You almost seem to be taking the position, John, that since before we had digital cameras and the internet (and hence the ability to analyse and communicate problems such as these), we wouldn't have known about the problems, therefore we shouldn't be making such a fuss now.

The thing is, the reason we're prepared to spend about £1000 on a hi-tech camera, is because we want all the advanced features, and we want them to work reliably.

Of course you can fiddle and tinker, and up to a point, work around the limitations of the technology.

But it shouldn't have to be like that.

And if testing the ability of the AF system to focus in a variety of light levels isn't beyond the ability of careful amateurs, how much less is there an excuse for the manufacturer not to, under the controlled conditions they supposedly use when they develop their equipment?

Heavens...

It's not a solution. It is a workaround.

And it's a rubbish workaround if the subject is moving and not giving you time to mess about with QS.

And if your eyesight is so good that you can QS accurately, reliably and quickly enough - down at EV 1.5 - then you're a better man than most of us that believe that a thousand pound camera should work.

Honestly, John, I know you're a Pentax fan, but to be so indifferent to this problem suggests an almost other-worldly attitude to expensive technology.

Guess they've improved the forum search a bit...

This was it, I think.

When I had quite a careful look at this, some time ago, the SR was not capable of eliminating the small amount of finger-induced shake (or possibly mirror-slap) that you inevitably get while pressing the button manually, even with the camera on a tripod.

With or without SR, it was still visible, not appreciably different either way.

Indeed. Regardless of the physics, as usual

Actually, this particular point hasn't been discussed at all, AFAIR. We've always argued about whether non-existent shake fools the SR into making some.

This is quite different - there is actual shake in this case, and the sensor SR is working *exactly* as it would without an IS lens on the front.

It really doesn't demonstrate this at all.

As I explained above, the SR is just moving the sensor in response to the accelerations it's detecting, just as it usually does.

It can't tell that the image isn't moving, because it doesn't work by detecting image movement.

It works by detecting camera body movement, and then on the assumption that this would move the image relative to the sensor, moves the sensor accordingly to compensate.

So it's not at all acting in a spurious manner.

Sorry John, but whether or not there's something funny about Pentax SR on a tripod, this video doesn't show it.

Was this a lens without IS?

And did the wander stop when you switched SR off?

If so, then this might be the first piece of actual, reproducible evidence (as opposed to the usual waffle and hearsay), to suggest that SR on Pentax cameras does introduce movement when there's actually none.


However, the K10D doesn't have live view.

Trying to use sensor SR at the same time as lens IS is futile, and a moment's thought shows why:

- Each system is using accelerometers to measure the amount of shake

- The design of each is independent.

So, while the lens IS is ensuring that the image that arrives at the sensor is still, the sensor SR is still experiencing the same accelerations whether the lens IS is on or off. Therefore it assumes that it still needs to move the sensor to compensate.

Clearly, moving the sensor around when the image isn't moving, is going to reintroduce the shake that the lens removed.

The sensor-based SR isn't watching the image move or not move, as we do on live view - it's measuring accelerations which are still happening, so of course it can't tell that the image has been stabilised already.

So sure, if you use SR when the camera is moving around, but the image is still, then you'll get shake in the image.


Nothing at all like the tripod situation, where the camera is still, and there aren't any accelerations to measure.

Phew. I was so tempted to get a second one!

No

It's a ball and socket joint so when you squeeze the grip, it releases, and you can position it anywhere you like. The slot in the side lets you tip it over on it's side at 90 degrees to the vertical too.

Then when you release the grip again, the spring holds it locked wherever you put it.

If it's as good as mine, the spring is incredibly strong, and allows no slippage at all.