K-5 Anti-shake (image stabilization) test

Skymist · 05-01-2011, 08:47 PM

If the SR sensor was contributing movement for cases where the camera was still, it would have shown up in the photo below. Here I was meteor hunting, and the photo is a 100% crop of a very small portion of the original, showing a satellite I caught. If the SR was doing things the trails of the stars and the satellite would be wavy or wiggly. This is an extremely sensitive test. The satellite image is barely 2 pixels wide, if that. I bring this up because I made the mistake of leaving SR on for this shot, and I thought, "Oh no, they will all be ruined" but they turned out fine. I could not have improved on the sharpness. This is better than any camera I've owned. Detail: Pentax K-5, 10s exposure by program delay, F/1.4, 50mm using Zeiss Planar T* 50mm F/1.4 ZK lens. Satellite at almost dead center, moving right to left. This image was checked with other sequential images to confirm it is a satellite.

Skymist · 05-01-2011, 10:02 PM

Satellite photo previous at 500%, 5 pixels per pixel. You can see that the track Fullwidth Half Max comes to 2 pixels or maybe just under that. I don't normally pixel peek but I was curious myself what it looked like.

30v iPentax · 05-01-2011, 11:04 PM

Ompa mielenkiintoista, mutta mihin tätä tarvitaan.

Jimfear · 05-01-2011, 11:35 PM

Originally posted by Skymist

If the SR sensor was contributing movement for cases where the camera was still, it would have shown up in the photo below. Here I was meteor hunting, and the photo is a 100% crop of a very small portion of the original, showing a satellite I caught. If the SR was doing things the trails of the stars and the satellite would be wavy or wiggly. This is an extremely sensitive test. The satellite image is barely 2 pixels wide, if that. I bring this up because I made the mistake of leaving SR on for this shot, and I thought, "Oh no, they will all be ruined" but they turned out fine. I could not have improved on the sharpness. This is better than any camera I've owned. Detail: Pentax K-5, 10s exposure by program delay, F/1.4, 50mm using Zeiss Planar T* 50mm F/1.4 ZK lens. Satellite at almost dead center, moving right to left. This image was checked with other sequential images to confirm it is a satellite.

Not to spoil the fun, but doesn't the K-5 automatically turn SR off if you're using timer to set it off. And it probably does so for the very reason that people sometimes forget to do it.

Gimbal · 05-01-2011, 11:42 PM

Originally posted by Skymist

I bring this up because I made the mistake of leaving SR on for this shot, and I thought, "Oh no, they will all be ruined" but they turned out fine. I could not have improved on the sharpness. This is better than any camera I've owned. Detail: Pentax K-5, 10s exposure by program delay,

The need to turn off SR when mounted on a tripod is greatly exaggerated IMHO. Unless you have a very large focal length combined with a very weak tripod, then it might be a problem, maybe.

And usually when using a tripod one also use a cable release, IR-remote or the built in shutter delay in which case the camera assume it is mounted on a tripod and automatically shuts off SR.

Edit, Jimfear was faster...

Skymist · 05-02-2011, 12:44 PM

Ok then, I suppose my time exposures do not prove anything about the stability of the SR module. I guess I need to repeat my original study with a tripod shot in which the shutter is hand actuated, to make sure that the SR is active. Let's see if there's any unwanted reaction of the SR module. Darn, I really need some neutral filters to do this right and I don't have any.

30v iPentax · 05-02-2011, 10:24 PM

Minlla on hyvä kuvanvakain Manfrotto 055XPROOB. Tulee hyviä ja teräviä kuvia.

Eruditass · 05-03-2011, 12:06 AM

Originally posted by HoundFrog

But what is clear from Falk's work, that SR design aggravates the shutter-induced blur.

I think you should re-read the conclusion or send Falk a pm.

kamisu · 03-26-2012, 12:17 AM

Although it is an old thread, but both labo fnac and digitalcamerainfo gave low scores for K-5 stabilization system.

Wheatfield · 03-26-2012, 06:25 AM

Originally posted by HoundFrog

Yes, but the reason of this blur is the particular SR design. Sensor is never hold firm, is floats in magnetic field that provides a limited torque. When shutter is released, mechanical impulse disturbs the pendant

If sensor is not floating, that impulse interacts with the mass of not only sensor, but with the mass of entire camera which is about 50-100 times bigger. So the effect is smaller proportionally.

This isn't actually correct. The blur these cameras can suffer from is caused by either the mirror or shutter movement, and has absolutely nothing to do with the SR mechanism. Many cameras have some harmonic vibration at certain speeds. Generally, 1/8 to 1/30 second are problematic with older Pentax film camera, for example.

Originally posted by Skymist

Ok then, I suppose my time exposures do not prove anything about the stability of the SR module. I guess I need to repeat my original study with a tripod shot in which the shutter is hand actuated, to make sure that the SR is active. Let's see if there's any unwanted reaction of the SR module. Darn, I really need some neutral filters to do this right and I don't have any.

Remember that you can never turn SR off completely, all you can do is disable it. The mechanism that moves the sensor for SR is the same one that is holding the sensor still when SR is disabled.

03-26-2012, 06:25 AM	#25
Wheatfield Moderator Join Date: Apr 2008 Location: The wheatfields of Canada Posts: 15,981	Originally posted by HoundFrog Yes, but the reason of this blur is the particular SR design. Sensor is never hold firm, is floats in magnetic field that provides a limited torque. When shutter is released, mechanical impulse disturbs the pendant If sensor is not floating, that impulse interacts with the mass of not only sensor, but with the mass of entire camera which is about 50-100 times bigger. So the effect is smaller proportionally. This isn't actually correct. The blur these cameras can suffer from is caused by either the mirror or shutter movement, and has absolutely nothing to do with the SR mechanism. Many cameras have some harmonic vibration at certain speeds. Generally, 1/8 to 1/30 second are problematic with older Pentax film camera, for example. Originally posted by Skymist Ok then, I suppose my time exposures do not prove anything about the stability of the SR module. I guess I need to repeat my original study with a tripod shot in which the shutter is hand actuated, to make sure that the SR is active. Let's see if there's any unwanted reaction of the SR module. Darn, I really need some neutral filters to do this right and I don't have any. Remember that you can never turn SR off completely, all you can do is disable it. The mechanism that moves the sensor for SR is the same one that is holding the sensor still when SR is disabled. Last edited by Wheatfield; 03-26-2012 at 06:30 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
image stabilization performance?	Shadowghost	Pentax K-r	6	02-03-2011 08:28 AM
Broken Anti-Shake results in Focus issues	volvomanny	Pentax DSLR Discussion	2	01-18-2011 07:33 AM
anti shake on or off w/monopod	Lance C	Photographic Technique	4	09-26-2010 08:52 AM
[K7]When absolutly NOT to use Image Stabilization?	Ivo_Spohr	Pentax DSLR Discussion	27	09-10-2010 12:49 AM
Anti-Shake/Tripod?	CPB.IV	Troubleshooting and Beginner Help	3	07-24-2009 10:28 AM

05-01-2011, 08:47 PM	#16
Skymist Senior Member Join Date: Apr 2011 Location: Monterey Bay, Callifornia Posts: 107 Original Poster	If the SR sensor was contributing movement for cases where the camera was still, it would have shown up in the photo below. Here I was meteor hunting, and the photo is a 100% crop of a very small portion of the original, showing a satellite I caught. If the SR was doing things the trails of the stars and the satellite would be wavy or wiggly. This is an extremely sensitive test. The satellite image is barely 2 pixels wide, if that. I bring this up because I made the mistake of leaving SR on for this shot, and I thought, "Oh no, they will all be ruined" but they turned out fine. I could not have improved on the sharpness. This is better than any camera I've owned. Detail: Pentax K-5, 10s exposure by program delay, F/1.4, 50mm using Zeiss Planar T* 50mm F/1.4 ZK lens. Satellite at almost dead center, moving right to left. This image was checked with other sequential images to confirm it is a satellite.

05-01-2011, 10:02 PM	#17
Skymist Senior Member Join Date: Apr 2011 Location: Monterey Bay, Callifornia Posts: 107 Original Poster	Satellite photo previous at 500%, 5 pixels per pixel. You can see that the track Fullwidth Half Max comes to 2 pixels or maybe just under that. I don't normally pixel peek but I was curious myself what it looked like.

05-01-2011, 11:04 PM	#18
30v iPentax Inactive Account Join Date: Apr 2010 Location: Suomi , Finland Posts: 368	Ompa mielenkiintoista, mutta mihin tätä tarvitaan.

05-01-2011, 11:35 PM	#19
Jimfear Veteran Member Join Date: Apr 2007 Location: Uppsala, Sweden Photos: Gallery Posts: 576	Originally posted by Skymist If the SR sensor was contributing movement for cases where the camera was still, it would have shown up in the photo below. Here I was meteor hunting, and the photo is a 100% crop of a very small portion of the original, showing a satellite I caught. If the SR was doing things the trails of the stars and the satellite would be wavy or wiggly. This is an extremely sensitive test. The satellite image is barely 2 pixels wide, if that. I bring this up because I made the mistake of leaving SR on for this shot, and I thought, "Oh no, they will all be ruined" but they turned out fine. I could not have improved on the sharpness. This is better than any camera I've owned. Detail: Pentax K-5, 10s exposure by program delay, F/1.4, 50mm using Zeiss Planar T* 50mm F/1.4 ZK lens. Satellite at almost dead center, moving right to left. This image was checked with other sequential images to confirm it is a satellite. Not to spoil the fun, but doesn't the K-5 automatically turn SR off if you're using timer to set it off. And it probably does so for the very reason that people sometimes forget to do it.

05-01-2011, 11:42 PM	#20
Gimbal Pentaxian Join Date: Apr 2007 Location: Sweden Posts: 2,106	Originally posted by Skymist I bring this up because I made the mistake of leaving SR on for this shot, and I thought, "Oh no, they will all be ruined" but they turned out fine. I could not have improved on the sharpness. This is better than any camera I've owned. Detail: Pentax K-5, 10s exposure by program delay, The need to turn off SR when mounted on a tripod is greatly exaggerated IMHO. Unless you have a very large focal length combined with a very weak tripod, then it might be a problem, maybe. And usually when using a tripod one also use a cable release, IR-remote or the built in shutter delay in which case the camera assume it is mounted on a tripod and automatically shuts off SR. Edit, Jimfear was faster...

05-02-2011, 12:44 PM	#21
Skymist Senior Member Join Date: Apr 2011 Location: Monterey Bay, Callifornia Posts: 107 Original Poster	Ok then, I suppose my time exposures do not prove anything about the stability of the SR module. I guess I need to repeat my original study with a tripod shot in which the shutter is hand actuated, to make sure that the SR is active. Let's see if there's any unwanted reaction of the SR module. Darn, I really need some neutral filters to do this right and I don't have any.

05-02-2011, 10:24 PM	#22
30v iPentax Inactive Account Join Date: Apr 2010 Location: Suomi , Finland Posts: 368	Minlla on hyvä kuvanvakain Manfrotto 055XPROOB. Tulee hyviä ja teräviä kuvia.

05-03-2011, 12:06 AM	#23
Eruditass Veteran Member Join Date: Jul 2009 Photos: Gallery Posts: 2,207	Originally posted by HoundFrog But what is clear from Falk's work, that SR design aggravates the shutter-induced blur. I think you should re-read the conclusion or send Falk a pm.

03-26-2012, 12:17 AM	#24
kamisu Senior Member Join Date: Feb 2012 Location: Iowa Posts: 130	Although it is an old thread, but both labo fnac and digitalcamerainfo gave low scores for K-5 stabilization system.