Image Stabilization vs. Shake Reduction

[Seafood]

I think I have decided on my camera....I think.

The Pentax K110d.

The savings on the 110 comes in the lack of Shake Reduction. But, does the K110 still have image stablization?

A salesman told me that it still has IS but the Shake Reduction is a step higher and is not included.

Please clarify what the K110 has and what it doesnt.

[AdamW]

The salesman was full of it.

Shake Reduction is what Pentax calls their technology that helps to counteract camera shake. It operates in the camera body, so it is available with any lens you use. Image Stabilization is what Canon calls their version of technolgy that helps to reduce camera shake. It operates in the lens, so it is only available on lenses that have it. Nikon's Vibration Reduction is also lens-based. Sony's technology is in-camera, like Pentax.

These are all variations on the same theme: counteracting blurry pictures due to camera movement (not subject movement.) In-camera systems are less expensive because they work with every lens. Lens-based systems are supposedly more effective. As far as I know, you can not combine the systems.

What does this mean to you? The K110 has no in-camera Shake Reduction. It can not use lenses with IS or AV. If you want a Pentax with a system to counteract camera shake, get the K100.

[Ray Schnoor]

Pentax calls its image stabilization "shake reduction", so they are really the same thing, and no, the K110D does not have image stabilization or "shake reduction".

From Jeff's review of the K100d, "inside the K100D is a CCD-shift image stabilization system. This system, which Pentax calls Shake Reduction...

To save $100 you can also check out the K110D (http://www.dcresource.com/reviews/cameraDetail.php?cam=873) model, which is the same as the K100D, except that it lacks Shake Reduction."

Ray.

[Vich]

What does this mean to you? The K110 has no in-camera Shake Reduction. It can not use lenses with IS or AV. If you want a Pentax with a system to counteract camera shake, get the K100.
Just to further clarify, both systems (in-camera and in-lens) use motion sensors combined with high-speed motors to "hold the image in place".

So obviously, if they are used together you'll just re-introduce the mess.

Perhaps some day a manufacturer will come out with a version with both, where the in-camera motors work together with the in-lens motors to perhaps triple the effects and achieve a 6 stop IS system. But to date, that's just pie-in-the-sky stuff. I should note; I don't know anything special, this is just layperson speculation.

I would see no reason that a 3rd party lens manufacturer couldn't make an in-camera IS system to function on the K110 (or any other non-IS camera), or for that matter even a K100 with the IS turned off.

In-Lens ones are more effective because it corrects the movement at a center point where only a slight tilt of a lens element has a large counteracting effect. The in-camera one has that sensor jumping around like a Jitterbug in Spring (is that how the saying goes?), with obvious noise, vibration, and potential mechanical movement ramifications. Far better to have slight movement of a refracting lens element.

[Ray Schnoor]

I would see no reason that a 3rd party lens manufacturer couldn't make an in-camera IS system to function on the K110 (or any other non-IS camera), or for that matter even a K100 with the IS turned off.
I'm sure that it is just a typo, but I assume that you mean an in-lens IS system.

Ray.

[swgod98]

I do believe Sigma (and maybe a few others) produce lenses with stabilization built in. An example is the APO 80-400mm F4.5-5.6 EX OS.

http://www.sigmaphoto.com/lenses/lenses_all_details.asp?id=3272&navigator=3

Personally, I think the K100D is worth the extra $100. SR with EVERY lens you put on it is worth it. And if you want in lens stabilization, well, you can always turn the camera SR off.

[Vich]

I'm sure that it is just a typo, but I assume that you mean an in-lens IS system.

Ray.

Nope; I meant to say "manufacturer", not "lens manufacturer". Since a single computer brain would have to manage both motors ....

Probably a stupid idea. A lens manufacturer would probably say "hey, we'll just add a 2nd moving element and accomplish the same thing with less problems".

[Vich]

I do believe Sigma (and maybe a few others) produce lenses with stabilization built in. An example is the APO 80-400mm F4.5-5.6 EX OS.

http://www.sigmaphoto.com/lenses/lenses_all_details.asp?id=3272&navigator=3

Personally, I think the K100D is worth the extra $100. SR with EVERY lens you put on it is worth it. And if you want in lens stabilization, well, you can always turn the camera SR off.
I'm really curious. Are 3rd party lenses with stabilization available for Pentax etc? I'm not up on the non-Canon lineups.

[Rhys]

I'm really curious. Are 3rd party lenses with stabilization available for Pentax etc? I'm not up on the non-Canon lineups.

Sigma has OS - optical stabilisation.

[SpecialK]

In-Lens ones are more effective because it corrects the movement at a center point where only a slight tilt of a lens element has a large counteracting effect. The in-camera one has that sensor jumping around like a Jitterbug in Spring (is that how the saying goes?), with obvious noise, vibration, and potential mechanical movement ramifications. Far better to have slight movement of a refracting lens element.

Sorry, wrong. The camera adjusts the SR function to the focal length of the attached lens. With older lenses you must enter the FL, however.

In-lens stabilization does make the image in the viewfinder more stable, however.

[cwphoto]

Sorry, wrong. The camera adjusts the SR function to the focal length of the attached lens.

I'm sure it does, but it's still less effective than in-lens designs - particularly as FL is increased (which is where it is most useful).

[coldrain]

In case it was not clear already:
The K100D has image stabilization in the body,
the K110D does NOT.

So, since you can not add it in future, GET THE K100D.

And tell the salesman he should be ashamed of himself.

[SpecialK]

I'm sure it does, but it's still less effective than in-lens designs - particularly as FL is increased (which is where it is most useful).

The previous assertion was in-body IS "jumped around like a jitterbug in spring", inferring it was not really controlled in any way. In fact, it only reacts as much as needed, based on the FL, and camera movement.

Whether IS is more important for long lenses or slow shutter speeds is really an opinion, since it is used for both.

I'd appreciate a link to the tests, though.

[coldrain]

The previous assertion was in-body IS "jumped around like a jitterbug in spring", inferring it was not really controlled in any way. In fact, it only reacts as much as needed, based on the FL, and camera movement.

Whether IS is more important for long lenses or slow shutter speeds is really an opinion, since it is used for both.

I'd appreciate a link to the tests, though.
IS os more important on longer focal lengths because of a very simple reason. The camera shake effect on a photo with longer focal lengths is a LOT bigger than with wide angle. This is the reason you need longer shutter speeds with longer focal lengths in the first place.

So, not only do you need IS a LOT sooner with longer shutter speeds, also the amount of correction needed with long focal lengths is a lot bigger. And you can only move the sensor a specific amount. With tele focal lengths the maximus vailable sensor correction is reached sooner.

With in-lens correction you can tweak the IS for exactly that lens, and with a smart placement of the IS lens, you can make it a lot more effective than in-body IS.

That said, not all IS lenses have great IS performance... the Canon EF 70-200 f2.8 L IS USM is not record breaking in IS performance for instance.

So... IS in-body is clearly an economical advantage. IS in-lens has definite technical advantages.

[cwphoto]

Whether IS is more important for long lenses or slow shutter speeds is really an opinion, since it is used for both.

I'd appreciate a link to the tests, though.

I don't get your first statement. The problem is camera-shake, the effect of which is related to both FL and SS.

I don't have a link, but it's a no-brainer that in-lens IS works better:
1) You can see the effect in the OVF.
2) It requires a relatively small movement in the IS optical elements over a sensor-based design.
3) The sensor stays within the image circle.

[Vich]

Sorry, wrong. The camera adjusts the SR function to the focal length of the attached lens. With older lenses you must enter the FL, however.

In-lens stabilization does make the image in the viewfinder more stable, however.
CW and Coldy have responded with nice explanations, excuse me if I'm a tad redundant.

The lens projects an image circle. The camera shakes moving that image around.

If I look through the viewfinder on my 1.6x crop camera with my 300mm lens attached (EFL 480mm), I can visibly see the image shaking around. I half depress the shutter button to activate IS and I hear the motor hum "a hand from above" comes down and visibly steadies it.

Rule of thumb says I would need about 1/500th shutter speed to take a decent shot at that FL. However I can get away with about 1/125th when the IS kicks in.

When I look through a 105mm lens I never visibly see camera shake.

Give it a try if you have a longer lens, it will be obvious why IS becomes more relevant at longer focal lengths.


As for In-Camera lens shaking around ... I really did not understand your response to that. OF COURSE it's not random shaking. The point was that the movement necessary to correct the exact same shake (on the exact same lens) is far greater when you're correcting the projected image rather than the projection point within the lens.

So; instead of having a less shaky image, the in-camera version has to move the projection screen to stay in front of the shaking image.

Also; one would imagine all the sensitive electronics hooked into the sensor and what the problems might be with shaking it all over the place constantly.

[michaelb]

I'm sure it does, but it's still less effective than in-lens designs - particularly as FL is increased (which is where it is most useful).

I would bet that in lens "IS/VR" is probably more effective than in body "SR", particularly at longer FL's, but I have still yet to see any comparative testing which "proves" this to be the case.

[SpecialK]

I don't get your first statement. The problem is camera-shake, the effect of which is related to both FL and SS.

I don't have a link, but it's a no-brainer that in-lens IS works better:
1) You can see the effect in the OVF.
2) It requires a relatively small movement in the IS optical elements over a sensor-based design.
3) The sensor stays within the image circle.

First statement: A photographer who is shooting low light with a wide angle (say in the 1/8-1/15th sec range) would be using IS primarily for the slow shutter speed aspect of it. That is certainly more important to him than the performance with a telephot. So, that seemed to be a generalization.

Just because you see less movement in the viewfinder doesn't mean it works better. You can close your eye and it affects nothing. It is an (operational) advantage, however, as I commented above.

How much movement the system uses - in itself - is meaningless.

I'm not arguing that either system works better that the other. I am only saying that some generalizations have been put forward, and it does not seem to always be so. And, no one aooarently has any real comparisons that I've seen anywhere.

[cwphoto]

First statement: A photographer who is shooting low light with a wide angle (say in the 1/8-1/15th sec range) would be using IS primarily for the slow shutter speed aspect of it. That is certainly more important to him than the performance with a telephot. So, that seemed to be a generalization.

Just because you see less movement in the viewfinder doesn't mean it works better. You can close your eye and it affects nothing. It is an (operational) advantage, however, as I commented above.

How much movement the system uses - in itself - is meaningless.

I'm not arguing that either system works better that the other. I am only saying that some generalizations have been put forward, and it does not seem to always be so. And, no one aooarently has any real comparisons that I've seen anywhere.

I've used both - in lens IS is a lot more effective.

There, feel free to quote me.

[T06]

First statement: And, no one aooarently has any real comparisons that I've seen anywhere.

http://http://www.dcresource.com/forums/showthread.php?t=29652