[falconeye]

Dear friends,

from a couple of sources, I have now compiled a study about the perfomance of the Pentax SR system and published the results

Please, go here: Falk Lumo: Pentax shake reduction revisited


It is based on data by P. Smith, ColorFoto and own tests and evaluations.


(Click for details)
The above chart shows good combinations of shutter speed and focal length (green) and bad ones (red), taking the SR system into account (mathematically modelled). One of the lines is nothing but the 35mm 1/f rule of thumb.

UPDATE: Chart updated in post #22!

My conclusion in a nutshell:

Pentax delivers a capable shake reduction system able to provide up to 4 stops stabilization. However, it is designed to work best at exposure time around 1/20s and therefore, is most useful for normal and wide angle lenses used at low light or in video. Starting at around 100mm focal length, it is increasingly unlikely to see a positive effect from the SR system and beyond 200mm, the SR system cannot be used anymore to produce tac-shap images at lower than usual exposure times.

Olympus shows that this isn't a principle limitation of sensor-shift by delivering best stabilization for longer focal lengths (as far as I am aware of tests). So, there is hope that a future installment of the Pentax SR system is more useful for long focal lengths.

I call it "Tele-SR" and say to Pentax: I want it and I want it now


Anybody wanting to know more, please first consult the blog article. Thank You.

---

Last edited by falconeye; 04-15-2010 at 05:50 PM.

[Parallax]

Thanks, falconeye. That is really interesting. I shoot a lot more wide-normal low light than I do telephoto so for me, it is not bad news but I'm sure others would wish it were the other way around.

[falconeye]

The first time I tried a Pentax with shake reduction (IIRC, a K100D) was with my A*600/5.6, and I was deliberately testing to see if the shake reduction worked with long lenses.
What I found was that it definitely did work, and worked very well.

I use math to help, not to bore you. So, never mind.

But it is interesting what you write. The point isn't in the modelling. It is simply that tests available to me and my own photographing experiments all tell one story: the SR mechanism is made for slower shutter speeds.

Of course, it works well with a A*600/5.6 if you shoot at less than 1/100s.

If however, you have real evidence that it works well with 1/500s (i.e., that SR makes a difference at all at that speed) then I would be highly interested in learning more about that.

If however, you found that you could shoot at 1/60s with a 600mm lens, then this wouldn't be in contrast with what I wrote. It's just not delivering tac-sharp images which, btw is easier with a K100D than a K-7.

[RioRico]

Two question sets:

1) Were longer lenses actually tested? Where can we find the protocols and data?

2) Do we need four active threads on SR in this sub-forum? Can any be combined?

[zackspeed]

Hey Falk,nice test.
In practice i love my SR in my K-7.
Perfekt use is possible with my darling FA*85.
I use this combination for my night citypics without Tripod.
Till 1/15 sec it´s no Problem to get pictures,good enough for 20x30 cm size and commercial homepages.
With the DA*300 is 1/60 no problem and 1/125 is absolut sharp on Display and print.
Sure,100% sharp is no picture without Tripod.
But what will i say?
I love to have my SR in my K-7,i use more than 80% telelenses and ranges between 85 and 300mm.

Best regards,Andy

[imtheguy]

Hello Falk,
Thank you for taking the time to assemble the research and post such an interesting topic. This is WAY more fun to pursue than the political/religious/social threads. As a long range shooter like Wheatfield I don't want to believe the extrapolation is true but I sure can't dismiss it either. You have given me something to work on after I get my taxes filed this week. I shoot 300mm-800mm every day and will experiment with a couple hundred shots to see if the SR effectiveness is obvious one way or another. If I can produce repeatable results I will post them here next week.

Until then, thank you for your efforts. Now hunker down while everybody argues and people look for "in the field" evidence to support or deny the graphic you posted.

[falconeye]

Were longer lenses actually tested? Where can we find the protocols and data?

Yes. Besides my own quick test at 300mm, I explicitely referred to the tests carried out by German magazine ColorFoto carried out in 2008 and 2010.

Their testing methodology was well documented. It is on par with the work by P. Smith (i.e., sub pixel accurate edge blur widths based on N=10 samples, static and blur w/o SR as reference), except that they used 23mm and 130mm focal lengths (35mm and 200mm equivalent for 35mm format). Additionally, they seem to be the only one who developped and built a shake test device which simulates the human tremor pretty well (no pure oscillations). This makes their tests very reproducible.

My blog contains the 2008 result. Here, I now additionally provide a scan of the 2010 result for 130mm:


The blue bar is the tripod reference, the red bar is without SR, thze yellow bar is with SR. "Kantenbreite" means "edge blur width" and "Belichtungszeit" means "exposure time".

It is obvious that the effect of SR at 1/50s and 1/100s is small and at 1/200s is not significantly different from zero. I.e., SR cannot render the 1/200s result as sharp as with a tripod. Of course, this would be desirable and some other tested devices could do exactly this (or at 1/100s or even at 1/50s).

As a long range shooter like Wheatfield I don't want to believe the extrapolation is true but I sure can't dismiss it either. [...]
If I can produce repeatable results I will post them here next week.

I found it hard to swallow as well. I always thought "my" SR would do better for long range tele photography. But I couldn't but post my findings as ... as I found them

I am really keen to learn about other tests which tell a different story.

Untill then, I find it reassuring that slower than 1/100s is a useful domain for tele photography as well and that some other cameras/lenses have a similiar characteristics.

[zackspeed]

Falk,that looks verry effectively,ore not??

[falconeye]

Falk,that looks verry effectively,ore not??

I can only guess what you mean here

Do you mean the chart I just posted. That SR at 130mm is pretty effective at 1/25s?

Yes, and I never said otherwise. But the images are not tac-sharp there as this would require <1px blur width and the SR cannot deliver it for 1/200s or slower.

This isn't called effective for long focal lengths in most people's book.

[zackspeed]

i mean the color foto chart

[zackspeed]

but,maybe i´m a wrong testperson.
cause i train sports shooting....i have no shake production in my hand

[falconeye]

i mean the color foto chart

So, I guessed correctly.

Andy, please read my interpretation in the post above yours. It isn't called effective for long focal lengths in most people's book.

P.S.
You don't need to make two consecutive postings as you can simply append to the first one.

[raider]

Thanks falcon for the article. I am sure your effort is greatly appreciated by the pentax community.

I have 1 question. I am not sure if you have made the same studies into the other type of SRs offered by the other manufacturers but I am keen to know if the canon/nikon's IS/VR system is "roughly" having the same issue as to maintaining the effectiveness at long focal length.

Appreciate your insights.

[twitch]

To me that chart shows SR is extremely effective at 200m which is considered a long focal length (I assume). As effective as a tripod, No. Effective compared to hand holding without SR, hugely so, especially at slow shutter speeds which is exactly when you need the SR.

[falconeye]

I have 1 question. I am not sure if you have made the same studies into the other type of SRs offered by the other manufacturers but I am keen to know if the canon/nikon's IS/VR system is "roughly" having the same issue as to maintaining the effectiveness at long focal length.

I say something about this in the blog article. Short answer is yes and no.

To me that chart shows SR is extremely effective at 200m which is considered a long focal length (I assume). As effective as a tripod, No. Effective compared to hand holding without SR, hugely so, especially at slow shutter speeds which is exactly when you need the SR.

I never said that SR won't work at 200mm. I said it works less well as anticipated at 1/200s.

Look, if you shoot at 200mm, you want blur to be constrained to about 1 or 1.5 pixels width. Otherwise, you could shoot at 150mm or 100mm in the first place!

So, the ability to nail sharpness is essential. With Pentax, whether SR be on or off, you'll have to shot faster than 1/300s with a 200mm lens. Many people would have liked to be able to deploy SR and use 1/100s or even 1/50s to nail sharpness at 200mm. That's my entire point.

And yes, some cameras / lenses out there nail sharpness like a tripod at 1/100s. At least according to the ColorFoto test results. So, it is doable!


BTW,
it can be useful to use 200mm with 2px blur. Because the larger lens will collect more light. Again, my comment applies to the "SR replacing a tripod" use case.