Presumably, Pentax came up with a fairly accurate model for real-world
camera shake, maybe basing it on measured data from one or several
real photographers.  Then they designed a shake reduction system to
correct this model camera-shake as well as possible.  Finding that
their system corrected the model camera-shake enough to allow 4 stops
longer exposure, they advertise it as such.

But of course, the camera shake from *me* holding the camera is
probably a little bit different from the camera shake when *you* hold
the camera, and both of us are probably not quite the same as the
model that Pentax used.  So results vary from individual to
individual.

Dan

Also, I choose a camera in with in body SR because I have a nerve disorder
that causes me to shake. No matter how hard I try to hold the camera steady
it shakes. I can promise you that Pentax didn't line up a bunch of people
with various nerve disorders to do lab testing. Benchmark testing has always
and still is a flawed way of proving a devices capabilities. Take a computer
for instance. The people running the benchmark test it with Windows running
alone, the Windows with Photoshop, then Windows, Photoshop and Norton
Anti-Virus. Well that is a very small amount of testing. They didn't test it
with Windows, Photoshop and MacAfee, some other anti-virus. They didn't test
it after running 3 years straight with normal user use (installing and
uninstalling software, etc.). What they test the system under in benchmark
testing is so small that it is worthless for real world comparison.

Even software reviews suffer from this. The results ones would get from
installing a software program on a clean system and using it for 20 minutes
is a far cry from someone installing it on a system with tons of other
stuff, using it for several times a day for 2 or 3 hours for several months.
The reviewer doing the test on a clean system is going to run in to far
fewer bugs, crashes, etc. than the person using the software in the real
world.

This is why benchmark testing shouldn't be used as any part of a purchasing
decision and reviews should be read from many sources, the results averaged
and used a small fraction of the information used to decide what program to
buy or not to buy. One of the best sources of buying information comes from
people that have bought and used the product in the real world. That is why
places like Circuit City, Best Buy, Amazon.com have user reviews/ratings.
However, these also have to be watched because too many people either don't
know what they are doing, don't know what their talking about, have a grudge
against a company, etc. So these user reviews need to be used a yet another
small fraction for a purchasing decision. Buy the time you get information
from all available sources you should have a fairly complete picture and
should have a much better chance of getting what you want and what you
expect.

Benchmarks however are 100% worthless.

Robert

Oh, and lets not forget the role of the marketing department which has a
tendency to make any thing weak seem like a much desired feature that
everyone should be clamoring to get. By the time you run the best possible
scenario the benchmark numbers are so far off they might as well have come
from another galaxy. Marketing makes most of the mess when trying to get to
the true about a product. The would hype ice to an ice cube and get it to
buy.

Robert

The Pentax/Samsung shake reduction works in a different way to other
in-camera systems.  The up-and-down and side-to-side movement is
compensating by moving the sensor up-and-down and side-to-side as in
other systems.  However there is additional compensation for rotating
the camera/lens combination about the lens axis, which the other
in-camera systems don't have.

If you jab rather than slowly squeeze the shutter button, the right
side of the camera goes down slightly, rotating the camera/lens
combination about the lens axis.  The Pentax/Samsung system
compensates by rotating the sensor about the lens axis in the opposite
direction.  It's very clever, and it seems to work very well.

The K10D sensor is an excellent sensor, capable of resolving fine
detail.  I don't think the K200D buyers should be too disappointed
with a K10D sensor at a much lower price.

When the collaboration was announced, both companies made great play
of the strengths that the other brought to the relationship.  At that
time, Pentax made it clear that Samsung's expertise was in sensor
design and manufacture.  Because all Pentax and Samsung DSLRs to date
have used Sony sensors, we have had to wait until now to see that
Samsung genuinely does have something to offer.

Until now, Sony has had a stranglehold on the sensor market, supplying
the majority of P&S digicam sensors plus DSLR sensors for Nikon,
Pentax, Samsung and Konica Minolta, whose DSLRs Sony took over.  The
profitability of Sony's digital imaging division has always been
underpinned by its huge earnings from selling sensors.

Now that Pentax has an alternative supplier for one DSLR model, you
can be sure there will be more Samsung sensors to follow soon.  Nikon
is designing its own sensors (although Sony currently manufactures
them) so you can see Sony being squeezed a little.  P&S sales have
reached a plateau and will drop.  And with very disappointing sales of
the Sony Alpha DSLRs, there is not much for Sony to fall back on.

I agree.  I think the K20D will be a very popular DSLR, helping to
increase Pentax's and Samsung's market share, and cementing Pentax's
place as No.3 in the DSLR sales statistics, with Sony, Olympus and
Panasonic trailing behind.