaning nothing should be resolved at all.... This theory seemed to make a ton of sense in the book, but when I tested it, I could not produce his results. While I do have reliable results with the standard DOF formulaes. Has anyone else tested this?

After reading the his book on tilts first, I was so sold on his knowledge, I almost automatically beleived anything he said..but as the saying goes in this field, test...test.. and then test....

-- Bill Glickman (bglick@pclv.com), August 10, 2000.



When I need very high precision, I use two different Palm Pilot programs (Vade Mecum and PCAM) to calculate stuff like this.

Vade Mecum (Robert Wheeler) program:

near: 10 ft far: 150 ft focus at: 18.73 ft

coc: .1 yields f=32 +3/10

coc: .075 yields f=45 +1/10

coc: .05 yields f=64 +3/10

PCAM gave approximately the same answers. Sure doesn't follow the 1/3 - 2/3 rule.

-- Larry Huppert (Larry.Huppert@mail.com), August 10, 2000.


1/3 applies to object space (it is actually an approximation which often breaks down), not to image space, ie your rail.

1/2 splits the sharpness evenly, but perceptually, you need more sharpness in your 150 object because it is smaller, so 2/3 appears more satisfying. See also II.3 in http://www.cs.berkeley.edu/~qtluong/photography/lf/how-to-focus.html

-- Q.-Tuan Luong (luong@ai.sri.com), August 10, 2000.


Moderation questions? read the FAQ