A reader asked “Isn't 24 megapixels inadequate in comparison
to the competition?”, to which I answered:
“The short answer is “no.” The megapixel race, while once important,
should have come to an end a while ago already.
But it is oh so salable. The
buying public in general has been conditioned to think that all it takes is
more megapixels to increase camera performance.
Nobody is ever talking about lenses, yet it is the lens that ultimately
delivers the image to the sensor. To
make sensors with a resolution that surpasses that of the lenses makes no
sense. The ever increasing number of
pixels of a full frame sensor, for instance, naturally necessitates making the
individual pixels smaller to allow them to be all fitted in the full frame
sensor space. At one point the
individual pixels become so small that they exceed the resolution capabilities
of the lenses used in conjunction with that sensor. With other words, we have a theoretical
resolution that cannot possibly be met.
At that point all that’s left is advertising potential.
I wrote about that very issue
a while ago in an article, titled:
“THE MEGAPIXEL RACE - ARE
THERE LIMITS?”
Go take a look. It should answer your question.”
That is very interesting. What is the resolution of Leica lenses, and how does that compare to the resolution of the human eye. With other words, aren't some lenses sharper than what the human eye can see?
ReplyDeleteThe 180mm f/3.4 Apo-Telyt-R is capable of a resolution of 600 lines per millimeter. That means the lines are 1/600 mm wide. That corresponds to 6 microns. The new Leica 24 megapixel sensor has an individual pixel size of 6x6 microns. Surprised? This effectively means that it requires the full resolution of the Apo-Telyt-R to take advantage of the sensor. It is generally assumed that this is the limit of resolution of most Leica lenses. That is a level of sharpness achieved by few if any other lenses on the market. With other words, any higher pixel count appears to be beyond what lenses can deliver. Of course there are losses that we have to accept. The bayer filter leads to certain losses as does interpolation, for instance.
ReplyDeleteAs for the human eye, the limit of resolution for perfect vision is 10 lines per millimeter (lpm). However, since perfect vision is something few people experience, 7 lpm is generally accepted as a more accurate figure. That is certainly substantially less than what camera lenses are capable of. However, we must also consider that the lenses create an image on a 24 x 35 mm area. For general viewing, these images have to be enlarged. Obviously, the bigger the enlargement, the less resolution in terms of lpm we have. The same is the case when cropping an image. Subsequently, a high resolution sensor or negative is always preferable.
Have any tests been published that support the 600 lines per millimeter figure for the Apo Telyt R?
ReplyDeleteYes, Joseph A. Schantz, Assistant Head of Research and Development Department at the Navel Photographic Center in Washington, DC. published a paper about the research that the US Navy conducted for a high resolution 35mm camera system. For that four lenses were developed by Ernst Leitz Canada, one of which was the 180mm f/3.4 Apo-Telyt-R. The lens was initially classified for several years. It was after its declassification that the lens became part of the lens line for the Leica reflex cameras. This was covered in an article on this blog titled:
DeleteMANUFACTURE AND PERFORMANCE OF PHOTOGRAPHIC LENSES
http://gmpphoto.blogspot.com/2012/02/by-heinz-richter-few-industrial.html