Does shake reduction affect image quality?

[Didache]

Hi all .. I have just got my first camera with shake reduction technology. My question is this: Does it affect image quality when it's turned on? Obviously it is going to improve quality in low light/hand-held/etc - but if the shutter speed is high enough not to gain any advantage, will the SR actually hurt the image slightly?

Cheers
Mike

[drg]

It will depend upon the circumstances and how the camera really uses an increased shutter speed. Should the exposure range be such that it can open the aperture and raise the shutter speed, it is a program shift. If the requirements are based on the focal length of the camera it may not have enough range left to make that adjustment and turn up the ISO.

Higher shutter speed is what of course everyone used prior to Image or Optical Stabilization and Vibration Reduction or the 'Active' forms of shake reduction.

I just, and I mean just, emailed P-J a review and that camera has very good high ISO performance, so turning up the shutter speed does not significantly reduce image quality.

A Samsung camera I played with in the past few months on the other hand, turned the ISO to high and the image suffered.

If your camera has an AUTO ISO feature that can be combined with SR, use it. That should limit the upper range of ISO that will be used so you can exert more control over the final image.

SR requires some experimentation to see what it does for you. Some cameras are just a lot easier to keep steady than others.

What camera are you using out of curiosity?

[Didache]

The K10D .. presumably the one you've just reviewed?

Mike

[mwfanelli]

Hi all .. I have just got my first camera with shake reduction technology. My question is this: Does it affect image quality when it's turned on? Obviously it is going to improve quality in low light/hand-held/etc - but if the shutter speed is high enough not to gain any advantage, will the SR actually hurt the image slightly?

Cheers
Mike

Image stabilization or anti-shake just lets you shoot handheld at a lower shutter speed than you normally could. I have used Canon IS lenses since 2000 and have never seen any change in image quality. If you are using a shutter speed at or faster than 1/FL then IS won't contribute anything. It makes no difference if you leave it turned on anyway.

[gryphonslair99]

Image stabilization or anti-shake just lets you shoot handheld at a lower shutter speed than you normally could. I have used Canon IS lenses since 2000 and have never seen any change in image quality. If you are using a shutter speed at or faster than 1/FL then IS won't contribute anything. It makes no difference if you leave it turned on anyway.

Oh yeah, tell that to my batteries. Leaving IS on all the time will wear shorten battery life.

[mwfanelli]

Oh yeah, tell that to my batteries. Leaving IS on all the time will wear shorten battery life.

Sorry, but that just isn't true for lens-based Canon IS. The IS system is ONLY on when the shutter button is depressed half-way down. It is off at all other times.

For Canon, at least, unless your shooting technique keeps your finger glued on that shutter release constantly the difference in battery usage is negligible. I have NEVER noticed a difference in batter life over the last seven years of use.

I have no idea how the sensor-based IS systems fare. I have seen no reviews indicating that there is a large drain on batteries. Could you refer me to some of these reviews? Thanks!

[drg]

No, I did not review the K10D. Darn!! I reviewed a Fuji Point and Shoot that employs a pretty straigtforward Shutter Speed increase technique.

The K10D employs of course the sensor shaker technology which has the benefit of working for every lens that gets mounted.

The Pentax has that bracketed Auto ISO feature as well which combined with Shake Reduction, has the potential to make for closer to optimum exposure while minimizing the effect of extra motion.

Whether moving the sensor affects image quality or not is going to be the subject of lot of further comparison testing I'm sure!

My feeling is that it may not affect the Image Quality as much as it may not be as effective as in the lens stabilization, at least in these early generations of use. There is a lot that has to be coordinated, and Pentax has certainly taken on an ambitious program with the three axis stabilization 'in the camera'.

I will want to see many photos with and without to really say whether it adversely affects in part of the photo other than just blur from camera motion.

[Loupey]

Without knowledge of the inner workings of the K10D, I can only comment from my experience with the Canon system:

Leaving the IS on even when using higher shutter speeds is more beneficial than detrimental. My reasoning is that if I can handhold well at 1/FL, then shouldn't the IS simply make it that much more stable?

There is one slight problem with this though - when a subject bolts away suddenly and you pan with it, the servos will initially try to keep the frame still and you'll notice a slight lag following it. But this is a very specific exception to the rule (imagine tracking a bee as it zips, lands, jerks, ...).

As for the battery drain, I see a significant drain - but that's because I like to keep the shutter button just a hair trigger from firing when I have the subject in my sights. I would imagine that the power consumption would be far less in stabilizing an image sensor compared to stabilizing the entire lens element.

[SmartWombat]

On my Canons I can hear the IS still running for a few seconds after the button is released.
The Panasonic LZ5 has two optical IS modes, full time and exposure only.
The Minolta A1 had the same.

I didn't find that IS on any of those cameras gave me an unacceptably short battery life, whether in-lens or on the sensor.

[Didache]

I'm going to try an experiment I think. I will wait for a pretty dull day, or alternatively dusk. I'll lock the camera up LOOSELY on a tripod and try some shots of a detailed scene with a longish shutter speed, with and without the anti shake on. Maybe I will be able to compare some blown up details to see whether or not the AS seems to affect the way fine detail is captured. Or, if anyone can think of any other way to objectively test this - I'd be glad to hear it. I'll post the results if anything interesting comes up.

Mike

[drg]

I will add a couple more thoughts to this discussion. The idea behind any anti-shake, stabilization, etc. is too improve the image by eliminating the visual recording of vibration and shake. So if all that is going on is stabilizing the recording of the image, then it will always be an improvement.

When you start tweaking other parameters including shutter speed and ISO then you are altering other factors which do have greater potential to alter Image Quality.

The question with the 'sensor shakers' has been about whether they break down at longer focal lengths and as they are trying to 'catch' the image (not the best euphemism perhaps) are they going to be as successful as a lens based gyro system?


On another note brought up, IS usage in Canon's CAN cause shortened battery life.

There is a big however, it usually is as a result of:

1. The lens needing some maintenance/repair. I had a rental situation with a 400/2.8 IS that was really eating batteries before it started to just not work very well at all.

2. I used a 24-105 f4 IS recently (since the beginning of the year) that seemed to make battery life drop precipitously. That lens only was used by me for about 250-300 shots so there may have been something else in play, though I kind of doubt it.

3. Auto Focus I believe, in AI Servo and certainly taking advantage of predictive AF, eats a lot more power than IS.

If you have IS turned on and are doing a lot of tracking but not making exposures, the actual shot count per charge will indeed drop.

An interesting test that I did some years back, was to use the popular 70-200 with IS turned on and turned off and shoot at one end of the zoom range and then the other to see if it made a difference in image count per charge. When using a Canon DSLR with the BP-5xx batteries, there was a decided difference of, if I recall fairly accurately on very similar shooting circumstances, over %20 reduction when shooting on the 200 end of the lens with IS on and tracking and switching from on subject to another. Not a terribly scientific test, but around 250+ images at each of the four combinations (70mm IS on, 70mm IS off, 200mm IS on, 200mm IS off) did produce results that indicated that IS and maybe something about the performance of the lens at its extreme consumed power at a greater rate.

[Sushigaijin]

I suggest that in-body stabilization is a huge boon to those of us who cannot afford $1000+ for long stabilized lenses. Even if it doesn't work as well. A $200 lens and a $1500 lens with a stop of useable difference is not worth the extra $1300. It is possible to build a large collection of quality lenses at a much better price point, and lose one stop. Big deal. Some stabilization is better than none, and people have been taking pictures since LONG before stabilization was even an option...and many of them did just fine.

[Copy_Kot]

If I use my Canon 300mm f/4L IS USM on a tripod with image stabilization turned on, I get a drift in the lens... as if the lens is still trying to compensate for assumed shake. The solution to fix this is very simple, I just turn off the IS. I wonder if cameras with built in stabilization do this as well (try to compensate even when the camera is motionless).