Pages: [1]   Go Down

Author Topic: I.B.I.S. and sensor shaking  (Read 6316 times)

allegretto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 660
I.B.I.S. and sensor shaking
« on: November 05, 2014, 07:30:10 am »

how does this work without degrading the image?

If it "knows" the frequency and amplitude of the shake it would be easy enough to track, but then how would that improve the photographer's stability?

Further it would be nice to know the G-forces since that has to take a toll on sensor life...No?

Thanks in advance...
Logged

Manoli

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2296
Re: I.B.I.S. and sensor shaking
« Reply #1 on: November 05, 2014, 08:53:13 am »

Further it would be nice to know the G-forces since that has to take a toll on sensor life...No?

I don't know about taking a toll on sensor life, why would it ? Sensors are normally 'screwed' into the body - fixed. Here you're suspending the sensor within the IBIS mechanism. I'd be far more concerned about how resistant the system would be to the odd accidental drop, knock etc.

Someone once knocked one of my Sony's 5 ft off a bar, straight drop onto a concrete floor. Didn't do the lens much good but the cam worked fine! Not sure that would have been the case if it had delicate internal quivering parts.

5-way IBIS is great for hand held shooting, but you need the mechanism to be A-OK, I'd hate to think what it'll do if it's out-of-alignment. Having said that, I've never heard of an Olympus EM user complain or suffer a failure.

Logged

allegretto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 660
Re: I.B.I.S. and sensor shaking
« Reply #2 on: November 05, 2014, 12:05:23 pm »

I don't know about taking a toll on sensor life, why would it ? Sensors are normally 'screwed' into the body - fixed. Here you're suspending the sensor within the IBIS mechanism. I'd be far more concerned about how resistant the system would be to the odd accidental drop, knock etc.





In my experience sometimes the g-force necessary to reach the goals can be hard on equipment. Consider the targeting mechanisms on the fast lasers. The wear themselves out and require constant maintenance. The mass here is much less and perhaps it's all trivial. But my question is how does the software get synched and interpret the data?
Logged

Telecaster

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3686
Re: I.B.I.S. and sensor shaking
« Reply #3 on: November 05, 2014, 05:09:10 pm »

My E-M5 & E-M1 have been all over the place—plane rides, car trips—over the past year + (20 months in the E-M5's case) without IBIS incident. I've never dropped either camera, though. In general I try not to do that.   :D  The system works in much the same way as optical IS but results in moving the sensor itself, rather than elements within the lens, in response to user motion. It's the most effective IS system I've come across so far.

-Dave-
Logged

spidermike

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 535
Re: I.B.I.S. and sensor shaking
« Reply #4 on: November 13, 2014, 11:37:54 am »

but then how would that improve the photographer's stability?



I am not sure what you mean by that
Logged

allegretto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 660
Re: I.B.I.S. and sensor shaking
« Reply #5 on: November 13, 2014, 06:05:02 pm »

I am not sure what you mean by that

 What was meant is that the IBIS increases the apparent stability of the photographer/camera.

It has no effect on the subject motion...
Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: I.B.I.S. and sensor shaking
« Reply #6 on: November 14, 2014, 01:01:10 am »

Do we ever know what part of a camera is most likely to fail? It could be memory or buttons or the LCD or ibis. My take is that we have to "trust" the engineers and manufacturers to do a good job, and vote wit hour wallet whenever they do a poor job.

-h
Logged

allegretto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 660
Re: I.B.I.S. and sensor shaking
« Reply #7 on: November 14, 2014, 01:28:14 am »

Do we ever know what part of a camera is most likely to fail? It could be memory or buttons or the LCD or ibis. My take is that we have to "trust" the engineers and manufacturers to do a good job, and vote wit hour wallet whenever they do a poor job.

-h

Well, to be precise I wasn't specifically asking about "failure", just noted that in some equipment that could happen

My original question is, "... how does shaking the sensor make the camera appear to be more stable...?"

Likely its some kind of self-synch algorithm or some more sophisticated derivative, but perhaps it's even more complex or simpler than that... I was just asking it anyone knew more about it...
Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: I.B.I.S. and sensor shaking
« Reply #8 on: November 14, 2014, 03:35:01 am »

Well, to be precise I wasn't specifically asking about "failure", just noted that in some equipment that could happen

My original question is, "... how does shaking the sensor make the camera appear to be more stable...?"

Likely its some kind of self-synch algorithm or some more sophisticated derivative, but perhaps it's even more complex or simpler than that... I was just asking it anyone knew more about it...
If some gyro/accelerometer says that the camera is rotated/moved so and so, there ought to be some countering sensor movement that (within the constraints of image circle) maintains a minimum smear projected on to the sensor?

Having more axes and more accuracy in motion estimation and sensor movement should only make this better?

-h
Logged

deejjjaaaa

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1170
Re: I.B.I.S. and sensor shaking
« Reply #9 on: November 14, 2014, 09:34:17 am »

Having more axes and more accuracy in motion estimation and sensor movement should only make this better?

as E-M1 example proved IBIS can't cope with shutter shake... only introduction of EFCS in E-M1 with a new firmware was able to fix that... IBIS can cope with slow frequency moves induced by your hands tremor only (ор from a finger movement to release the shutter)... move to a territory of shutter impact and all your axes are hopeless, too slow to react and/or not capable to react with proper direction/precision (even when camera's firmware or whatever sensor/ASIC inside knows that shutter was just released and shake from its blade might be forthcoming)
« Last Edit: November 14, 2014, 09:37:00 am by deejjjaaaa »
Logged

allegretto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 660
Re: I.B.I.S. and sensor shaking
« Reply #10 on: November 14, 2014, 10:18:07 pm »

as E-M1 example proved IBIS can't cope with shutter shake... only introduction of EFCS in E-M1 with a new firmware was able to fix that... IBIS can cope with slow frequency moves induced by your hands tremor only (ор from a finger movement to release the shutter)... move to a territory of shutter impact and all your axes are hopeless, too slow to react and/or not capable to react with proper direction/precision (even when camera's firmware or whatever sensor/ASIC inside knows that shutter was just released and shake from its blade might be forthcoming)

so you're saying it's best a high-amplitude, low-frequency motion

makes sense of course.

but is it just attaching the sensor to some electro-gyros? and they seek a single vectored axis orientation?
Logged

schrodingerscat

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 374
Re: I.B.I.S. and sensor shaking
« Reply #11 on: November 17, 2014, 02:29:48 pm »

As a tech for the last 30 years, maybe I can inject a little empirical knowledge.

There is no "Sensor Shake" and wondering if people are confusing sensor IS with vibrating the low pass for dust removal. IS via the sensor is done through an X/Y movement facilitated by magnets and suspending the sensor assembly. Lens IS works much the same, but shifts an element assembly instead. The camera's processor and firmware determine how to move the sensor/elements in relation to the motion detected by the sensors. This movement is planar, and some Z axis movement may come into play due to manufacturing and design tolerances.

Shutter vibration should have little or no effect on IS, tho a side effect may be relying on longer speeds due to the benefits of IS. Mechanical shutters move mass at a high rate of speed, that is then stopped short. Inertia then comes into play. Modern vertical travel blade type shutters have to have their curtain springs wound rather tightly to squeak 1/2000 and higher out of them. Was the same in the film days.
Logged

deejjjaaaa

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1170
Re: I.B.I.S. and sensor shaking
« Reply #12 on: November 17, 2014, 02:53:20 pm »

There is no "Sensor Shake"

call it whatever you want = http://www.dpreview.com/forums/thread/3651827

you might have such an oscilloscope and reproduce the experiment  ;)
Logged

deejjjaaaa

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1170
Re: I.B.I.S. and sensor shaking
« Reply #13 on: November 17, 2014, 02:55:30 pm »

IS via the sensor is done through an X/Y movement



Logged

allegretto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 660
Re: I.B.I.S. and sensor shaking
« Reply #14 on: November 18, 2014, 08:51:43 am »

with ya so far... but how does it "know" how much

it can't focus nearly as fast. Lens elements no doubt are no doubt more mass than a sensor

but how does it approximate proper displacement

or is it independent of the motion of the holder and a fixed rate of "refresh"
Logged

deejjjaaaa

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1170
Re: I.B.I.S. and sensor shaking
« Reply #15 on: November 18, 2014, 09:43:40 am »

Lens elements no doubt are no doubt more mass than a sensor

why no doubts

so how many lens elements OIS uses and what is their size & weight exactly ? no it is not the front element... and why do you think that OIS element is always made of glass ? and IBIS does not move a sensor (silicone chip itself) alone, it moves it with a lot of things making a sensor assembly and that includes things like IR/UV/AA/etc filters present of top of CFA in sensor assembly - and those are not plastic... so why no doubts ?
Logged

allegretto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 660
Re: I.B.I.S. and sensor shaking
« Reply #16 on: November 18, 2014, 08:22:48 pm »

why no doubts

so how many lens elements OIS uses and what is their size & weight exactly ? no it is not the front element... and why do you think that OIS element is always made of glass ? and IBIS does not move a sensor (silicone chip itself) alone, it moves it with a lot of things making a sensor assembly and that includes things like IR/UV/AA/etc filters present of top of CFA in sensor assembly - and those are not plastic... so why no doubts ?

I see

so you feel it likely that a sensor weighs more than a lens element?

and you appear to feel that the Ibis must move several items with extreme accuracy

but you don't really know how the algorithm works

is that what you're saying...?

just trying to be clear...
Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: I.B.I.S. and sensor shaking
« Reply #17 on: November 20, 2014, 02:25:47 am »

with ya so far... but how does it "know" how much
I don't get what you are continually asking for. My mobile phone "knows" where it is in the world. It "knows" it when I shake it. It has a lot of knowledge about orientation and movement recorded by low-cost sensors. One would assume that camera makers are able to integrate similar sensors?

Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: I.B.I.S. and sensor shaking
« Reply #18 on: November 20, 2014, 02:31:41 am »

so you feel it likely that a sensor weighs more than a lens element?
In order to stabilize a given camera movement "x", you could either move the sensor by "y" or some optical element by "z".

As long as no one has offered estimates on how large those movements might be, the mass that needs to be moved, or the intrinsic limitations to either stabilization approach (image circle etc), I feel that concentrating on _one_ parameter is a bad idea. Hint: moving a large mass by a small amount need not be any "harder" than moving a small mass by a large amount.

Would it not be more fruitful to discuss actual performance of various stabilization systems? I believe that there are standardized tests out there?

-h
Logged

allegretto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 660
Re: I.B.I.S. and sensor shaking
« Reply #19 on: November 21, 2014, 01:22:00 pm »

Well it would be if that's the question I wondered about.

Used IS from Oly, Canon, Nikon and Sony and they all seem to help in the proper setting. If one is a stop or two better, fine. But my question was not "DO" or "HOW MUCH". My question was "HOW"

No problem. No one who's answered seems to really know but likely different manufacturers use different ones unless they license.

Secondly, there is a great deal of difference in a system moving a small mass far and a large mass a very small distance. Yes, it's possible to do the same "work", but the Engineering is very different. Anyway I suspect they are moving a very small mass a very small amount


In order to stabilize a given camera movement "x", you could either move the sensor by "y" or some optical element by "z".

As long as no one has offered estimates on how large those movements might be, the mass that needs to be moved, or the intrinsic limitations to either stabilization approach (image circle etc), I feel that concentrating on _one_ parameter is a bad idea. Hint: moving a large mass by a small amount need not be any "harder" than moving a small mass by a large amount.

Would it not be more fruitful to discuss actual performance of various stabilization systems? I believe that there are standardized tests out there?

-h
Logged
Pages: [1]   Go Up