View Full Version : 7 lines per mm?

02-14-2006, 03:50 PM
Is it possible to make a pinhole camera which will give you the magical figure of 7 lines per mm needed for a sharp print?

I know that ive been discussing various aspects of this in other threads, anyone not bored with the topic of resolution optimisation (which is a side issue as the many beautiful images posted on this site testifies) please pass on to more fun questions... thank you!

02-14-2006, 04:01 PM
you know what i ment, right :)

02-14-2006, 06:42 PM

Pinhole photography does not lend itself to absolute measurements of resolution (e.g. lines / mm). The smallest granule of detail on the film image is limited by the diameter of the pinhole - there cannot be detail in the image smaller than the hole. Where pinhole image resolution is impressive is in relative resolution, that is the amount of information per sqare inch of film area. Since the optimal pinhole size, as limited by diffraction, increases proportionally less than the area illuminated as the pinhole to image distance increases, the relative resolution goes up as the distance and image size goes up. In order to get 7 lines per mm, you would have to have a pinhole .143 mm or less, and that limits your pinhole to film distance to just over 10 mm, which results in a pretty small image. But a pinhole photo on 8X10 at say, 75 to 150 mm distance, with an optimal pinhole has many times the information in the image than the small image, which at an appropriate viewing distance, make it look sharp to the eye.

02-14-2006, 07:06 PM

thanks for your infomation. basically ive been having a good natured debate with a friend who shoots 8x10 lens photography. i shoot 8x10 with a pinhole and we both contact print. we were interested if one of my contact prints could theoretically approach one of his contact prints for resolution and sharpness such that an objective viewer would not be able to tell the two prints apart. i think it would be possible at 20x24 but im not sure if its possible at 8x10. not that i think that pinhole is in any way inferior; what my friend would give for my DOF!

02-14-2006, 07:16 PM
The absolute shaprness of an 8X10 pinhole photograph can never be compared to an 8X10 negative made with a lens. It can be compared to an 8X10 enlargement made from a much smaller negative, though. Pinhole photographs are not sharp in the absolute sense - they are just sharp enough too look great at a normal viewing distance. And as you say, the sharpness of the parts of the pinhole image that are outside of the depth of focus of the lens are more than comparable.

02-14-2006, 10:28 PM
I thought lenses were diffraction limited after stopping down.
Formula was something like 1600/(f/stop).
So at f/64 the best you could hope for would be 1600/64 or 25 lines.

This formula would yield 7 lines at f/228 - which is definately pinhole territory.
At some point does the glass become useless?
If the glass ever becomes an impediment - then that would mean that there is a region where
pinhole would outperform glass.

I was looking at http://nvl.nist.gov/pub/nistpubs/jres/104/5/j45mie.pdf and promptly got a headache.

02-14-2006, 11:32 PM
Hmm. Well, oversimplifying (by ignoring diffraction), you'd need a hole of around .07 mm diameter to give 7 lp/mm resolution. That's optimum for about a 3 mm focal length, suggesting there'd be problems with image circle size... ;)

02-14-2006, 11:44 PM
Uh, yup. That's the nut I'm trying to crack but it's so @*#@ abstract it's hard to make enough sense out of.

What I got out of this article is that it's hard to distinguish between the various subtle choices of 'optimal' constants (Rayleigh vs. others) because it's so subjective, but that it IS definitely in one's interest to make an effort to be in the 'ballpark'. But you already knew that before trying to read this article.

I found another article that said contrast is a factor and some images look sharper with an unexpected diameter than predicted...but he used computer simulations instead of real images...I don't know if that's right.

I didn't know one could get 7 lp/mm with pinhole. I have been looking for articles that discuss this quantitatively, and thus far I didn't think I had seen anything referenced higher than 2.5-3 lp/mm.

I wondered if reducing an 8x10 to 4x5 does this because we see some tricky images posted that look TOO sharp to be credible pinhole images sometimes :)

I think that diffraction limiting with too-small an aperture and glass is still a different animal than a pinhole because I think (my opinion) the pinhole can only produce a remotely acceptable resolution when it is close to it's optimal circumstances...diameter, spacing , wavelength.

A lens is still doing some refracting but has diffractive problems at too small an aperture. I don't know if diffraction limiting is one of the evil 7 aberrations. Smaller apertures correct a number of those aberrations independently (mathwise).

What I'm still working is understanding how image formation shares similarities in both pinhole and glass...both are described (apparently) by the same 2-D Fourier Transform model. It seems that pinhole can only produce a decent image where this model is optimized, yet the 2DFT is relevant (that's all I understand so far) for image formation with large apertures in glass far from the diffraction limit.

If I get that part, I have my math class paper started.

I think the choice if 1600 or whatever number one chooses/f-stop has alot of assumptions and qualifiers to go with it. I don't know how far one can run with it, and I have to keep re-reading why format plays a role in determining where diffraction limit is...if it's based on the sliding circle of confusion scale, that may be why it's so slippery....it's based on an assumption of acceptance and maybe viewing distance...I'm not sure

02-15-2006, 04:41 AM
Hi, the math is certainly confusing however there are definately some pinhole photographers that seem to atchive sharp(ish) prints without cheating eg. dianne boss and martha casanave...

eg. one of dianne boss's recent ones... Attached files http://f295.f295.org/uploads/dianneboss1_9897.jpg (http://f295.f295.org/uploads/dianneboss1_9897.jpg)

02-15-2006, 04:45 AM
...and one from martha casanave, beautiful photographs im sure you will all agree... just wish i knew how they do it!

Attached files http://f295.f295.org/uploads/marthacasanave1_5409.gif (http://f295.f295.org/uploads/marthacasanave1_5409.gif)

02-15-2006, 05:05 AM
This looks like what I ohh and ahh over and find out later are reductions from larger negs...I don't if this is the case for this work, however.

I think the dramatic contrast (filtered?) enhances the overall discrimination between tones...I don't know if this appears sharper or gives more visual information to look at, drawing one away from sharpness question.

02-15-2006, 06:41 AM
This would be fun:

Shoot with an 8x10 pinhole camera.
Reproduce onto 35mm film with reproduction camera.
Compare to 35mm lens shot of the same scene.

02-15-2006, 06:51 AM
The dianne boss photograph i posted is at 20x20 size and is selling for $1500!


02-15-2006, 08:06 AM
"we see some tricky images posted that look TOO sharp to be credible pinhole images sometimes"


are you suggesting that sharp pinhole pictures can only be produced by cheating?

02-15-2006, 08:50 AM
mesmer - I suggest that you stop thinking about theory and resolution and other people's images and GO OUT AND SHOOT SOME FILM!

02-15-2006, 08:58 AM

I do... and I like talking about theory and resolution and other peoples photographs which I thought was the purpose of this forum?

02-15-2006, 09:04 AM
... although I do think to much for my own good ... blame years of academia ...

02-15-2006, 01:48 PM
Message deleted for lack of empirical evidence.

02-15-2006, 07:50 PM
"are you suggesting that sharp pinhole pictures can only be produced by cheating?"

NAH, but there are some that just nail it and it makes me wonder what is different...psychooptics, maybe.

No, I'm not calling you a psycho OR a cheater. :o

02-15-2006, 07:59 PM
I agree with the above interpretation. I wanted to explore a 'bias' toward a smaller constant in the aperture diameter optimization formula, because there IS a subjective factor there in the choice.

Apparently one factor (not sure if attributable to Rayleigh directly) is the claim that the smaller ideal apertures produce impractically long exposures. It occurred to me that this need not be a restriction with modern films...in the 19th century, yes.

Mlelenz took the trouble to recalculate the dreaded table of functions & graphs at Lommel did before Rayleigh because computers and electricity have an advantage over the inkwells and candles of the day.

I am working two jobs & goingto school right now, so that's why I ain't shootin' more film. I'm very curious whether people willing to blaspheme (reduce) the assumed optimization constants will get better resolution...living with the longer exposures. We have the advantage of fast emulsion and understanding reciprocity failure correction...or, bypass that by using Tmax or chromogenic C-41 film in robust lighting.

02-15-2006, 08:42 PM
This is getting a bit too speculative.
We need empirical evidence!

Here is a shot you have all seen before. This was with the 57mm f/207 and was originally scanned at 2400 dpi. I wanted to take a look at the detail of the railing, which has nice vertical elements every 8 or 12 inches. They are about 30 yards away from the camera’s position. Now I believe 2400dpi is really close to 95dpmm, so I set the burn tool to a 95 pixel circle and clicked it a few times in the detail crop so we can all see just what is inside that little millimeter.

Does this look like I just might be getting more that 3 lines per mm? The diffraction limit for lenses at F/200 is 8 lines.

Oh, wait a minute! – isn’t there someone in the mathematics department that can help me here? There are numbers to be crunched! I could be WRONG! Somebody please prove me wrong, as it seems sometimes to be the only way that I can learn something. DOH!

I believe we have a new project for someone better qualified than myself:
Resolve this resolution question!

We need a volunteer!
Do we have any takers?
How about taking some exposures of resolution charts in a controlled environment?
The USAF 1951 chart is at http://medfmt.8k.com/mf/USAF.pdf
While you’re at it, take some shots with and without a filter!

Get to work!
Attached files http://f295.f295.org/uploads/frame4c72stmarysorthodoxrail_4819.jpg (http://f295.f295.org/uploads/frame4c72stmarysorthodoxrail_4819.jpg) http://f295.f295.org/uploads/stmaryorthoframe4c2400raildetail_1070.jpg (http://f295.f295.org/uploads/stmaryorthoframe4c2400raildetail_1070.jpg)

02-18-2006, 03:05 AM
I'll bet you had better resolution on the film than the scanner showed you.

Local lab has a $55k scanner. I can zoom about 6 x higher in PS, etc. with their scans than with my old consumer one, althought I'll bet my Agfa is better than the last three cheapest ones that I could find on sale.

Chris Perez has some interesting resolution test info on the web (hevanet site).

And here's an easy to read article that pitches an argument on optimizing pinhole for MTF rather than resolution. www.biox.kth.se/kjellinternet/Pinhole.pdf 2.56MB pdf

Looks like his pinhole formula constant is 2*0.78 = 1.56 vs 2.0 many people use.

02-18-2006, 05:43 AM

Thanks for the excellent and also funny paper!

Have you photographed using a pinhole designed around the smaller constant? What kind of result did you get?

Its going to take me some time to digest the details but it also looks like the larger optimum pinholes associated with longer focal lengths and larger film sizes (ie a 20x24 'normal' camera) would give the sharpest results. I know of one photographer who has converted his wheelie (trash) bin to hold a sheet of 5x3 ft photograpic paper who claims that he can produce images as sharp as a 35mm!

02-18-2006, 05:59 AM
The photographer with the 5x3 foot bin camera is justin quinnell. Its obviously a bit hard to scan a 5x3 ft photograph so I cant verify his claims that he gets the same sharpness as a 35 mm camera but hes a pretty well known figure over here. Still, his experiment needs to be replicated!

02-18-2006, 06:04 AM
the details on justins bin camera are at


02-18-2006, 11:52 AM
APUG has a similar thread running on pinhole resolution at:


Jim Jones posted this:

"Donald Qualls' comments agree with my research. The basic simplicity of pinhole photography has lured many practitioners into believing that precision is unnecessary. Where optimum sharpness is needed, the diameter of the pinhole should be calculated, and the pinhole fabricated (or purchased) within a few percent of the ideal size. When the diameters of the pinhole and the Airy disc are nearly the same size, there is a distinct increase in on-axis resolution. Then, line pairs smaller than the pinhole diameter can clearly be resolved. The image suffers from reduced resolution and slight astigmatism at angles away from the axis. Maximum pinhole image quality is obtained with large format wide angle cameras. Enlarging the pinhole of a wide angle camera fairly well evens out the sharpness across the image at a considerable loss of on-axis resolution. Fortunately, a good photographer can often create fine pinhole photos with little regard to the quest for optimum resolution.

Traditional hyperfocal and DOF formulae don't always apply to pinhole photography. The discrepancies between the many charts and formulae for pinhole diameters should be resolved by each photographer based on his own preferences. Someone with a modest understanding of physics and math and a drive for perfection may find an in-depth study of pinhole optics rewarding. Others may find just making pinhole photos more useful."

I included the entirety of his comment, but my interest was his statement concerning an increase in on-axis resolution, near the ideal pinhole size, such that line pairs smaller than the aperture size can apparently be imaged.

I suppose there's no way of knowing if this is true "for sure" without having a solid grasp of the physics involved. Perhaps someone could comment about this.

02-18-2006, 07:05 PM
I included the entirety of his comment, but my interest was his statement concerning an increase in on-axis resolution, near the ideal pinhole size, such that line pairs smaller than the aperture size can apparently be imaged.

I suppose there's no way of knowing if this is true "for sure" without having a solid grasp of the physics involved. Perhaps someone could comment about this.


I agree with your skepticism. It doesn't make sense when I think about it in the abstract. I read through Murray's cited article with great interest, though - after all I have all of these 'too small' pinholes to try to find something to do with. In the short run, I am very interested in the physics (which I am not equipped to understand, but which fascinate me immensely), but in the long run, I am most interested in the images. I guess that I don't care how many lines per mm I can image as long as I can make pictures that I like. The calculations for me are indications of how far I can push things before I can expect the image to go all wambly on me - and that might be just what I am looking for.

02-20-2006, 03:54 PM
. . .

And here's an easy to read article that pitches an argument on optimizing pinhole for MTF rather than resolution. www.biox.kth.se/kjellinternet/Pinhole.pdf 2.56MB pdf

Looks like his pinhole formula constant is 2*0.78 = 1.56 vs 2.0 many people use.

Thanks for that link Murray! That's a keeper.

02-22-2006, 05:35 AM
going to try a smaller-than-optimum pinhole in my camera and post the results....

02-22-2006, 05:40 AM
alternative explanation for why smaller than optimum pinholes give sharper photographs already posted...

Since paper and ortho film are sensitive to blue light, and relatively insensitive to red, wouldn't this automatically be like using a blue filter? Maybe that's why the pictures with my 250mm camera seem so sharp even though I'm using a smaller than optimum pinhole?