How would one best describe/measure KM's antishake ability?

[ron c sunshine coast,qld,a]

I'm trying to get confimation/correction on an idea.

<BR>Read <A HREF="http://forums.dpreview.com/forums/read.asp?

forum=1024&message=10451070">this post</A> to see how that guy

describes how little sensor movement is needed to acheive

stabilization with the antishake system.

<P>I think he has got it completely wrong but i'm struggling just a

little to be absolutally sure.

<P><P>Would it be correct to describe the antishake system in terms

of it's 'shift' ability -just like a shift lens' ability is measured

in mm of image shift?

<p>For example,KM have said that their sensor can move up to 10mm (i

don't know if it often moves that far or if that's just promotional

talk,but let's just say 10mm maximum for the sake of the argument)

<BR>Would it then make complete sense to say that the movable sensor

could correct for 10mm of image movement (during the time the

shutter is open)?

 

<P>This would tie in very well with the rumours/testing that say

that antishake works best with wider angle lenses.One could very

easily imagine moving a handheld telelens by that much ,but shaking

a wide angle by that much would mean you'd have to turn a huge amount

[frank uhlig]

And who, pray, is KM?

 

Of course telelenses need much more momentum to be stabilized; they are heavier, stick further away from the body etc than the smaller shorter wide ones. A bit of the 1/focal length rule applies here as well.

[erin.e]

Konica Minolta

[greg_erker]

Somewhere on the web is a movie of a partially disassembled 7D

that shows the CCD moving. When powered down the CCD is parked

in the upper left corner. At power up it does some quick movements

then centers itself. From the movie I would guess it has 10mm

total movement. So during used the CCD is centered and you would

have +/- 5mm of movement.

 

If I find the link I'll post it.

 

Greg

[michael_hohner]

If the lens/imaging system has an angle of view of x, and the amplitude of the shaking is within an angle of y, you need to move the sensor by y/x*(sensor size) to compensate for the image shift. For example, if the (horizontal) angle of view is 50 degrees and the (horizontal) shake is within 1 degree, you have to move the (APS-C size) sensor (horizontally) by about 1/2 mm to compensate for the shift of the image.

<p>

Another parameter is the reaction time of the system, ie. time from beginning of movement to completion of compensating sensor move. The image displacement during that time limits the effectiveness of the anti-shake mechanism.

<p>

The size and weight of the lens is no factor for AS, but the size and weight of the sensor is.

<p>

Things get more complicated when you want to compensate also for lateral movement of the camera in addition to angular movement. In that case you do not only have to factor in the angle of view but also the focus distance. Lateral movement has a stronger effect on image displacement for closer focus distances. For example, at 1:1 magnification, 1 mm of lateral movement results in 1 mm of image shift, while near infinity, a lateral movement of 1 mm has nearly no effect.

[ron c sunshine coast,qld,a]

Interesting comments Michael.It looks like i'm on the right track.

<BR>Your description on reaction time is also good-something i'm just starting to get my head around now.I guess it would be especially difficult for the sensor to respond as it nears it's full 5mm-each-way movement .

 

<P>Yes Frank ,KM is a dumb name.I'm getting used to it unfortunately.

<BR>I have my brother's old manual pentax model KM here at the moment.For awhile i had sitting on the dresser party covered by something so that the pentax logo wasn't showing.Every time i saw it as i walked past i thought in the back of my mind 'that's odd-when did konica-minolta ever make a manual camera?' LOL