PDA

View Full Version : DOF of a 90mm lens on 4x5

hnaa
27-Apr-2011, 16:15
Hello all, sorry if this is a stupid question. I have become a bit confused as to matters of depth of field on 4x5 cameras. In particular, I am wondering if a 90mm lens has enough depth of field for landscape photography without using Sheimpflug tilting? If I understand it correctly (not a given), the Sheimpflug principle lets the focus plane drop into a wedge shape away from the camera. I am going to be making some pictures in wooded areas, and for that I was wondering if "conventional" (without tilt) depth of field at, say f22, would give me a good depth of sharpness? I would probably be making prints at about 20x25 or 24x30 inches.

Leigh
27-Apr-2011, 16:26
DoF is a very subjective subject. ;)

You can use the calculator at DOFMaster http://dofmaster.com/dofjs.html to estimate real numbers.

Calculations use a larger Circle of Confusion (CoC) for LF formats than for MF or 35mm, based on the assumption that the magnification required in printing will be less. You might want to adjust the CoC for your calculations, or select an MF format.

Even for 24" x 30" prints your magnification is only 6x.

- Leigh

Bill_1856
27-Apr-2011, 18:45
Yes.
Focused at 30 feet everything from about 6 feet to infinity is in focus. Hyperfocal distance is about 8 feet.

rdenney
27-Apr-2011, 19:19
DoF is a very subjective subject. ;)

You can use the calculator at DOFMaster http://dofmaster.com/dofjs.html to estimate real numbers.

Note that DOFMaster is using an 8x10" print as the standard across the different formats. If you want to make big prints, then I suggest setting the circle of confusion manually to something like 0.015mm.

Rick "who has it on his iPhone" Denney

Nathan Potter
27-Apr-2011, 19:46
As Leigh says a 24X30 in. print is about 6X from a 4X5. If you want about 300DPI on that 24X30 print (nominally a 3 mil pixel size (75 um)) then the COC at the film would want to be about 1/6 of the 3 mils or about 15 um, as master Denny says. That gives you, at the hyperfocal lens setting, about 6 ft. to infinity within a 15 um COC or better focus range. And you can ignore Herr Scheimpflug and achieve the 15um COC over the whole frame.

Nate Potter, Austin TX.

hnaa
28-Apr-2011, 00:49
Thanks a bunch guys, the reason I asked is that maybe I have the opportunity of borrowing a 4x5 Cambo Wide with a 90mm lens, so I would have rise and fall, but no tilt. I still intend to see if I can scrape together the cash for an Ebony RSW45 or a Walker XL Wide, but a used Cambo could be an option. I don't have a car, so something rather more portable than the Sinar F1 (I had for a while before having to sell it) would definately get me out making pictures more often.

dave_whatever
28-Apr-2011, 01:03
In particular, I am wondering if a 90mm lens has enough depth of field for landscape photography without using Sheimpflug tilting?

If you plan on doing compositions with a strong front-to-back foreground element then probably not, although you might be able to scrape by at f/45 some of the time. With really deep compositions you won't have a prayer without tilt.

If will depend on how sharp you want your shots to be though, or rather how much unsharpness you're prepared to tolerate in the name of Depth-of-field. If you're really strict on getting maximum sharpness then f/22 will only get you from inifinity to 6metres. f/32 would get you down to 3 metres. But for many purposes this will be overkill and way to strict, as they say "your mileage may vary".

hnaa
28-Apr-2011, 01:54
Hmm,
DOFmaster gives a hyperfocal distance for a 90mm at f/22 at about 12 feet, this other (very technical) site: http://www.cambridgeincolour.com/tutorials/DOF-calculator.htm gives a hyperfocal distance of about 30 feet (9 meters). I am confused again... I don't think I would want to stop down further than f/22.

ic-racer
28-Apr-2011, 09:02
If you have to ask this question, then I'd recommend keeping the standards parallel and you should be fine for almost all typical landscape subjects.
In many cases the amount of front tilt is so small it falls right on the edge of the detent on my camera. You will likely need a magnifier to know how much front tilt (if any) you need. On that same topic, make very sure the detent for "0" on your front and rear are parallel from side-to-side. If they are not perfect, one side of your horizon is always going to be out of focus on close examination.

If you are contact printing, ignore all that and just start shooting ;)

In terms of the different hyperfocal distances you have encountered. The hyperfocal distance will be different based on a myriad of conditions, most importantly the distance from the viewer to the print and the print size. The hyperfocal distance with that lens at f22 for a 4x5 contact print framed, on a wall in a museum, with a potted plant on the floor where you would like to stand, will be about 1 foot :)

hnaa
1-May-2011, 12:06
OK, sorry to beat a dead horse here, but I tried putting the CoC at 0.015mm in the DOF master calculator, which gives, for a 90mm lens on 4x5 at f22, a hyperfocal distance of 78.6 feet, with a near limit of acceptable sharpness of 39 feet. Does sound a bit restrictive. Going to f32 gets a hyperfocal distance of 55 feet, with a near limit of 28 feet. Would you say this fits with your experience? That for making a 24x30 print you'd need to consider a CoC of around 0.015mm? Obviously considerations of sharpness is a very subjective factor, and not necessarily the most important (which I would say, for me, would be composition).
I hope I don't sound too silly, I am just trying to understand all these figures.

BetterSense
1-May-2011, 12:34
I made a box camera with a 90mm angulon. I planned to have it fixed focus, but found that DOF was just not enough for many subjects if you are picky about OOF foregrounds and such. I added a focusing helical which makes the camera quite useful but now I wish I had some built-in perma-tilt; then I could orient the camera itself which ever way would work best.

rdenney
2-May-2011, 21:55
OK, sorry to beat a dead horse here, but I tried putting the CoC at 0.015mm in the DOF master calculator, which gives, for a 90mm lens on 4x5 at f22, a hyperfocal distance of 78.6 feet, with a near limit of acceptable sharpness of 39 feet. Does sound a bit restrictive. Going to f32 gets a hyperfocal distance of 55 feet, with a near limit of 28 feet. Would you say this fits with your experience? That for making a 24x30 print you'd need to consider a CoC of around 0.015mm? Obviously considerations of sharpness is a very subjective factor, and not necessarily the most important (which I would say, for me, would be composition).
I hope I don't sound too silly, I am just trying to understand all these figures.

If you can see 5 lines/mm on a print with good modulation transfer using only the naked eye, then you can multiply that by the enlargement factor to determine what the resolution at that modulation transfer needs to be. If the enlargement is 10 times the negative size, for example, then the negative needs to show good MTF at 50 lines/mm to provide that performance on the print.

If the spot is an arbitrarily small detail, then the fuzziness around it is similar to the black line in the line pair. For the line to look like a line and not like a patch of fuzzy gray, the modulation transfer has to be sufficient to clearly distinguish black from white. The circle of confusion for that spot is therefore about the size of the black line in those pairs of alternating black and white lines.

So, for that 10x enlargement, we'll need 50 line pairs, which means the black line will be 1/100 of a millimeter wide. If the circle of confusion is 1/50 mm, then the lines and spaces between them will merge into gray. So, the circle of confusion has to be 1/100 mm. That's 0.01mm. That's what it takes for something to look arbitrarily sharp on a 10x enlargement when studied up close with the naked eye. Yes, that is a demanding standard.

0.015 mm circle of confusion will support a 6.7x enlargement to that same standard. That's a print about 27x34 inches, which is quite a large print.

0.015 is a useful limit--it falls in the general vicinity of the limit of lenses. Some lenses resolve better and some less so, but this is in the middle of the pack.

For a 16x20 print, you need only 0.025 circle of confusion to provide that same standard of sharpness for the 4x enlargement from 4x5.

DOFMaster uses a circle of confusion of 0.1 mm, which corresponds to a 1x enlargement from 4x5 using this tough standard. That is a contact print.

You don't, of course, have to use a standard that tough. If you back up from the print far enough to see all of it, and if the print is about 16x20 or larger, then you won't be able to resolve five lines/mm on the print.

The point of all of this is to reinforce what is really true--you and only you can set your own standard of sharpness. If a calculation seems too tough, it may be that it reflects a higher requirement for sharpness than is important to you. We can't tell you where that limit is; we can only identify the boundary conditions.

I have made many photos, though, where I compromised depth of field or focus management because I thought the aperture just too small. The result was a limitation on how big the print could be. I'm limited on print size, so this is not usually that big a problem. At some point when stopping down, diffraction starts to create bigger fuzzy spots than being off the focus plane, and then diffraction is what is limiting print size. But in general I would rather be limited by diffraction than insufficient depth of field--uniform fuzziness being preferable in many cases to apparent loss of focus for foreground and background objects.

So, you don't necessarily have to set the circle of confusion standard based on an arbitrarily large print size, unless you are required to achieve a certain print size for a given image. You can just stop it down to the point where it looks sharp on the ground glass with a high-power loupe and leave it at that. Or you can make some test prints to find out where your personal standard has to be.

Generally, there's a reason scale-focus point-n-shoot large-format cameras use really short lenses. In that company of cameras, 90 mm is on the long side.

If these calculations seem senseless, and if DOFMaster's depth of field didn't demonstrate acceptable sharpness with its 0.1mm circle of confusion, pick something smaller and test again. You'll find what you need after a couple of tries.

Rick "whose biggest candidate aperture is f/22 most of the time" Denney

engl
3-May-2011, 02:44
OK, sorry to beat a dead horse here, but I tried putting the CoC at 0.015mm in the DOF master calculator, which gives, for a 90mm lens on 4x5 at f22, a hyperfocal distance of 78.6 feet, with a near limit of acceptable sharpness of 39 feet. Does sound a bit restrictive. Going to f32 gets a hyperfocal distance of 55 feet, with a near limit of 28 feet. Would you say this fits with your experience? That for making a 24x30 print you'd need to consider a CoC of around 0.015mm? Obviously considerations of sharpness is a very subjective factor, and not necessarily the most important (which I would say, for me, would be composition).
I hope I don't sound too silly, I am just trying to understand all these figures.

Yes, that does match my experience. I've been doing some shooting without tilt using scale focus at night (where the GG is too dark to confirm DOF stopped down), and a CoC of 0.015mm has worked well as a sharpness criterion at F16-22. That is of course my own subjective interpretation of "good sharpness", used for prints 24x30 to 30x40.

If your goal is big, sharp prints, I'd recommend getting a camera with tilts in order to get the best results with the widest range of compositions. Getting a very portable camera suitable for what you want to do does not have to be very expensive, look at something like a Shen-Hao TFC45-IIB.

hnaa
3-May-2011, 05:04
First off,
thanks to Rick Denney for the very good reply to my question. And to "engl", I guess you are right, a camera with tilts is preferrable, since I don't want to go wider than 90mm. I have actually been looking at the Shen Hao TFC45-IIB (had aimed for Ebony RSW45, but the money didn't stretch that far). Since I don't have the option of buying (or trying) it here in Denmark, I am still undecided. My main concern would be the stability of the camera, especially in regards to keeping the front and back parallel.

engl
3-May-2011, 07:12
Well, when you are tilting the idea is to not have the lens and film parallel :) But yeah, stability/rigidity is important, and the front and back standards should of course be parallel when you expect them to be. I'd recommend asking the people here on this forum about experiences with that camera, especially with 90mm lenses. Make sure to ignore anyone with the "It is not an Ebony so it is not good"-mentality.

Another good option might be a used Wista VX. Rigid metal camera, front tilt/rise geared, fairly lightweight, does not need drop bed with 90mm, can be fitted with bag bellows if you need major movements. The SP is slightly heavier but adds rear micro-adjustments.

hnaa
3-May-2011, 08:37
:D Well, I meant vertically parallel :D

hnaa
3-May-2011, 12:53
:D Well, I meant vertically parallel :D

Jeeez, I guess I am a bit stressed these days, horizontically parallel obviously:o :o

Leonard Evens
9-May-2011, 14:22
I question the rationale for reducing the coc in the negative significantly when making larger prints. You have to take into account the viewing conditions.

Suppose you have found that a given coc works well for you for 8 x 10 prints viewed at 10 to 12 inches, a normal viewing distance for that size print. If you plan to make a 16 x 20 print, which is double the size, linearly, of an 8 x 10 print, you would halve the coc only if you assumed viewers were going to view the print from the same distance, 10 - 12 inches. The argument would be that the coc in the negative would be enlarged twice as much in the larger print .

But the more reasonable assumption is that viewers will look at a 16 x 20 print from about twice as far as they will view an 8 x 10 print. In that case the angle subtended at the eye by the doubled coc in the larger print is the same as that subtended by the eye by the coc in the 8 x 10 print. So you should make no change in the coc you assume for the negative.

If you assume that all prints are viewed by putting one's nose right against the print, you might as well give up using DOF as a useful concept. If nothing else, diffraction will become a significant issue even for moderate f-stops if you view the print very closely.

rdenney
9-May-2011, 16:59
I question the rationale for reducing the coc in the negative significantly when making larger prints. You have to take into account the viewing conditions.

I did in my post. This is, of course, a decision for the photographer to make as to how he and others should be able to view his prints, and what standard of sharpness that imposes.

I don't know about you, but I can't see a thing with my nose against the print, even through the bottom lenses in my trifocals. But I often approach a larger prints as close as I can focus, to see if the apparently limitless detail holds up. If it gets fuzzy as you get close, the effect is lost. That effect is a fundamental tenet of large-format work for many photographers.

Rick "noting that many museums light so dimly that sharpness can't be enjoyed from any distance" Denney

eddie
10-May-2011, 05:31
Get a chamonix. I love mine.

Will the borrowed camera allow you to tilt the camera down and then use rear tilts? This may give you a bit if slemflug.

hnaa
10-May-2011, 07:06
Hello all,
thanks for all the great replies. It seems fate made my decision for me, when a used (EX++) Ebony RSW45 turned up in the UK at 850£ (about 1300\$), which is almost exactly what it would cost me to get a Shen Hao with Ebony screen and bagbellows shipped to Denmark. So I jumped, and the camera should get here tomorrow, boy am I looking forward to getting that beauty! I'll wait with the Cambo Wide until I can afford a Wide DS with 35mm Rodenstock on a tilt-board, and a digital back. Shouldn't take me more than a century or so to save up to that outfit.
Also, it turns aout that a photography club here in Copenhagen has a complete darkroom, for large format. That would certainly shave off some of the costs of developing (and scanning). I am happy as a pig in s... :D
Now I just need to get a 90mm, some good film holders, a spot meter, and I'm ready to go (luckily I have a very good, sturdy tripod already, and a couple of packs of velvia in the fridge). Thanks again everybody.

Leonard Evens
10-May-2011, 12:56
I did in my post. This is, of course, a decision for the photographer to make as to how he and others should be able to view his prints, and what standard of sharpness that imposes.

I don't know about you, but I can't see a thing with my nose against the print, even through the bottom lenses in my trifocals. But I often approach a larger prints as close as I can focus, to see if the apparently limitless detail holds up. If it gets fuzzy as you get close, the effect is lost. That effect is a fundamental tenet of large-format work for many photographers.

Rick "noting that many museums light so dimly that sharpness can't be enjoyed from any distance" Denney

"nose agains the print" was figurative. But before I had my cataracts removed, I was so nearsighted that I could actually do that by taking my glasses off.

I have a pair of glasses I had my otician make specially for viewing the 4 x 5 screen at about 6 inches. That gives me a filed of view comparable to what I see when I view an 8 x 10 print at about 10 - 12 inches. Using those glasses with the near point/far point method, I can usually focus quite well without needing to use a loupe. I have yet to find an enlarged print made from a 4 x 5 negative or transparency which doesn't begin to break down when viewed from 6 inches.

BetterSense
10-May-2011, 13:08
What is the "near point/far point method"?