[er1kksen]

A combination of the K-x's SR and good handholding technique gave me sharp shots at 135mm and 1/30th-1/40th second shutter speeds shooting a show tonight (this was at f2.8 and ISO 2500, mind you, and looking a tad underexposed. Rather dark venue)... SR gives you leeway beyond what can be achieved through good technique alone.

I just have to shut it off on my K-x when using shutter speeds above 1/80th. Not that that's a real problem, and it's just my unit and a few others.

[Wheatfield]

well... no. Bad technique is bad technique, and SR isn't there to fix that. I just think it's unfair to claim that SR is only good for compensating for bad technique. SR does give a 2-4 stop advantage, and that's 2-4 stops that can be used by any photographer before a tripod becomes necessary.

I don't think anyone has ever claimed that all it is good for is compensating for bad technique, at least not until you brought it up.
Certainly not in this thread.

However, there can be no doubt that (sufficiently) bad technique will reduce or remove it's effectiveness.
The only way it can give a 4 stop improvement, for example, would be to use very good technique, and I suspect that if a person was exceptionally steady, one could get more than 4 stops out of the Pentax SR system.

[Adam]

SR makes shutter speeds as low as 1/6 or even 1/5 feasible, so to answer the question directly- YES!

[Jasvox]

SR is Shake Reduction, not Shake Proof.

Jason

[GoremanX]

GoremanX, could you try the same with SR off? Or at least find the slowest speed you can achieve without SR?
Just to know how many stops we are really talking about here...
With my 200mm, I'm able to go from 1/30 (SR off) to 1/5 (SR On), with about the same success rate (about 75% sharps). That's nearly 3 stops...

I just took 90 pictures at different shutter speeds with SR on and off to see how much improvement I really get out of it.

For all of these, I was standing about 30 feet from my subject, which was fairly well lit with 2 x 100watt equivalent CFLs and 1 x 60watt equivalent CFL. All shooting was done at ISO 800, and for every stop of shutter speed I added, I removed one stop of aperture. My stance involves having both elbows resting against my chest, aiming carefully at a spot, half-pressing the shutter to auto focus, waiting for the SR icon to come on (if SR is turned on), and then gently pressing the shutter button the rest of the way to take a picture. My left hand is holding the lens by its zoom ring, and my right hand is holding the camera's body, with my thumb under the camera to help keep it stable when I press the shutter button with my index finger. I aimed at a slightly different spot of my scene for each shot, and always re-focused.

The camera is a K-7 with firmware 1.03. The lens is a DA* 50-135mm set to 135mm. I took 9 pictures at each setting to see how many of them would turn out "good". Why 9? Because that fills up a 9x9 grid on the screen in playback, which made it easier to keep count. I only counted the ones with excellent sharpness as "good". Any amount of visible camera shake meant I didn't count that picture. The scene I picked included a lot of text, which made it easy to determine how much camera shake was present.

Is that description detailed enough?

Here are my results:

SR Disabled
1/100 and F2.8 - 7 good (always out of 9)
1/50 and F4.0 - 3 good
1/25 and F5.6 - 1 good
1/13 and F8.0 - 0 good
1/6 and F11 - 0 good

SR Enabled
1/100 and F2.8 - 8 good
1/50 and F4.0 - 8 good
1/25 and F5.6 - 8 good
1/13 and F8.0 - 4 good
1/6 and F11 - 3 good

These 90 photos span 5 different shutter speeds and 4 full stops. I think those results are pretty good, they certainly show SR in a positive way. And they're much better than the results I used to get with the old firmware. I couldn't get so many good pictures at 1/6 before, especially not at 135mm.

Here are a few 100% crops of the shots that I considered "good" (remember I was always focusing in a slightly different place):


(that far right one is borderline, that was the worst of the "good" ones)

And these are 100% crops of some of the shots I considered "bad":


This is the scene I was shooting, aloways focusing somewhere on the box to make sure the text was in focus:

[GoremanX]

I don't think anyone has ever claimed that all it is good for is compensating for bad technique, at least not until you brought it up.
Certainly not in this thread.

I dunno, I guess that's just the impression I got from some of the posts on this thread. But you're right, though, it wasn't ever specifically stated that way. My bad.

However, there can be no doubt that (sufficiently) bad technique will reduce or remove it's effectiveness.
The only way it can give a 4 stop improvement, for example, would be to use very good technique, and I suspect that if a person was exceptionally steady, one could get more than 4 stops out of the Pentax SR system.

Yes, bad technique will defeat any SR/VR/IS/OS technology. I don't see SR as a crutch to rely on when I'm drunk. I see it as a tool to get a higher percentage of usable pictures during difficult shooting sessions where I can't use a tripod. And in my experience, Pentax's SR performance is right up there with the alternatives. I tried out a Nikon 70-300 VR lens on a D90 a week ago at a Ritz store, and I was getting somewhat better results with my K-7/50-135mm combo than with the D90/70-300mm. I had both cameras side by side and was swapping back and forth a lot while the sales guy got annoyed because he knew I wasn't there to buy anything. The difference wasn't huge, but at similar focal lengths, my K-7 was giving me a few more keepers than the Nikon was.

Incidentally, I also compared auto focus performance in daylight and in poor light, and my K-7/50-135mm didn't fare nearly so well there But that's not something for this thread.

[labnut]

My apologies. I was going to post a link to the study but I am glad someone else has already done that. Thanks for the many comments and insights.

Just to add to the information, below are links to some of the key Pentax patents in this field. They are very helpful for understanding how shake reduction is implemented.

Pentax Shake Reduction Patent Original US7224893
Pentax Shake Reduction Patent 7224893
Pentax Shake Reduction Patent 20080226276
Pentax Shake Reduction Patent 20080226276 Original

[distudio]

Just to add to the information, below are links to some of the key Pentax patents in this field. They are very helpful for understanding how shake reduction is implemented.

Very interesting, thanks for that, the description of SR operation in terms of the third handshake angle (roll/z) in the patent 2008/0226276 does indicate that SR in the z plane may be implemented. However I find it curious that this roll compensation functionality isn't featured in the text or diagrams for the K7 on the Japanese Pentax site

K-7?????????????PENTAX

Cheers,

[labnut]

Very interesting, thanks for that, the description of SR operation in terms of the third handshake angle (roll/z) in the patent 2008/0226276 does indicate that SR in the z plane may be implemented. However I find it curious that this roll compensation functionality isn't featured in the text or diagrams for the K7 on the Japanese Pentax site

Rob, as near as I can make out the roll they refer to is the rotation detection and horizon levelling function that Pentax pioneered with the K7. They specifically refer to the roll induced by depressing the shutter release button. I think the roll is rotation around the lens axis. So I think what they describe is already implemented in the K7. But patents make for hard reading and I will have to read it again more carefully.
Regards, Peter.

[LRNav]

Flyer: A nice piece of work by P. Smith. Thanks for posting the link.

Dick K.

[Ira]

Goreman, don't you think it's time to put away the Christmas stuff?

[GoremanX]

Goreman, don't you think it's time to put away the Christmas stuff?

It is put away... it's in the basement, where I throw all the Christmas crap out of sight. No-one goes down there except me ever since it got flooded and he floor got all warped (hence rendering the massive, expensive pool table unusable). That fugly Santa just got picked up from the pile of other Christmas crap on the floor and thrown into a low-light "scene" that I use whenever I need some colourful thing to shoot at 2am.

[Rondec]

I can't compare in lens VR/IS with sensor SR, but I can say that I get at least 3 stops of improvement in my hand holding ability. Every so often it gets shut off for some reason and then I start to notice. I particularly notice the benefit with my longer lens (longest is the DA 50-135), just because it is so easy to have my shutter speed drop into the 1/20 or 1/30 range.

There was a thread where people posted long exposures demonstrating SR effectiveness.

[LRaven]

Here it is comparison optics IS with Pentax built in SR: RiceHigh's Pentax Blog: When OIS Lenses are Put onto Sensor-shift IS Bodies..
From my side I should say that SR in modern Pentax k-x less effectiveness than 10 years old camera Canon S1, with optic's image stabilizator IS.

[GoremanX]

Here it is comparison optics IS with Pentax built in SR: RiceHigh's Pentax Blog: When OIS Lenses are Put onto Sensor-shift IS Bodies..

While everyone seems to agree that there's a minimum 2 EV improvement from Pentax's in-body shake reduction, this guy says the improvement is "marginal at best". Also, last I checked, the OS lenses automatically disable the in-body SR when OS is activated. This has been confirmed by numerous other OS lens owners. So this guy's results are dubious at best.