Pinhole cameras

[thebbqguy]

I know pinhole cameras are the most basic things, but I've seen a large variety in results.

Some photos are quite sharp compared to the typical lomo results I see most often.

Are there threads here about pinholes? I've looked a lot but didn't see any specific to pinhole alone.

Any suggestions?

Here's an example of pinhole photos I really like.

http://www.kirtecarterfineartphotography.com/#/gipfel/

I'd like to learn to take photos like these.

[Rayt]

I have the Harmon Walker 4x5 pinhole and was surprised by the quality of the images, not fuzzy at all. Those images in that website look digital sharp.

[thebbqguy]

Yes. I've noticed some very high quality images from some users of pinholes.

That Harmon 4x5 pinhole is not inexpensive, but it seems very effective.

[John Kasaian]

All I can say is to never make one from a frozen turkey. Ever.

[Scyg]

A properly sized pinhole on an 8x10 can give you pretty decent resolution. Here's one of mine done recently:
203282
The blurriness of the branches at the top is mostly motion, not distortion.
Here's a 1:1 300dpi fragment:
203283
These are paper negatives, so the tonality is limited, but judging by film I've shot on similar formats, it can be really nice.
The camera is a scratch-built box (a.k.a. Tree Camera™) with lots of rise for those frog's eye perspectives:
203281
I can't help thinking that the site in your original post uses some kind of computer juju to sharpen them up. They seem awfully crisp for 4x5s.


All I can say is to never make one from a frozen turkey. Ever.
Now that you've said it out loud, I was thinking...

[thebbqguy]

Yes. I contacted the photographer. He confirmed that those images were processed in Photoshop to clean them up.

[Jim Jones]

Designing pinhole cameras involves compromises. The Harman Titan is optimized for the sharpest practical 4x5 pinhole images. This means wide angle coverage which is not ideal for many subjects. 4x5 cameras with increasingly long focal lengths have increasing loss of detail. Decades ago I fabricated a pinhole camera using 4x5 and 5x7 film and with a focal length of 25 feet to photograph a solar eclipse. The pinhole diameter was about .11 inches, and the negative was so unsharp that scanning it at eight pixels/inch might have retained all the detail of the 2.5 inch diameter image of the moon. In comparison, a 4x5 Harman Titan would require scanning at maybe 100 pixels/inch to retain all the image detail. That is a more extreme focal length range than we encounter in practical pinhole cameras, but it demonstrates the effect of focal length on sharpness.

[reddesert]

The sharpness of a pinhole image is limited by
1) the diameter of the pinhole - the image will have a blur spot size of at least the pinhole diameter
2) diffraction, which makes a spot size proportional to the f-number. f-number = focal length/aperture diameter.

That means for any focal length, there is an optimum pinhole diameter - too large and you get more blur from the pinhole, too small and you get blur from diffraction. Additionally, it works out in favor of larger formats and relatively small enlargements, when comparing the same angle of view. That is, if you want a normal-wide field of view, you could get a blurry Lomo-pinhole-ish image from a 35mm camera and a passable image from a 4x5 (but don't expect to enlarge it to make a mural print).

The pinhole calculators you can find on the web take these effects into account.

[Scyg]

The pinhole calculators you can find on the web take these effects into account.

These things aren't quite as cut-and-dried as a lot of web-based wisdom would make it seem. There's a different quality to the softness you get from a big hole and from diffraction. I've always had best luck with pinholes just a hair smaller than the "scientific" optimum given by web calculators.

[reddesert]

These things aren't quite as cut-and-dried as a lot of web-based wisdom would make it seem. There's a different quality to the softness you get from a big hole and from diffraction. I've always had best luck with pinholes just a hair smaller than the "scientific" optimum given by web calculators.

I guess I didn't mean to say that the pinhole calculators are exact. Different ones give slightly different answers, and there's very little exactitude in pinhole photography.

In the ideal case, the illumination produced by the pinhole is a circle of uniform illumination with sharp edges, the blur made by diffraction is an Airy function (which has a peak and an outer ring), and the spot actually produced is the convolution of the two functions. I think that most of the online calculators use some approximation for the diameter of the Airy spot, and that hardly anybody who writes these calculators has actually tried integrating the two functions.

In practice, pinholes (especially homemade) are rarely perfectly sharp-edged circles, the thickness of the pinhole material should matter a little for the diffraction, and so there are a number of reasons why the calculators would be just a guideline. You can play with them to see the effects of making the diameter, focal length, or the format larger or smaller, but achieving optimum results will take, as you suggest, some experimenting.

I once made a six foot long pinhole camera for viewing (not photographing) a solar eclipse. I used a cardboard box for skis and cut a window at one end so people could look in to see the projection of the sun's image. It wasn't the most perfect image, but it was a good educational tool, in part because everyone was curious about the giant box.

[thebbqguy]

Is it unreasonable to enlarge a "good" 4x5 pinhole image to 18 x 18 for printing?

[Scyg]

Define "unreasonable". I've enlarged 35mm pinholes to 8x10, which would be a far greater enlargement factor. They certainly didn't look like 8x10 contact prints, but that wasn't the point of doing it.

[Jim Jones]

The sharpness of a pinhole image is limited by
1) the diameter of the pinhole - the image will have a blur spot size of at least the pinhole diameter
2) diffraction, which makes a spot size proportional to the f-number. f-number = focal length/aperture diameter. . . .

An interesting phenomena occurs when a pinhole diameter is near the crossover size for minimum diffraction limited and geometry limited image blur. At that point, diffraction actually enhances sharpness, and the image of a 1951 USAF test chart will resolve line pairs finer than either diffraction or geometry predict. This is explained better here: http://inside.mines.edu/~mmyoung/PHCamera.pdf. This is most true for images on or near the pinhole axis. From near the edges of pinhole images, the pinhole appears elliptical, not round, and this introduces astigmatism as well as increased diffraction limitation. This can be seen in the chart below:
203371

[Scyg]

With very wide angle cameras it is also worth considering the ratio of the pinhole's diameter to its thickness - you're effectively passing light through a tube, and the longer the tube, the more constrained the light path towards the edges of the image. For my panoramic cameras I ended up using .0005" (that's half a thousandth of an inch) stainless steel shim stock, which significantly improved performance at the edges.

[Greg]

An interesting phenomena occurs when a pinhole diameter is near the crossover size for minimum diffraction limited and geometry limited image blur. At that point, diffraction actually enhances sharpness, and the image of a 1951 USAF test chart will resolve line pairs finer than either diffraction or geometry predict. This is explained better here: http://inside.mines.edu/~mmyoung/PHCamera.pdf. This is most true for images on or near the pinhole axis. From near the edges of pinhole images, the pinhole appears elliptical, not round, and this introduces astigmatism as well as increased diffraction limitation. This can be seen in the chart below:
203371

Thanks your for post. For my pinhole cameras I usually make around 20 pinholes and test them at the focal length that I am going to use them at with different frames on a roll of 35mm film. Inevitably, even though they differ by the slightest in diameter, one pinhole always produces an image that is superior to the others. That is the pinhole that I use on my LF or ULF pinhole camera. Instead of a USAF test, no longer have one, I use a stop sign about 40-50 feet away.

[Jim Jones]

Greg -- You may find this link useful: https://jimdoty.com/learn/lenses/usaf_test/usaf_test.html.