PSA: Sigma lens IS doesn't turn off when you turn it off?!

[ahsanford]

Dude had a shoot ruined by the IS on his Sigma not really being off:

https://petapixel.com/2018/03/08/sigma-art-lens-messed-ferrari-photo-shoot/

"A few days later, I had a chat with SIGMA service personnel. They explained to me that even though you switch stabilization on the lens (OS) to off, it doesn’t actually mean it’s off… The stabilization element is kept in its position by a magnetic field. Whenever the camera goes to sleep, the element moves down. After waking up the camera, it takes a moment to reposition itself to the proper position. And it’s even doing this during exposures!"

- A

 

[9VIII]

...shouldn’t this appy to all stabilized lenses?
This should be the same on Canon lenses when you turn the camera off, you probably can’t count on a clear image if you want to go from “off” to shooting within a few seconds (maybe more relevant to mirrorless?), and any lens with IS that gets adapted to a body without electronic contacts will just have a de-centered element the whole time.

 

[jd7]

If I understand the article correctly, the issue is that when the lens has no power, the OS lens group moves to a "parked" position that isn't ready to go for the next shot, so when the lens is turned on, there is a short delay before the OS group is back in position and ready to go ... but then it stays in the same place until the lens goes to sleep again. (And with OS turned off, the lens doesn't try to steady the shot - so it is off.)

If that is correct, the photographer just needs wake up the camera/lens a few moments before starting an exposure, if the camera has gone to sleep between shots. I therefore don't understand why the author was so quick to sell the lens. It was a lens he liked and it was working fine, it just operates in a way he hadn't previously appreciated.

I believe all IS/VR/VC systems need to be activated a short period before taking a shot to give them a chance to activate properly, so as has been suggested above, perhaps the stabilisation systems from all manufacturers have the same issue that Sigma's OS has in this regard?

 

[hne]

If I understand the article correctly, the issue is that when the lens has no power, the OS lens group moves to a "parked" position that isn't ready to go for the next shot, so when the lens is turned on, there is a short delay before the OS group is back in position and ready to go ... but then it stays in the same place until the lens goes to sleep again. (And with OS turned off, the lens doesn't try to steady the shot - so it is off.)

The problem presented in the linked article is as such: The Sigma lens has a floating IS element when IS is turned off. If you take a series of exposures on a sturdy tripod, they will not be aligned.

If that is correct, the photographer just needs wake up the camera/lens a few moments before starting an exposure, if the camera has gone to sleep between shots. I therefore don't understand why the author was so quick to sell the lens. It was a lens he liked and it was working fine, it just operates in a way he hadn't previously appreciated.

I believe all IS/VR/VC systems need to be activated a short period before taking a shot to give them a chance to activate properly, so as has been suggested above, perhaps the stabilisation systems from all manufacturers have the same issue that Sigma's OS has in this regard?

Yes, optical image stabilisation systems tend to jump a bit on start up. The author had image stabilisation turned off and still got a jump, causing different apparent camera position from shot to shot.

 

[jd7]

The problem presented in the linked article is as such: The Sigma lens has a floating IS element when IS is turned off. If you take a series of exposures on a sturdy tripod, they will not be aligned.

Hi hne

Could you point me to the part of the article which says that? To me, the key part seemed to be: "Whenever the camera goes to sleep, the element moves down. After waking up the camera, it takes a moment to reposition itself to the proper position." Given that, I cannot see why you would expect any problem taking shots on a tripod unless you let the camera go to sleep between shots ... and didn't give it time to get back in position after waking it up again before starting the next exposure. Am I missing something?

 

[Mt Spokane Photography]

The problem presented in the linked article is as such: The Sigma lens has a floating IS element when IS is turned off. If you take a series of exposures on a sturdy tripod, they will not be aligned.

Hi hne

Could you point me to the part of the article which says that? To me, the key part seemed to be: "Whenever the camera goes to sleep, the element moves down. After waking up the camera, it takes a moment to reposition itself to the proper position." Given that, I cannot see why you would expect any problem taking shots on a tripod unless you let the camera go to sleep between shots ... and didn't give it time to get back in position after waking it up again before starting the next exposure. Am I missing something?

The problem with the post, is that it leaves people wondering exactly what the issue was, he does not say exactly why the issue, it may have been obvious to him, but you can see that comments from readers are guesses and assumptions.

I would have not picked a 24-105 for a high end shoot.

 

[hne]

"When checking my shots in Capture One (as I always shoot tethered), I realized that the car was jumping up and down in the frame"

and

"The stabilization element is kept in its position by a magnetic field. Whenever the camera goes to sleep, the element moves down. After waking up the camera, it takes a moment to reposition itself to the proper position. And it’s even doing this during exposures!"

So, the image stabilization has no mechanical stable mode even when turned off and the stabilizing element repositions itself every time you wake it up plus every few seconds (once in a 5s exposure at least).

 

[lexaclarke]

Isn't this normal? I have the Canon 70-200 f/4 IS and I get a slight sense of something moving within the lens when the IS is off. I assume the IS elements can't just suddenly be glued in place so there's bound to be some give. I thought that was why some companies say IS lowers image quality, because there will always be an element you can't guarantee is in exactly the right position.

 

[jd7]

Isn't this normal? I have the Canon 70-200 f/4 IS and I get a slight sense of something moving within the lens when the IS is off. I assume the IS elements can't just suddenly be glued in place so there's bound to be some give. I thought that was why some companies say IS lowers image quality, because there will always be an element you can't guarantee is in exactly the right position.

That is my understanding also.

 

[jd7]

"When checking my shots in Capture One (as I always shoot tethered), I realized that the car was jumping up and down in the frame"

and

"The stabilization element is kept in its position by a magnetic field. Whenever the camera goes to sleep, the element moves down. After waking up the camera, it takes a moment to reposition itself to the proper position. And it’s even doing this during exposures!"

So, the image stabilization has no mechanical stable mode even when turned off and the stabilizing element repositions itself every time you wake it up ...

Agreed.

I would be very interested to know whether other manufacturers have stabilisation systems which have mechanically lock the IS lens group in the neutral position (if that is the correct term) when the stabilising system is turned off.

... plus every few seconds (once in a 5s exposure at least).

I have to say that is not my reading of the article. I agree with Mt Spokane Photography that the article isn't very clear about the problem the author experienced, but my interpretation was he was saying the problem occurred only at the start of an exposure, eg the article says that the front of a car was not sharp while the rest was. Since he was light painting the car during the 5 second exposure, I assume he started light painting at the front of the car, and there was no problem with the rest of the car because by the time he got to it the lens had settled down.

 

[Quackator]

Looks like the problem is more that the D850 sleep mode
doesn't keep the stabilizer powered during long exposures.

Remember Nikon sued Sigma over VR. They might as well
try to sabotage Sigma. Keeping things proprietary already
killed their waferstepper business.

They are recycling this strategy now.

Meanwhile, Sigma Global Vision lenses for Canon receive
firmware updates, adding onboard correction data and
greatly improving AF performance.

Blaming Sigma seems to be a pretty lame excuse.

 

[pwp]

Have used Canon IS glass since the 1990's starting with the very likeable 300 f/4IS and never experienced anything like the issue described. Maybe because I hardly ever turn IS off. :

@Mt Spokane, agree, a Sigma 24-105 does seem like an odd choice for a what he describes as a high-end shoot.

I've made impulse Sigma buys from time to time, done the tests, and never actually trusted any one of them on a commercial project. Money down the drain in my case. They just couldn't be trusted to deliver the goods for paying clients. I'm sure there are entirely dependable Sigmas out there. Just that none came my way...

Like Willy Wonka's (and the Chocolate Factory) dentist Dad famously said, "why take the chance?"

-pw

 

[hne]

"When checking my shots in Capture One (as I always shoot tethered), I realized that the car was jumping up and down in the frame"

and

"The stabilization element is kept in its position by a magnetic field. Whenever the camera goes to sleep, the element moves down. After waking up the camera, it takes a moment to reposition itself to the proper position. And it’s even doing this during exposures!"

So, the image stabilization has no mechanical stable mode even when turned off and the stabilizing element repositions itself every time you wake it up ...

Agreed.

I would be very interested to know whether other manufacturers have stabilisation systems which have mechanically lock the IS lens group in the neutral position (if that is the correct term) when the stabilising system is turned off.

... plus every few seconds (once in a 5s exposure at least).

I have to say that is not my reading of the article. I agree with Mt Spokane Photography that the article isn't very clear about the problem the author experienced, but my interpretation was he was saying the problem occurred only at the start of an exposure, eg the article says that the front of a car was not sharp while the rest was. Since he was light painting the car during the 5 second exposure, I assume he started light painting at the front of the car, and there was no problem with the rest of the car because by the time he got to it the lens had settled down.

The person writing the article stated that in a 5s long light painting exposure, the rear of the car was sharp and the front blurry, and added this picture to the article:

 

[old-pr-pix]

Agree with MSP the post isn't very clear. AFAIK all IS systems that have moving parts (lens element or IBIS) need power to 'center' the element. Hence there are two very different off conditions - off with no power where the element is allowed to go where ever gravity takes it - and off with power where it should be centered but not allowed to react to any motion inputs. It seems reasonable that any camera going to sleep would un-power the attached lens. It's an interesting question whether any IS/VR/IBIS system has a power off mode with some type of physically centered, locked-in-place condition?

As to the 24-105 lens, the Sigma Art variety is known to be much higher resolution than Canon variety as well as 2/3 stop brighter. But why risk the shoot with a zoom of any type? Slightest bump of the zoom ring would throw images out of alignment from change in FL. Then he would have had real problems.

 

[hne]

Agree with MSP the post isn't very clear. AFAIK all IS systems that have moving parts (lens element or IBIS) need power to 'center' the element. Hence there are two very different off conditions - off with no power where the element is allowed to go where ever gravity takes it - and off with power where it should be centered but not allowed to react to any motion inputs. It seems reasonable that any camera going to sleep would un-power the attached lens. It's an interesting question whether any IS/VR/IBIS system has a power off mode with some type of physically centered, locked-in-place condition?

According to Canon:
"During typical operation, when the camera’s meter system turns off, the IS system is signaled by the camera to lock and center the correction elements and using power from the camera, the IS system performs this task."

Source: https://www.learn.usa.canon.com/resources/articles/2014/is_lens_rattling.htmlp

 

[Mt Spokane Photography]

As to the 24-105 lens, the Sigma Art variety is known to be much higher resolution than Canon variety as well as 2/3 stop brighter. But why risk the shoot with a zoom of any type? Slightest bump of the zoom ring would throw images out of alignment from change in FL. Then he would have had real problems.

There is not enough difference to bother with,

http://www.opticallimits.com/canon_eos_ff/864-sigma24105f4eos?start=2

 

[jd7]

According to Canon:
"During typical operation, when the camera’s meter system turns off, the IS system is signaled by the camera to lock and center the correction elements and using power from the camera, the IS system performs this task."

Source: https://www.learn.usa.canon.com/resources/articles/2014/is_lens_rattling.htmlp

Interesting! Presumably if Canon's IS system parks the IS lens group in the centred position when the IS system is on and the camera powers down, that is also where the IS lens group is parked when the IS system is off, so it sounds like Canon's IS system does work differently from Sigma's OS in that regard. I wonder why Sigma's system parks the IS lens group in a non-centred position (assuming I am correct in understanding that is how Sigma's system works)?

 

[AlanF]

As to the 24-105 lens, the Sigma Art variety is known to be much higher resolution than Canon variety as well as 2/3 stop brighter. But why risk the shoot with a zoom of any type? Slightest bump of the zoom ring would throw images out of alignment from change in FL. Then he would have had real problems.

There is not enough difference to bother with,

http://www.opticallimits.com/canon_eos_ff/864-sigma24105f4eos?start=2

Lensrentals in its usual rigorous analysis has the Sigma rather better at shorter focal lengths. The Nikon 24-120mm seems the best.
https://www.lensrentals.com/blog/2018/02/mtf-testing-for-the-sigma-24-105mm-f4-dg-os-hsm-art-series-lens/

 

[lexaclarke]

According to Canon:
"During typical operation, when the camera’s meter system turns off, the IS system is signaled by the camera to lock and center the correction elements and using power from the camera, the IS system performs this task."

Source: https://www.learn.usa.canon.com/resources/articles/2014/is_lens_rattling.htmlp

Interesting! Presumably if Canon's IS system parks the IS lens group in the centred position when the IS system is on and the camera powers down, that is also where the IS lens group is parked when the IS system is off, so it sounds like Canon's IS system does work differently from Sigma's OS in that regard. I wonder why Sigma's system parks the IS lens group in a non-centred position (assuming I am correct in understanding that is how Sigma's system works)?

I think the point is Sigma's doesn't 'park' the element at all. If there is power it keeps it floating in place, and if there's not power then it just settles wherever it likes. At least I think that's what this is all getting at.

 

[hne]

According to Canon:
"During typical operation, when the camera’s meter system turns off, the IS system is signaled by the camera to lock and center the correction elements and using power from the camera, the IS system performs this task."

Source: https://www.learn.usa.canon.com/resources/articles/2014/is_lens_rattling.htmlp

Interesting! Presumably if Canon's IS system parks the IS lens group in the centred position when the IS system is on and the camera powers down, that is also where the IS lens group is parked when the IS system is off, so it sounds like Canon's IS system does work differently from Sigma's OS in that regard. I wonder why Sigma's system parks the IS lens group in a non-centred position (assuming I am correct in understanding that is how Sigma's system works)?

I think the point is Sigma's doesn't 'park' the element at all. If there is power it keeps it floating in place, and if there's not power then it just settles wherever it likes. At least I think that's what this is all getting at.

That's what I tried to convey. Thank you for clarifying.