Started 2 months ago | Questions thread

JimKasson wrote:

Erik Kaffehr wrote:

Hi Jim,

The stuff you are sharing is very good. But, it is based on a basic understanding of basic principles of photographic imaging.

Back in 1913, when Oscar Barnack developed the first Leica he counted what kind of resolution was needed to achieve the quality of a printed postcard of that time. He arrived at about 1/1500 of the image diagonal. The original Leica had 24x36 mm image size with a 43 mm diagonal. 43 / 1500 -> 0.028, or something like 0.030 mm, this is the magic number that DoF calculators use.

So, that essentially says that if we photograph a star image, a spot on the sensor 0.03 mm wide would be perceived as sharp.

Now, digital cameras have pixels. Pixels are mostly rectangular. So how large would that pixel be?

Let's calculate the area of the 0.03 mm spot, that would be Pi * (0.03 / 2) ^2 -> 0.000707 square mm.

How large would a pixel be that had the same area? sqrt (0.000707) -> 26.6 microns.

So a 24x36 mm sensor with 27 micron pitch would match Oskar Barnacks criterion.

How many megapixels would that be? The answer is:

24 / 0.027 * 36 / 0.027 / 1e6 -> 1.2 MP. So, we need like 1.2 MP for a postcard size image. That was true in Oskar Barnack's time and it is true in our time.

So, what is different now? Some things, but not much.

If we look at a 'Bluray' movie, it has a resolution of 1920 x 1080 pixels. That is around 2.1 megapixels twice the information Oskar Barnack found 'good enough' back in 1913.

Just to say, most viewers find 24" full HD screens sharp enough. If we judge sharpness on a 24" monitor we see around 2MP. Oskar Barnack's 1500 CoC across the diagonal shake a bit.

But, now we go 40-50 MP on 24x36 mm or 33x44 mm. The old requirements for sharpness are as valid as ever.

Except, we print (or display) large and view close. A 4K 24" monitor will be not more demanding than a full HD (2K) 24 monitor, unless we view it at half distance.

So, now we have high resolution, we need to change the way we see images to make best use of it. The recipe is to print (or display) large and view close.

Once we do it, we also need to change our criteria for good sharpness.

I can't disagree with any of that.

Thanks Jim!

My point is that you cannot understand the future without understanding the past.

Many of the recommendations, like 0.030 mm CoC are based on the past. There is nothing wrong with that, those recommendations still serve us well.

But, even cell phones are good enough to fulfill those recommendations. I would say that cell phones are just great for 'f/8 at 1/125s and be there' kind of images.

But, if we want to deliver beyond 'f/8 at 1/125s and be there' we need to learn and understand new criteria, leading to new approaches.

Just to say, 'f/8 at 1/125s and be there' has given us a lot of great images, think David Douglas Duncan, Cornell Capa, Elliot Erwitt or Ernst Haas. Well, regarding, Ernst Haas was probably not in the f/8 at 1/125s club...
With the new camera systems, we may be in search of something else, like 'excellence'. Looking at excellence, we may need to reexamine the rules.

Playing the old f/8 at 1/125s rules on modern cameras may be suboptimal use.

Best regards

Erik

-- hide signature --

Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles

Complain
Post ()
Keyboard shortcuts: