About Diffraction

[Leonard Evens]

Same analysis for 4 x 5?

Indeed, as far as diffraction is concerned, the degree of enlargement before seeing a difference doesn't depend on the format. So wg's remarks would still apply for 4 x 5. The only way the format enters into the discussion is in the degree of enlargement necessary to produce a given size print. Assuming the same size print, you need to enlarge about half as much with 8 x 10 as with 4 x 5, so you are able to stop down further before the effect of diffraction would be apparent.

Let me quibble about a couple of other points.

First the reciprocal rule is not the only way to combine resolutions. Many people prefer to use the root mean square method: square the reciprocals, sum them, take the square root and then take the reciprocal of that. In reality, all these methods must be considered rough rules of thumb. None of them has any justification from optical theory. The only way to do it accurately is to multiply MTF functions for the different components of the optical system, but that is much too hard to do routinely. Jacobson has an example of such an analysis in his Lens Tutorial, which can be found at the photo.net website. Hence, you should only use such rules as a rough guide. In critical situations, you have to test and see what happens.

Second, do you really want to base your decisions on what the print looks like from 25 cm? Some people do have an irresistible urge to get their noses right up against any print, no matter how large. But, if you think about it, that places a severe upper limit on the size of the print, even if you use an 8 x 10 negative. Normal viewing distance for a print is usually taken to be the diagonal of the print. So an 8 x 10 print might be viewed from about 12 inches and a 16 x 20 print at 24 inches. As you get further away from the print, you can tolerate progressively less resolution, and if an 8 x 10 print looks sharp at 12 inches, a 16 x 20 print, made from the same negative, will look sharp at 24 inches. An 8 x 10 print will not show appreciable diffraction effects even at f/64, and it follows that any larger print, viewed from the appropriate distance will also not show diffraction effects at usable apertures.

A small peeve. There was lots of helpful advice, but I had to go to the sixth response to find one that actually answered the question. This seems a common occurence. Someone asks a sensible question, and people answer by explaining why the question is foolish and why he/she should be doing something else. I just did it myself in the above paragraph. I don't object to that kind of advice, but perhaps one should also make an effort to answer the question first.

[bglick]

Now that the question has been answered, a few tid bits to tidy this up.....

Chad, glad we got you in the ballpark...and as I have mentioned, and Leonard supported me, in the ballpark is at best + / - 50% , more on this later......

Henry, I think Leonard answered your 4x5 question.... All things remain equal, IF, you open up 2 stops for each format, (as its half the fl) which then gets eaten up by the extra 2x enlargement required to get back to the same final size, this is at the plane of sharp focus.

Leonard, as for the reciprocal method vs. root mean square method.... agreed, i have quibbled with both myself..and in the end, they produce relatively similar answers, so not a huge deal, I stick with 1/R as its supported by Fuji, and it's their film I use :-) But regardless, since the purpose is only trying to land in the ballpark, it's not that signficant.

Also, Leonards comments on viewing distance are VERY important consideration often overlooked, in my example, I assumed the closest focus distance, whatever that is for the viewer, 10 - 18" based on how close the person can focus. But as a general rule, you double the viewing distance, you can half the resolution and have the same net effect of apparent sharpness. This is why 35mm film can be used to shoot billboards. The ideal viewing distance is often the image diagonal. But that does not stop someone from walking up close to a 60" print, it's just human nature to inspect even further.... specially in this high tech era where everyone has a mangifying glass on everything, computing speed, storage capacity, Mega Pixels, etc. etc.....

As for the application of this math to photo optics..... when i started in photography, I too was fooled into the application of otpical math to determine outcomes. And although there is some very basic and fundamental uses for the math, which can at least get you in the ballpark, there is rarely a precise means. As Leonard pointed out, having detailed information on the lens MTF data and the film MTF data is the closest you can get to achieve a "non tested" outcome. However, rarely if ever, in the photographic world do these two pieces of information exist complete enough form, to arrive at such a conclusion. And even then, it's not quite clear how to combine the two to get the optimum outcome, although there is several theories on such. However, the MTF theory is the ONLY method that ties together YOUR lens and YOUR film, and although it does not include many of the other factors in the image taking process, it does give you some ceilings to work with.

The application of optics math to generic lenses, well, this is where things fall apart. If you take the standard Depth Of Field equation to find out your focus point, so the near and far acheive a certain resolution....well, now IMO, you have enter voodo optical math. As for the formula, it doesn't know if you are using the sharpest lens made with the sharpest film made of the modern era, or 1800's vintage lenses and film.... it does not know if you are using a lens optimized for close up work, but yet shooting at infinity. It does not know if you are shooting at an f stop the lens was NEVER designed to be used at, as all lenses are designed to reach their optimum performance over a 2 - 4 f stop range. So what do you end up with? At best, you end up with some theoretical value that is often useless, and this is where testing becomes mandatory. The DOF tables or scales used with certain cameras were sometimes created with testing, and sometimes theory, different enlargement factors were considered, different films were used throughout different eras, etc. etc... hence why this has become an ever changing and evolving area of photography, but fortunately for the better.

Although I was quite befuddled without the joy of pluggin a few numbers into a spreadsheet and seeing the outcome of a given photographic situation, I have finally come to accept it... very painful indeed.... but, now I work in a reality based world.... :-)

[Capocheny]

Leonard, wg...

Thank you... I appreciated your responses.

Cheers