Pinhole Camera Question...

[Capocheny]

Hi All,

I had posted a question back a short while ago in regards to pinhole cameras. I received a lot of great suggestions from some of the more knowledgeable folks on this forum. To them... I raise a glass of wine to say, "Thank you!"

I've recently gotten ahold of a standard wooden graftlock 4x5 back and an appropriate pinhole (f350 - produced with a Number 8 needle.) Now, my question is, "what depth should I build this camera (from pinhole to film plane)? What about the height and width? The plan is that the camera will be built out of mahogony or oak so it'll be fairly stable (as opposed to a shoe box or similar container.)

Ideally, I'd like the entire sheet of film to fall within the image circle of the lens thus avoiding the vignetting look of some images I've seen.

I'll apologize in advance if the answers to my questions "should be" absolutely apparent but this line of thinking is something I'm entirely uneducated about. Has anyone on this forum built a camera with a 4x5 back on it before? What were the dimensions of your camera? I've seen some of the cameras on the various links previously recommended but I'm still not entirely sure of the appropriate dimensions required for a 4x5 filmback.

BTW, John Kassian, I noticed that you had done some experimenting on a pinhole camera back awhile ago... how did things turn out for you. Are you still using it? :)

Thanks in advance.

Cheers

[Mark Sawyer]

If the f/stop is really f/350, then the distance to the film plane should be 350 times the diameter of the pinhole. Coverage depends on how thick the pinhole material is and whether there are any burrs pushed back by the needle pushed through it. Theoretically, infinitely thin material with a clean pinhole would have 180 degrees coverage. There may also be some light fall-off, though. You could figure this by measuring the distance from the pinhole to the center of the film and to the corners of the film, and figuring a "bellows extension factor." (Anybody ever use a center filter with a pinhole?)

[Emmanuel BIGLER]

Henry.

Short anwser first : if your pinhole is an optimum f/350 pinhole for 4"x5" very probably the best distance is 150 mm, see below but you can experiment with a shorter distance and a wider angle of view. The best is to know the actual pinhole diameter in order to better predict the optimum distance and corresponding f-number.

Now a detailed answer. ;-);-)

There are many interesting issues in your question.

The first issue is : given a certain pinhole diameter, what is the optimum distance to film for best image quality ? the answer is well-known and is given, for example, in Leslie Stroebel's book

Leslie D. Stroebel, "View Camera Technique'', 7-th Ed., ISBN 0240803450, (Focal Press, 1999)

Page 50 you find examples of images showing the effect of best pinhole size and page 49 you get the useful formula relating the pinhole diametre d_opt to the best film-to-pinhole distance D :

d_opt(in mm) = (1/28).D(in mm)^(1/2)

where "d" is the pinhole diameter and "D" the best pinhole-to-film distance
some examples :

D = 45 mm d_opt = 0.24mm ; theoretical angle of view on 4"x5" (diagonal = 150mm) 120 degrees
D = 90 mm d_opt = 0.34mm ; angle of view on 4"x5" 80 degrees
D = 150mm d_opt = 0.44mm ; angle of view on 4"x5" 53 degrees
D = 300mm d_opt = 0.62mm ; angle of view on 4"x5" 28 degrees

The corresponding f-numbers are given by N = D/d_opt = 28.(D in mm)^(1/2); to find the exposure time you have to extrapolate from a given (T, N) combination to (T', N') as T' = T(N'/N)^2

For example the sunny-16 rule says : 100 ASA film on bright sunshine, T=1/100s = 0.01 s @ f/16 (T =1/ASA seconds for N=16), hence we get :

D= 45 : N_opt=188 T = 1.4 s
D= 90 : N_opt=265 T = 2.7 s
D=150 : N_opt=343 T = 4.5 s
D=300 : N_opt=485 T = 9.1 s

If we assume that your pinhole is an optimum pinhole "f/350" we should be close to D=150mm and d_opt = 0.44 mm hence the best distance is like the standard focal length for 4"x5", focal = diagonal = 150 mm (plus or minus a few mm ;-) But in order to apply Stroebel's formula we need to know the diameter, or at least know if f/350 refers to an optimum pinhole.

Now the second issue is : what is the maximum angular field of view ? this cannot be determined easily, it depends on how thick the pinhole material is and whether the edges of the pinhole are tapered or not. A well designed pinhole can deliver some useful rays above 90 degrees of angle.
I would even say that pinhole images make sense to me for wide-angle fields of view since there is no distorsion at all, even for a 90 degrees field of view.

A last question is : given a 0.44 mm pinhole used at its optimum distance of 150mm, what is the corresponding image quality ? here... be brave because the image of a distant point source cannot be smaller in diameter than the pinhole size itself. So in your case, about 0.4 mm. If we fill a 4"x5" format frame with rows and columns of .4mm adjacent spots we get a total number of spots of 250x320 points for a total of... 80,000 points, 80 kilo-pixels, not 80 mega-pixels as we could scan any reasonable 4"x5" image taken with a modest glass optics ......

Another remark : you are not at all bound to Stroebel's formula, you can do as you wish, simply for a given diameter any distance different from the optimum distance will deliver images of lesser quality than the optimum. Distances shorter than the optimum deliver more illumination of film at the expense of less image points in the image whereas a bigger distance than optimum does not improve image quality (the total number of equivalent points will be constant vs. distance D beyond the optimum) but yields a much, much smaller equivalent number N, so exposure times become terribly long for no increase in image quality.

Last remark regarding the even-ness of light distribution in the image field. A pinhole camera obeys the simple cos^4(theta) law where theta is the angle of a ray plotted from the pinhole to a given point on film and measured with respect to the optical axis.

at +-45degrees form the axis (total angle of view : 90 degrees) the cos^4(theta) factor equals 1/4 i.e. two f-stops slower than on-axis. For a 53 degrees angle no problem the light fall-off effect is not actually negligible , 0.64 ~= -2/3 stop, but the effect is similar in most standard view camera lenses.

[Mark Sawyer]

" I would even say that pinhole images make sense to me for wide-angle fields of view since there is no distorsion at all, even for a 90 degrees field of view."

Unless you have spherically-curved film, there will be some noticeable "stretching" of the image at the outer edges in a wide-angle pinhole photograph. There is also a phenomenon that looks very much like distortion, most easily recognizable in the curving of straight lines, that happens when you condense a very wide field of view into a small rectangle.

[Emmanuel BIGLER]

Mark, you are absolutely right about 'distorsion' ; I meant : conventional barrel or pincushion distorsion is, as far as I can imagine, non-existent in a pinhole image. Of course the projection of a collection of spheres looks very strange in the edges of very wide field, but as far as I know, the projection of a flat grid-like object paralell to film should not exhibit any distorsion, a rectangular grid should not be changed into anything but a similar grid by the pinhole shadow-cast. The projection of a 3-D object is another story.

About using a centre filter : this raises an interesting but somewhat academic question since you probably do not want to purchase a centre filter which would cost five times the price of the whole pinhole camera ;-)

Modern wide angle lenses exhibit a light fall-off effect which is not as fast as the classical cos^4 (theta) law. So a centre filter even if it is placed in the right position (my guess would be : at the right distance with respect to the entrance pupil, i.e here : the pinhole itself) may slightly under-correct a cos^4 light fall-off effect but could be useful, at least to somebody who would be happy to have a spare centre filter ;-)

[Mark Sawyer]

" the projection of a flat grid-like object parallel to film should not exhibit any distorsion, a rectangular grid should not be changed into anything but a similar grid by the pinhole shadow-cast."

Let's say the grid is on a big wall that runs way off into the distance to the left and the right. Stand right in front of it, maybe five to twenty feet away, look straight at it and your eyes will tell you it's a grid of straight parallel lines. Look to your left, and all those straight, parallel lines will converge towards a vanishing point. Look to your right, again, you see straight parallel lines converging towards a vanishing point.

Now take a photo (pinhole or otherwise) with a wide agle lens pointed straight at the wall. In the middle of the photo, the lines will be straight and square. To both the left and right edges, you'll see the lines begin to curve in convergence towards their vanishing points.

The way our peripheral vision and visual perception in the brain functions, we can't see it happen in our field of view looking at the wall. (I've tried!) But when you condense a wide view into a print, you see it all condensed into a small enough area to see the effect.

Wanna see it happen? Find a big brick wall (for the grid) and photograph it from close up with a wide angle pinhole!

"About using a centre filter : this raises an interesting but somewhat academic question since you probably do not want to purchase a centre filter which would cost five times the price of the whole pinhole camera ;-)"

Closer to 500 times the cost! For anyone wanting a cheap, easy-to-make pinhole camera, get a metal cannister with the removeable push-on top (the kind butter cookies or some types of candy come in), paint the inside black, and drill about a 1/2" to 1" hole in the side. Tape a square of aluminum foil over it, and poke a small hole with a needle in the foil. I use these for my photo classes, and they work great! I usually buy the cans for 25 to 50 cents at the thrift stores, and they come in a wide array of "formats," including panoramic!

[Emmanuel BIGLER]

To both the left and right edges, you'll see the lines begin to curve in convergence towards their vanishing points.
Mark. We agree that we do not speak about what the eyes sees but what is actually recorded on film. If the film is plane and perfectly paralell to the wall, and if the effect that you describe is true, it means that perspective rendition in a pinhole camera in the edges does not obey the rules of simple conical geometrical projection like you can plot it by a geometrical drawing of a single ray crossing the centre of the pinhole. In such a geometrical projection the image of a circle drawn on the wall remains a circle even in the edges, and all straight and parallel lines on the wall remain perfectly straight and parallel on film if the film is parallel to the wall. Now what you say implies that this model fails in the real world of a wide-angle pinhole camera.
I'm wondering in an actual wide angle pinhole what would actually be the explanation for this apparent behavior departing from pure geometrical conical projection. Since a pinhole is a diffractive optical system and since diffraction departs from geometrical optics, should we be surprised to see effects departing from a geometrical model ? probably not ! As you know using the camera obscura was the key to re-discover conical projection in art after centuries of approximate perspective rendition but pinhole cameras used at the time were certainly not wide angle pinhole cameras.

[Capocheny]

Hi Mark and Emmanuel,

Thank you both for all the information... I think what I need to do is to gather up some good reading material on "the art of the pinhole camera." There's definitely a LOT more to a pinhole camera than just a simple box with a pinhole for a lens.

I've been told that using a #8 needle produces a diameter of .023" (061 mm) providing a Normal (7") focal length and an f-stop of 330. So, if I wanted to obtain the sharpest image with the least amount of distortion... this sounds like the way to go then. Would I be correct in this assumption? I'm under the impression that the smaller the hole, the sharper the image. Therefore, perhaps, a wide-angle is the way to go?

I was also told that using a #10 needle to produce a diameter of .018" (0.46 mm) would produce a wide-angle (5") focal length with an f-stop of 280. According to the discussion above... can I expect a fair amount of distortion at the edges then?

You both sound like you've had a great deal of experience with using pinhole cameras. Are you still shooting with one? What are your cameras like?

Mark... I'll be making the pinhole in a .002 brass shim and removing the burrs by means of sanding it down.

Emmanuel... thank you for the Stroebel reference but to be honest, I've never really been much of a math student. However, I'll try and go through the info.

Thanks again

Cheers

[Emmanuel BIGLER]

Henry. No maths are required. Stroebel's formula is a good starting point.
If the pinhole becomes too small, diffraction occurs and light rays open like a fan and the projected image is blurred. If the hole is too big you get more light but the image is blurred as well, so there exist an optimum.

I have attended a fascinating conference about the history of pinhole cameras at my local school of fine arts two years ago during the "world pinhole day" next event will be in May I think.
Pinhole photography means absolute freedom at the expense of a lack of sharpness. So, enjoy the freedom of pinhole photography : may be the most impressive pinhole image I've seen at the conference was a self-portrait of a pinhole photographer taken through a small hole in the middle of the fingers of his hand. In this image, emotion is here and maths, really far away.

[Emmanuel BIGLER]

Henry. No maths are required. Stroebel's formula is a good starting point.
If the pinhole becomes too small, diffraction occurs and light rays open like a fan and the projected image is blurred. If the hole is too big you get more light but the image is blurred as well, so there exist an optimum.

I have attended a fascinating conference about the history of pinhole cameras at my local school of fine arts two years ago during the "world pinhole day" next event will be in May I think.
Pinhole photography means absolute freedom at the expense of a lack of sharpness. So, enjoy the freedom of pinhole photography : may be the most impressive pinhole image I've seen at the conference was a self-portrait of a pinhole photographer taken through a small hole in the middle of his fingers. In this image, emotion is here and maths, really far away.

[Capocheny]

Hello Emmanuel,

Thank you... math makes my head hurt! :)

I'll just have to play with the design and see what comes out of it. I must admit I'm really quite excited about the prospects of seeing my first image.

The reason for this intrest in pinhole photography is that I went to a gallery opening a few weeks back. I was absolutely stunned by the beautiful images I saw and how much depth of field there was in those images. The photographer is Diane Bos and she shot the image of the gargoyle figure for the UPS ads that you see on television and in newspapers every so often. I'm not sure if the exhibition is still on at the gallery or not but you may be able to see some of her images online at www.kostiukgallery.com

Again... thanks kindly for the information. It's greatly appreciated.

Are you still shooting with a pinhole camera? I do use a Sinar X 4x5 but it's really not all that conducive to carrying around out in the field. Hence, I thought I'd give pinhole LF a try. After I succeed with the 4x5 version... I want to move ahead with an 8x10 version using photographic paper instead of film. :)

Cheers

[Emmanuel BIGLER]

I gave a try to pinhole cameras a long time ago but did not continue, attending this local pinhole conference recently made me think again to the physics of the device and I'm considering to practise it again with my children.

Using B&W photographic paper instead of film works fine except for two points
- classical graded B&W paper is blue-sensitive and not panchromatic ; variable contrast papers have a somewhat extended sensitivity to green but I do not know the results on the actual contrast of the image. Allow an exposure index of ISO 1 to ISO 10 in daylight. In artificial light actual sensitivity is smaller because there are much less blue rays emitted by a tungsten bulb. Papers are now rated with and ISO-P (like P-aper) sensitivity index which is roughly 100 times the value of the film ISO sensitivity. an ISO P 500 paper has roughly an ISO 5 sensitivity if used like a film in a camera.



- the image will be.. a negative of course and will be left-right reversed except if you add a mirror in front of the camera. this point is very minor ; you can scan the image with a flatbed scanner and do both the negative/positive 'reversal' processing as well as correcting the left-right symmetry by simple digital image post-processing. But you may of course prefer to work like Fox Talbot and make paper-to-paper contact prints, the final print will be a positive and will be left-right corrected like the original ;-)

[Emmanuel BIGLER]

Information about the World Pinhole Day
www.pinholeday.org (http://www.pinholeday.org)

The Worldwide Pinhole Photography Day is held each year on the last Sunday in April.
The next Worldwide Pinhole Photography Day is on Sunday, April 24, 2005.

[Capocheny]

Ennanuel,

Good to hear that you're considering getting back into pinhole photography again... did you, per chance, take a peek at the link I posted for Diane Bos's images? IMHO, they were simply beautiful images and that's what is inspiring me to experiment.

According to some exposure tables I've seen from the internet...they recommend using classical graded paper at an ASA rating of 6 on a handheld meter (and adjusting for reciprocity failure.)

I'm hoping to get my pinhole camera designed and built for the next Worldwide Pinhole Photography Day... it should be great fun.

Cheers

[Rob Hale]

Hi Emmanuel

Am I completly up the creek or are these hole sizes and f numbers back to front
i.e. should the larger pinholes not be a smaller f number or is it a relationship to the focal length?

“I've been told that using a #8 needle produces a diameter of .023" (061
mm) providing a Normal (7") focal length and an f-stop of 330. So, if I
wanted to obtain the sharpest image with the least amount of distortion...
this sounds like the way to go then. Would I be correct in this
assumption? I'm under the impression that the smaller the hole, the
sharper the image. Therefore, perhaps, a wide-angle is the way to go?

I was also told that using a #10 needle to produce a diameter of .018"
(0.46 mm) would produce a wide-angle (5") focal length with an f-stop of
280. According to the discussion above... can I expect a fair amount of
distortion at the edges then?”

Regards Rob.