Win a Pentax Limited Lens: Raffle Now Live!

stewart_photo · 08-02-2007, 10:37 AM

The German print magazine, ColorFoto (colorfoto.de), recently did a test of various image stabilization systems (17 in all used by nine manufacturers). The Pentax K10D with kit lens was included in the test. Other DSLR's in the same general price range were the Nikon D80 with 18-200 VR (Vibration Reduction) lens, Canon EOS 30D with 17-55 IS (Image Stabilizer) lens, and Sony Alpha DSLR-A100 with kit lens.

The shake reduction system employed by the Pentax K10D came out on top of the four with a score of 85% (described as excellent for this price range by the magazine), followed by the Nikon D80/18-200 VR with a score of 78%, the Canon 30D/17-55 IS with a score of 77%, and the Sony A100 with a score of 51%. Each score is in comparison to 100% with the unshaken camera.

If you read German, the entire article can be purchased and downloaded here...

colorfoto.de - Anti-Wackelsysteme im Vergleich

All cameras were tested at the same focal length of 35mm (equivalent), using exposure times of one half and one quarter seconds. The tests were done using a laboratory shaker at frequencies of 2 Hz for normal conditions and 4 Hz to provide extreme conditions. With each camera, an image was first taken without movement and then again with movement & shake reduction. To eliminate camera variables, images from each camera were compared only against images taken by the same camera.

To determine the basic test conditions (2 & 4 Hz), the Bradley J. Davis/John O’Conell method of amplitude measurement of human physiological tremor in regard to holding photographic cameras was adopted. To eliminate human varibles in the final image evaluations, the Siemens Spatial Frequency Response (SFR) method was used for blur estimation. This method was developed in cooperation between Dietmar Wueller of Image Engineering and the Cologne University of Applied Sciences. A complete abstract of the testing method (PDF in English) can be see here...

http://digitalkamera.image-engineering.de/downloads/EI_2007_6502_22.PDF

In addition to proving the SR system employed by the K10D is indeed very nice, the results also directly contradict commonplace notions that in-lens stabilization is superior to in-body stabilization. In-body stabilization can be superior to in-lens stabilization in some situations.

stewart

Last edited by stewart_photo; 08-02-2007 at 02:38 PM. Reason: Clarification: ...2 Hz "for normal conditions"

Peter Zack · 08-02-2007, 11:38 AM

Pentax is getting quite a few nice feathers in it's cap. Another positive review. Wonder if RH is reading?

kenyee · 08-02-2007, 01:41 PM

This was covered a bit on DPR. Conclusion was in-body VR is more effective than in-lens VR at smaller focal lengths, so the test only confirms this. If you use longer lens, the in-lens VR should be better, so I hope they'll test that case as well...

pmperry · 08-02-2007, 04:06 PM

Originally posted by kenyee

This was covered a bit on DPR. Conclusion was in-body VR is more effective than in-lens VR at smaller focal lengths, so the test only confirms this. If you use longer lens, the in-lens VR should be better, so I hope they'll test that case as well...

Yeah but interestingly enough the K10D / GX-10 is the only IS enabled body, that I know of, in which you can choose the focal length of the stabilisation system which is huge if you ask me because one thing Canon was very right about, "It takes more IS for a longer lens than it does for a shorter lens".

Oh and if you've ever turned the dust cleaning device on you can hear the SR mechanism jolt and to me it seems to move more than the IS elements in my old 70 - 200 F/2.8 IS L USM ever did. So, with that said I can't imagine the K10D being any worse than the competition out to 400mm.

mattdm · 08-02-2007, 06:07 PM

Originally posted by pmperry

Yeah but interestingly enough the K10D / GX-10 is the only IS enabled body, that I know of, in which you can choose the focal length of the stabilisation system which is huge if you ask me because one thing Canon was very right about, "It takes more IS for a longer lens than it does for a shorter lens".

Are you sure the Sony cameras just don't have the feature for setting the focal length for old manual lenses? I'm not sure how'd they'd properly work at all without knowing the focal length somehow.

silverbullet · 08-02-2007, 08:54 PM

If you own a Pentax, Canon or Nikon, you should be happy with the SR results with standard focal length lenses. I really don't know if these results are different enough to equate to any real world difference. If you own a Sony...sorry.

Entropy · 08-02-2007, 08:59 PM

Originally posted by mattdm

Are you sure the Sony cameras just don't have the feature for setting the focal length for old manual lenses? I'm not sure how'd they'd properly work at all without knowing the focal length somehow.

I would bet on this. For an IS system to work well, it MUST be tuned based on the lens focal length.

As to the in-body element moving more than the IS element in an in-lens system - If the IS element (whether in-lens or in-body) is placed properly, far less movement is needed to achieve the same effect. My guess is that when Canon bashes in-camera IS, they're referring to this fact.

Personally, I'd love to see the ability to use in-camera or in-lens (if available) IS depending on what lens was installed. In-camera for short lenses and anything that doesn't have IS, and in-lens for really long lenses.

carpents · 08-02-2007, 09:12 PM

Originally posted by mattdm

Are you sure the Sony cameras just don't have the feature for setting the focal length for old manual lenses? I'm not sure how'd they'd properly work at all without knowing the focal length somehow.

The Minolta Anti-Shake defaults to roughly 50mm shake reduction with any (every) manual lens. Because they changed their mount in the modern era, it was assumed by Minolta that most users would be using the Minolta A-mount lenses, which have long had electronics (including the focal length) transmitted. Hence no need to account (too much) for manual lenses.

stewart_photo · 08-03-2007, 02:31 AM

Originally posted by kenyee

This was covered a bit on DPR. Conclusion was in-body VR is more effective than in-lens VR at smaller focal lengths, so the test only confirms this. If you use longer lens, the in-lens VR should be better, so I hope they'll test that case as well...

Sadly, DPReview won't let me sign on (doesn't like my yahoo email address), so I very rarely ever visit that site. Anyway, there were reasons why that particular focal length was chosen. One being that prior testing of many lenses revealed very little difference in sharpness between lenses at this focal length (varying sharpness could skew the results). There were others, but I don't remember what they were and don't have the magazine anymore. I wrote the message above about a week ago while a friend was visiting for several days and that friend has now returned home with the magazine. Regardless, since all advanced image stabilization systems, including that in the K10D, compensate to some extent for focal length of the lens, I also doubt there would be that much of a difference in the results if a greater focal length were used. Of course, I can't prove that so we'll just have to agree to disagree with each other's opinions.

stewart

axl · 08-03-2007, 07:55 AM

Originally posted by pmperry

Yeah but interestingly enough the K10D / GX-10 is the only IS enabled body, that I know of, in which you can choose the focal length of the stabilisation system...

let me think...
K100D maybe

Deni · 08-03-2007, 12:32 PM

Originally posted by axl

let me think...
K100D maybe

And probably the k100d super as well

Entropy · 08-03-2007, 07:21 PM

Originally posted by stewart_photo

Sadly, DPReview won't let me sign on (doesn't like my yahoo email address), so I very rarely ever visit that site. Anyway, there were reasons why that particular focal length was chosen. One being that prior testing of many lenses revealed very little difference in sharpness between lenses at this focal length (varying sharpness could skew the results). There were others, but I don't remember what they were and don't have the magazine anymore. I wrote the message above about a week ago while a friend was visiting for several days and that friend has now returned home with the magazine. Regardless, since all advanced image stabilization systems, including that in the K10D, compensate to some extent for focal length of the lens, I also doubt there would be that much of a difference in the results if a greater focal length were used. Of course, I can't prove that so we'll just have to agree to disagree with each other's opinions.

stewart

I think the one possibility might be that the image displacement induced by camera shake on a really long lens might "peg" the mechanics of an in-body SR system, whereas if the stabilization were in the lens it could be placed in a part of the optical path where a small amount of correction element travel would result in a large correction.

axl · 08-12-2007, 09:51 AM

Originally posted by Deni

And probably the k100d super as well

you got me on that one

SteveB · 08-12-2007, 10:45 AM

Does in lens IS twist the lens, i.e. just move one side? If so, and pictures I've seen seem to suggest that's how it works, then some of the image must go out of focus, surely, as it's simulating a misaligned lens that owners send back to be repaired because one side of the image is soft!

There's a comparison test of in-lens vs. in-body IS in What Digital Camera magazine September issue. Not very scientific as vastly different lenses are used on each camera although they are all used at around 100mm FL (35mm equiv), but the K10D rating is 'very good' at 1/15sec, 'useful' at 1/8sec and ineffective at 1/4sec using a 24-90mm lens but it was tested at 115mm FL. The best performance was the Nikkor 55-200VR tested at 83mm FL (hmm... do I detect bias?), working down to 1/4sec which no others did. The Canon was one of the worst, being ineffective at 1/8sec.

Those tested are......
(Optical IS)
Canon EOS-30D with 70-200mmf4 IS
Panasonic DMC-L1 with Leica 14-50mmf2.8-3.5
Nikon D40x with 55-200mmf4-5.6VR
(In body)
Pentax K10D with 24-90mmf3.5-4.5 AL IF
Olympus E-510 with Zuiko 50-200mmf2.8-3.5
Sony Alpha A100 with Sony 75-300mmf4.5-5.6

arbutusq · 08-12-2007, 12:45 PM

I can't say which system is better as i only have access to in body, but I can get sharp results at 1/50 and reasonably consistent ones at 1/30 at 450 EFL, that's 3-4 stops advantage and good enough for me.

Each person has a slightly different shooting style and will get very different results with each system. I prefer in body because I get disoriented when the image is stabilized in the viewfinder.

08-02-2007, 10:37 AM	#1
stewart_photo Veteran Member Join Date: Apr 2007 Location: Heidelberg, Germany Posts: 1,864	Test Results: K10D SR Superior - Even Compared to Some In-Lens IS Systems The German print magazine, ColorFoto (colorfoto.de), recently did a test of various image stabilization systems (17 in all used by nine manufacturers). The Pentax K10D with kit lens was included in the test. Other DSLR's in the same general price range were the Nikon D80 with 18-200 VR (Vibration Reduction) lens, Canon EOS 30D with 17-55 IS (Image Stabilizer) lens, and Sony Alpha DSLR-A100 with kit lens. The shake reduction system employed by the Pentax K10D came out on top of the four with a score of 85% (described as excellent for this price range by the magazine), followed by the Nikon D80/18-200 VR with a score of 78%, the Canon 30D/17-55 IS with a score of 77%, and the Sony A100 with a score of 51%. Each score is in comparison to 100% with the unshaken camera. If you read German, the entire article can be purchased and downloaded here... colorfoto.de - Anti-Wackelsysteme im Vergleich All cameras were tested at the same focal length of 35mm (equivalent), using exposure times of one half and one quarter seconds. The tests were done using a laboratory shaker at frequencies of 2 Hz for normal conditions and 4 Hz to provide extreme conditions. With each camera, an image was first taken without movement and then again with movement & shake reduction. To eliminate camera variables, images from each camera were compared only against images taken by the same camera. To determine the basic test conditions (2 & 4 Hz), the Bradley J. Davis/John O’Conell method of amplitude measurement of human physiological tremor in regard to holding photographic cameras was adopted. To eliminate human varibles in the final image evaluations, the Siemens Spatial Frequency Response (SFR) method was used for blur estimation. This method was developed in cooperation between Dietmar Wueller of Image Engineering and the Cologne University of Applied Sciences. A complete abstract of the testing method (PDF in English) can be see here... http://digitalkamera.image-engineering.de/downloads/EI_2007_6502_22.PDF In addition to proving the SR system employed by the K10D is indeed very nice, the results also directly contradict commonplace notions that in-lens stabilization is superior to in-body stabilization. In-body stabilization can be superior to in-lens stabilization in some situations. stewart Last edited by stewart_photo; 08-02-2007 at 02:38 PM. Reason: Clarification: ...2 Hz "for normal conditions"

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Wireless P-TTL test results	adamfogerty	Pentax Camera and Field Accessories	3	06-20-2010 05:18 PM
Test results: resurrected Zykkor 135mm Manual Lens	ismaelg	Pentax SLR Lens Discussion	4	09-27-2009 07:36 PM
Test results: Pentax SV / SuperTak 55 1.8	ismaelg	Film SLRs and Compact Film Cameras	2	07-31-2009 05:19 PM
slrgear.com test results: Tamron 28-75 f2.8: best lens... ever (even full frame).	123K10D	Pentax SLR Lens Discussion	22	11-15-2007 08:06 AM
Need help evaluating results from focus test	pdmlmember	Pentax DSLR Discussion	19	09-11-2007 02:09 AM

08-02-2007, 11:38 AM	#2
Peter Zack Pentaxian Moderator Emeritus Join Date: May 2007 Location: Edmonton Alberta, Canada Photos: Gallery \| Albums Posts: 10,643	Pentax is getting quite a few nice feathers in it's cap. Another positive review. Wonder if RH is reading?

08-02-2007, 01:41 PM	#3
kenyee Veteran Member Join Date: Mar 2007 Location: Boston, PRofMA Photos: Albums Posts: 3,026	This was covered a bit on DPR. Conclusion was in-body VR is more effective than in-lens VR at smaller focal lengths, so the test only confirms this. If you use longer lens, the in-lens VR should be better, so I hope they'll test that case as well...

08-02-2007, 04:06 PM	#4
pmperry Forum Member Join Date: Jul 2007 Location: Tampa, Florida Posts: 50	Originally posted by kenyee This was covered a bit on DPR. Conclusion was in-body VR is more effective than in-lens VR at smaller focal lengths, so the test only confirms this. If you use longer lens, the in-lens VR should be better, so I hope they'll test that case as well... Yeah but interestingly enough the K10D / GX-10 is the only IS enabled body, that I know of, in which you can choose the focal length of the stabilisation system which is huge if you ask me because one thing Canon was very right about, "It takes more IS for a longer lens than it does for a shorter lens". Oh and if you've ever turned the dust cleaning device on you can hear the SR mechanism jolt and to me it seems to move more than the IS elements in my old 70 - 200 F/2.8 IS L USM ever did. So, with that said I can't imagine the K10D being any worse than the competition out to 400mm.

08-02-2007, 06:07 PM	#5
mattdm Veteran Member Join Date: Feb 2007 Location: Boston, MA Photos: Gallery Posts: 2,948	Originally posted by pmperry Yeah but interestingly enough the K10D / GX-10 is the only IS enabled body, that I know of, in which you can choose the focal length of the stabilisation system which is huge if you ask me because one thing Canon was very right about, "It takes more IS for a longer lens than it does for a shorter lens". Are you sure the Sony cameras just don't have the feature for setting the focal length for old manual lenses? I'm not sure how'd they'd properly work at all without knowing the focal length somehow.

08-02-2007, 08:54 PM	#6
silverbullet Senior Member Join Date: Jan 2007 Location: Edmonton Alberta Posts: 212	If you own a Pentax, Canon or Nikon, you should be happy with the SR results with standard focal length lenses. I really don't know if these results are different enough to equate to any real world difference. If you own a Sony...sorry.

08-02-2007, 08:59 PM	#7
Entropy Veteran Member Join Date: Jun 2007 Location: Owego, NY Posts: 976	Originally posted by mattdm Are you sure the Sony cameras just don't have the feature for setting the focal length for old manual lenses? I'm not sure how'd they'd properly work at all without knowing the focal length somehow. I would bet on this. For an IS system to work well, it MUST be tuned based on the lens focal length. As to the in-body element moving more than the IS element in an in-lens system - If the IS element (whether in-lens or in-body) is placed properly, far less movement is needed to achieve the same effect. My guess is that when Canon bashes in-camera IS, they're referring to this fact. Personally, I'd love to see the ability to use in-camera or in-lens (if available) IS depending on what lens was installed. In-camera for short lenses and anything that doesn't have IS, and in-lens for really long lenses.

08-02-2007, 09:12 PM	#8
carpents Veteran Member Join Date: Sep 2006 Location: West Chester, PA Photos: Gallery \| Albums Posts: 1,420	Originally posted by mattdm Are you sure the Sony cameras just don't have the feature for setting the focal length for old manual lenses? I'm not sure how'd they'd properly work at all without knowing the focal length somehow. The Minolta Anti-Shake defaults to roughly 50mm shake reduction with any (every) manual lens. Because they changed their mount in the modern era, it was assumed by Minolta that most users would be using the Minolta A-mount lenses, which have long had electronics (including the focal length) transmitted. Hence no need to account (too much) for manual lenses.

08-03-2007, 02:31 AM	#9
stewart_photo Veteran Member Join Date: Apr 2007 Location: Heidelberg, Germany Posts: 1,864 Original Poster	Originally posted by kenyee This was covered a bit on DPR. Conclusion was in-body VR is more effective than in-lens VR at smaller focal lengths, so the test only confirms this. If you use longer lens, the in-lens VR should be better, so I hope they'll test that case as well... Sadly, DPReview won't let me sign on (doesn't like my yahoo email address), so I very rarely ever visit that site. Anyway, there were reasons why that particular focal length was chosen. One being that prior testing of many lenses revealed very little difference in sharpness between lenses at this focal length (varying sharpness could skew the results). There were others, but I don't remember what they were and don't have the magazine anymore. I wrote the message above about a week ago while a friend was visiting for several days and that friend has now returned home with the magazine. Regardless, since all advanced image stabilization systems, including that in the K10D, compensate to some extent for focal length of the lens, I also doubt there would be that much of a difference in the results if a greater focal length were used. Of course, I can't prove that so we'll just have to agree to disagree with each other's opinions. stewart

08-03-2007, 07:55 AM	#10
axl Veteran Member Join Date: Mar 2007 Location: Nove Zamky, Slovakia Photos: Gallery \| Albums Posts: 7,183	Originally posted by pmperry Yeah but interestingly enough the K10D / GX-10 is the only IS enabled body, that I know of, in which you can choose the focal length of the stabilisation system... let me think... K100D maybe

08-03-2007, 12:32 PM	#11
Deni Veteran Member Join Date: May 2007 Location: Tirana, Albania, South Europe, Planet Earth Photos: Gallery \| Albums Posts: 621	Originally posted by axl let me think... K100D maybe And probably the k100d super as well

08-03-2007, 07:21 PM	#12
Entropy Veteran Member Join Date: Jun 2007 Location: Owego, NY Posts: 976	Originally posted by stewart_photo Sadly, DPReview won't let me sign on (doesn't like my yahoo email address), so I very rarely ever visit that site. Anyway, there were reasons why that particular focal length was chosen. One being that prior testing of many lenses revealed very little difference in sharpness between lenses at this focal length (varying sharpness could skew the results). There were others, but I don't remember what they were and don't have the magazine anymore. I wrote the message above about a week ago while a friend was visiting for several days and that friend has now returned home with the magazine. Regardless, since all advanced image stabilization systems, including that in the K10D, compensate to some extent for focal length of the lens, I also doubt there would be that much of a difference in the results if a greater focal length were used. Of course, I can't prove that so we'll just have to agree to disagree with each other's opinions. stewart I think the one possibility might be that the image displacement induced by camera shake on a really long lens might "peg" the mechanics of an in-body SR system, whereas if the stabilization were in the lens it could be placed in a part of the optical path where a small amount of correction element travel would result in a large correction.

08-12-2007, 09:51 AM	#13
axl Veteran Member Join Date: Mar 2007 Location: Nove Zamky, Slovakia Photos: Gallery \| Albums Posts: 7,183	Originally posted by Deni And probably the k100d super as well you got me on that one

08-12-2007, 10:45 AM	#14
SteveB Veteran Member Join Date: Mar 2007 Location: U.K. Posts: 685	Does in lens IS twist the lens, i.e. just move one side? If so, and pictures I've seen seem to suggest that's how it works, then some of the image must go out of focus, surely, as it's simulating a misaligned lens that owners send back to be repaired because one side of the image is soft! There's a comparison test of in-lens vs. in-body IS in What Digital Camera magazine September issue. Not very scientific as vastly different lenses are used on each camera although they are all used at around 100mm FL (35mm equiv), but the K10D rating is 'very good' at 1/15sec, 'useful' at 1/8sec and ineffective at 1/4sec using a 24-90mm lens but it was tested at 115mm FL. The best performance was the Nikkor 55-200VR tested at 83mm FL (hmm... do I detect bias?), working down to 1/4sec which no others did. The Canon was one of the worst, being ineffective at 1/8sec. Those tested are...... (Optical IS) Canon EOS-30D with 70-200mmf4 IS Panasonic DMC-L1 with Leica 14-50mmf2.8-3.5 Nikon D40x with 55-200mmf4-5.6VR (In body) Pentax K10D with 24-90mmf3.5-4.5 AL IF Olympus E-510 with Zuiko 50-200mmf2.8-3.5 Sony Alpha A100 with Sony 75-300mmf4.5-5.6

08-12-2007, 12:45 PM	#15
arbutusq Veteran Member Join Date: Sep 2006 Location: Vancouver BC canada Photos: Gallery Posts: 380	I can't say which system is better as i only have access to in body, but I can get sharp results at 1/50 and reasonably consistent ones at 1/30 at 450 EFL, that's 3-4 stops advantage and good enough for me. Each person has a slightly different shooting style and will get very different results with each system. I prefer in body because I get disoriented when the image is stabilized in the viewfinder.

Test Results: K10D SR Superior - Even Compared to Some In-Lens IS Systems

Win a Pentax Limited Lens: Raffle Now Live!