As I understand it, the resolution in the final image is the lower of
the sensor resolution or the lens resolution.
With film, by using a larger negative, you needed less grains per inch
to get the same final resolution. Likewise, your lens needed to
resolve fewer lines per inch.
With digital, using a larger sensor means that you can either cram
more pixels in at the same density, or get larger pixels with the same
image resolution. If you use larger pixels, you get more dynamic
range, and likewise your optics don't need to resolve as sharply to
match, or exceed, the resolution of the sensor.
To a first approximation, the cost of sensors follow Moore's law (or
more accurately, his observation). In short, as time goes on, sensor
performance gets cheaper.
With glass, Moore's law affects the design phase of the lens. It
becomes cheaper to design better optics. In 1973, I remember Dad
bragging about how much time on a Cray was used in designing the
Series-1 70-210 that he bought for his Spotmatic. There's a good
chance that if they were properly ported I could run all of those
calculations on my cell phone in a matter of hours.
There have likewise been other advances in the materials and
manufacturing of optics, but not at the exponential rate of the
sensors. My guess is that the price/performance of glass is improving
at a linear rate, not the exponential rate of sensors.
In short, silicate performance used to be cheaper than silicon
performance, but soon the exponential curve will cross the linear
curve and silicon performance will be cheaper than silicate
performance. I'll put myself out on a limb and say that the next step
will be more and more correction of optical problems in post processing.
However, as to the sensor size question, the most efficient use of
your performance dollar will change based on what you're trying to do,
and when you try to do it.
If this seemed to be a long winded nattering around the issue, just be
thankful that I didn't sink my teeth into the topic of "good enough"
performance.
--
Larry Colen [email protected] sent from i4est
--
PDML Pentax-Discuss Mail List
[email protected]
http://pdml.net/mailman/listinfo/pdml_pdml.net
to UNSUBSCRIBE from the PDML, please visit the link directly above and follow
the directions.
