5-axis IBIS in a P&S

Started Aug 8, 2013 | Discussions
tjuster1
Senior MemberPosts: 1,512
Like?
5-axis IBIS in a P&S
Aug 8, 2013

Just saw this on dpreview's front page: http://www.dpreview.com/news/2013/08/08/hold-steady-casio-exilim-ex-zr800-gains-five-axis-stabilization-16MP-BSI-CMOS-25-400mm-zoom#specs

The most interesting thing to me is that Casio has managed to cram 5-axis IBIS into a body just a hair larger than PM2. Now admittedly I don't know how good it is, or how much smaller the IBIS has to be for the smaller sensor, but at least this hints at the possibility of someday seeing superior IBIS in a truly small body, something I've been hoping for for a long time.

Bizzarrini
Regular MemberPosts: 482Gear list
Like?
Re: 5-axis IBIS in a P&S
In reply to tjuster1, Aug 8, 2013

I'm not so sure it actually works like the Oly system. From the press release:

"HS Anti Shake performs 5 - axis, 5 - stop*5 image stabilization

Camera shake which occurs in the five axes*5 of pitch, yaw, vertical shift, horizontal shift and roll is precisely corrected. Casio has employed an optical lens that achieves stabilization equivalent to approximately 2 stops*6 slower in terms of shutter speed. Bringing this effect together with combined high - speed burst images results in image stabilization equivalent to about 5 stops slower*5 in terms of shutter speed. This allows users to minimize blur while holding the camera to take pictures, which can easily result in camera shake, even for shots of night scenes and when zooming. (Results may vary depending on the shooting environment and camera settings). In addition, the EX-ZR800 employs a lens-shift optical lens, which further reduces image blur on the LCD screen when recording movies or holding the camera."

 Bizzarrini's gear list:Bizzarrini's gear list
Panasonic Lumix DMC-ZS7 Fujifilm FinePix X100 Olympus Tough TG-2 Panasonic Lumix DMC-GF1 Fujifilm X-Pro1 +22 more
Reply   Reply with quote   Complain
ntsan
Contributing MemberPosts: 723Gear list
Like?
Re: 5-axis IBIS in a P&S
In reply to tjuster1, Aug 8, 2013

Panasonic and Canon also have "5-axis" IS too on their DC, but their one is combining Lens IS + Electronic IS, where else Olympus one is full mechanical 5 axis IS

 ntsan's gear list:ntsan's gear list
Nikon Coolpix 5400 Panasonic Lumix DMC-FZ1000 Panasonic Lumix DMC-LX100 Panasonic Lumix DMC-GH4 Olympus M.Zuiko Digital ED 12mm 1:2 +2 more
Reply   Reply with quote   Complain
Anders W
Forum ProPosts: 19,066Gear list
Like?
Re: 5-axis IBIS in a P&S
In reply to ntsan, Aug 8, 2013

ntsan wrote:

Panasonic and Canon also have "5-axis" IS too on their DC, but their one is combining Lens IS + Electronic IS, where else Olympus one is full mechanical 5 axis IS

As it sounds, the Casio does what Panasonic and Canon have already done: Ordinary lens-based IS combined with bursts of underexposed shots at high shutter speeds which are then merged and aligned via in-body processing. The latter technique is of course available to MFT users as well, on top of what regular IBIS/OIS can do. It's only that the body won't do it for us. We have to shoot a burst and then merge/align in PP by means of software like LR/Enfuse or PhotoAcute.

One might add that correction for at least one of the five axes, roll, is in all likelihood electronic only in these compacts. To my knowledge, lens-based IS can't correct for roll. Possibly, correction for vertical and horizontal shift is purely electronic is well. While it is not impossible to correct for that via lens-based IS, I know of only one lens that does it (Canon 100/2.8L IS USM Macro).

 Anders W's gear list:Anders W's gear list
Panasonic Lumix DMC-G1 Olympus OM-D E-M5 Olympus E-M1 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G Vario 7-14mm F4 ASPH +21 more
Reply   Reply with quote   Complain
tt321
Senior MemberPosts: 4,150Gear list
Like?
Re: 5-axis IBIS in a P&S
In reply to Anders W, Aug 8, 2013

Anders W wrote:

ntsan wrote:

Panasonic and Canon also have "5-axis" IS too on their DC, but their one is combining Lens IS + Electronic IS, where else Olympus one is full mechanical 5 axis IS

As it sounds, the Casio does what Panasonic and Canon have already done: Ordinary lens-based IS combined with bursts of underexposed shots at high shutter speeds which are then merged and aligned via in-body processing. The latter technique is of course available to MFT users as well, on top of what regular IBIS/OIS can do. It's only that the body won't do it for us. We have to shoot a burst and then merge/align in PP by means of software like LR/Enfuse or PhotoAcute.

One might add that correction for at least one of the five axes, roll, is in all likelihood electronic only in these compacts. To my knowledge, lens-based IS can't correct for roll. Possibly, correction for vertical and horizontal shift is purely electronic is well. While it is not impossible to correct for that via lens-based IS, I know of only one lens that does it (Canon 100/2.8L IS USM Macro).

How does the E-M5/E-P5 roll correction work? Does it work differently for horizontally and vertically held shots?

Reply   Reply with quote   Complain
Anders W
Forum ProPosts: 19,066Gear list
Like?
Re: 5-axis IBIS in a P&S
In reply to tt321, Aug 8, 2013

tt321 wrote:

Anders W wrote:

ntsan wrote:

Panasonic and Canon also have "5-axis" IS too on their DC, but their one is combining Lens IS + Electronic IS, where else Olympus one is full mechanical 5 axis IS

As it sounds, the Casio does what Panasonic and Canon have already done: Ordinary lens-based IS combined with bursts of underexposed shots at high shutter speeds which are then merged and aligned via in-body processing. The latter technique is of course available to MFT users as well, on top of what regular IBIS/OIS can do. It's only that the body won't do it for us. We have to shoot a burst and then merge/align in PP by means of software like LR/Enfuse or PhotoAcute.

One might add that correction for at least one of the five axes, roll, is in all likelihood electronic only in these compacts. To my knowledge, lens-based IS can't correct for roll. Possibly, correction for vertical and horizontal shift is purely electronic is well. While it is not impossible to correct for that via lens-based IS, I know of only one lens that does it (Canon 100/2.8L IS USM Macro).

How does the E-M5/E-P5 roll correction work? Does it work differently for horizontally and vertically held shots?

It works the same way in horizontally and vertically held shots: By rotating the sensor around the optical axis. You can't rotate a lens and achieve the same effect. If you do, nothing happens to the image since the lens is radially symmetric.

The four other "axes" of the five-axis IBIS are accomplished by by only two sensor movements. Pitch requires the same up-down sensor movement as vertical shift and yaw the same left-right sensor movement as horizontal shift. So in this case, all you have to do in order to add vertical and horizontal shift to the two axes any IS system already has is a) add additional movement sensors and b) make sure the processing has information on focus distance since correction for vertical and horizontal shift requires that info.

 Anders W's gear list:Anders W's gear list
Panasonic Lumix DMC-G1 Olympus OM-D E-M5 Olympus E-M1 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G Vario 7-14mm F4 ASPH +21 more
Reply   Reply with quote   Complain
azazel1024
Senior MemberPosts: 1,102Gear list
Like?
Re: 5-axis IBIS in a P&S
In reply to tt321, Aug 8, 2013

One thing to keep in mind, the smaller the camera, the less optical image stabalization can help. This is, actually, to a degree a handicap for m4/3. Granted, I don't really know of any IBIS on larger formats...but larger formats COULD have better IBIS because of their size.

If the E-M5 has, (making something up) 5mm of sensor shift to compensate for shake, a FF sensor might be able to have more because of the larger sensor size combined with a bigger image circle to start with. So if you are trembling around and would be reaching the limits of the sensor shift, FF IBIS might be able to compensate further. Something like a P&S with its tiny sensor, likely isn't going to be able to shift much at all, limiting its utility.

Same with in lens, though in lens you could conceivable have the elements shift as much as you want if you design the thing however you want...though I am sure there is some kind of upward bound where you couldn't design the optical system to have the elements shift anymore and have even remotely a practical design.

-- hide signature --

Many things dealing with Olympus and their OM and Pen cameras, plus my general photography and musings http://omexperience.wordpress.com/

 azazel1024's gear list:azazel1024's gear list
Sigma 30mm F2.8 EX DN +10 more
Reply   Reply with quote   Complain
tt321
Senior MemberPosts: 4,150Gear list
Like?
Re: 5-axis IBIS in a P&S
In reply to Anders W, Aug 8, 2013

I see. But of course. Thanks.

Reply   Reply with quote   Complain
Anders W
Forum ProPosts: 19,066Gear list
Like?
Re: 5-axis IBIS in a P&S
In reply to azazel1024, Aug 8, 2013

azazel1024 wrote:

One thing to keep in mind, the smaller the camera, the less optical image stabalization can help. This is, actually, to a degree a handicap for m4/3. Granted, I don't really know of any IBIS on larger formats...but larger formats COULD have better IBIS because of their size.

If the E-M5 has, (making something up) 5mm of sensor shift to compensate for shake, a FF sensor might be able to have more because of the larger sensor size combined with a bigger image circle to start with. So if you are trembling around and would be reaching the limits of the sensor shift, FF IBIS might be able to compensate further. Something like a P&S with its tiny sensor, likely isn't going to be able to shift much at all, limiting its utility.

Same with in lens, though in lens you could conceivable have the elements shift as much as you want if you design the thing however you want...though I am sure there is some kind of upward bound where you couldn't design the optical system to have the elements shift anymore and have even remotely a practical design.

-- hide signature --

Many things dealing with Olympus and their OM and Pen cameras, plus my general photography and musings http://omexperience.wordpress.com/

On the contrary, smaller formats are if anything at an advantage when it comes to IS. If an MFT sensor has to move 2 mm in an instant to make up for the shake, an FF sensor has to move 4 mm during the same instant (twice as fast) to make up for the same shake. On top of that, the components that have to move, whether glass or the sensor, are bigger and heavier for FF.

 Anders W's gear list:Anders W's gear list
Panasonic Lumix DMC-G1 Olympus OM-D E-M5 Olympus E-M1 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G Vario 7-14mm F4 ASPH +21 more
Reply   Reply with quote   Complain
tt321
Senior MemberPosts: 4,150Gear list
Like?
Re: 5-axis IBIS in a P&S
In reply to Anders W, Aug 8, 2013

Anders W wrote:

tt321 wrote:

Anders W wrote:

ntsan wrote:

Panasonic and Canon also have "5-axis" IS too on their DC, but their one is combining Lens IS + Electronic IS, where else Olympus one is full mechanical 5 axis IS

As it sounds, the Casio does what Panasonic and Canon have already done: Ordinary lens-based IS combined with bursts of underexposed shots at high shutter speeds which are then merged and aligned via in-body processing. The latter technique is of course available to MFT users as well, on top of what regular IBIS/OIS can do. It's only that the body won't do it for us. We have to shoot a burst and then merge/align in PP by means of software like LR/Enfuse or PhotoAcute.

One might add that correction for at least one of the five axes, roll, is in all likelihood electronic only in these compacts. To my knowledge, lens-based IS can't correct for roll. Possibly, correction for vertical and horizontal shift is purely electronic is well. While it is not impossible to correct for that via lens-based IS, I know of only one lens that does it (Canon 100/2.8L IS USM Macro).

How does the E-M5/E-P5 roll correction work? Does it work differently for horizontally and vertically held shots?

It works the same way in horizontally and vertically held shots: By rotating the sensor around the optical axis. You can't rotate a lens and achieve the same effect. If you do, nothing happens to the image since the lens is radially symmetric.

The four other "axes" of the five-axis IBIS are accomplished by by only two sensor movements. Pitch requires the same up-down sensor movement as vertical shift and yaw the same left-right sensor movement as horizontal shift. So in this case, all you have to do in order to add vertical and horizontal shift to the two axes any IS system already has is a) add additional movement sensors and b) make sure the processing has information on focus distance since correction for vertical and horizontal shift requires that info.

By the way, what does this rotational compensation deal with any movements of the rotational axis? The camera holder is not obliged to rotate around the centre of the sensor, or even around a point within the sensor's boundaries. Is this axis position shift dealt with by the shifting parts of the IS?

Reply   Reply with quote   Complain
Anders W
Forum ProPosts: 19,066Gear list
Like?
Re: 5-axis IBIS in a P&S
In reply to tt321, Aug 8, 2013

tt321 wrote:

Anders W wrote:

tt321 wrote:

Anders W wrote:

ntsan wrote:

Panasonic and Canon also have "5-axis" IS too on their DC, but their one is combining Lens IS + Electronic IS, where else Olympus one is full mechanical 5 axis IS

As it sounds, the Casio does what Panasonic and Canon have already done: Ordinary lens-based IS combined with bursts of underexposed shots at high shutter speeds which are then merged and aligned via in-body processing. The latter technique is of course available to MFT users as well, on top of what regular IBIS/OIS can do. It's only that the body won't do it for us. We have to shoot a burst and then merge/align in PP by means of software like LR/Enfuse or PhotoAcute.

One might add that correction for at least one of the five axes, roll, is in all likelihood electronic only in these compacts. To my knowledge, lens-based IS can't correct for roll. Possibly, correction for vertical and horizontal shift is purely electronic is well. While it is not impossible to correct for that via lens-based IS, I know of only one lens that does it (Canon 100/2.8L IS USM Macro).

How does the E-M5/E-P5 roll correction work? Does it work differently for horizontally and vertically held shots?

It works the same way in horizontally and vertically held shots: By rotating the sensor around the optical axis. You can't rotate a lens and achieve the same effect. If you do, nothing happens to the image since the lens is radially symmetric.

The four other "axes" of the five-axis IBIS are accomplished by by only two sensor movements. Pitch requires the same up-down sensor movement as vertical shift and yaw the same left-right sensor movement as horizontal shift. So in this case, all you have to do in order to add vertical and horizontal shift to the two axes any IS system already has is a) add additional movement sensors and b) make sure the processing has information on focus distance since correction for vertical and horizontal shift requires that info.

By the way, what does this rotational compensation deal with any movements of the rotational axis? The camera holder is not obliged to rotate around the centre of the sensor, or even around a point within the sensor's boundaries. Is this axis position shift dealt with by the shifting parts of the IS?

Yes. Unless my geometric intuition fails me, the three kinds of movement of which the sensor is capable (up-down, left-right, and rotation about the optical axis) should jointly be able to cope with the scenario you describe.

 Anders W's gear list:Anders W's gear list
Panasonic Lumix DMC-G1 Olympus OM-D E-M5 Olympus E-M1 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G Vario 7-14mm F4 ASPH +21 more
Reply   Reply with quote   Complain
Guy Parsons
Forum ProPosts: 19,488Gear list
Like?
Rock and roll sensor......
In reply to tt321, Aug 8, 2013

Some videos may help....http://www.engadget.com/2012/02/08/olympus-om-d-e-m5-five-axis-sensor-shift-image-stabilization-han/ but not working for me at the moment.

Also..... http://www.youtube.com/watch?v=TVNB3dayL-A

There is quite distinctly an X-Y shift and Rotation movement, so combining those will cope with a mess of shake problems.

The Casio method is a cheaty method involving using multiple frames at 40 fps and combining them. They use the same combining for low light low ISO improvements and also long tele shots in dim light to reduce shake. Nice camera though, my wife uses the earlier ZR200 and the trickier ZR1000.

As an aside.... Casio have a neat custom setup method that is worth copying by Olympus, used to allow 999 custom setups, latest models now only 12 custom setups. Setups are easily downloaded to PC and restored or translated using third party software (or sometimes just by filename changes) to another Casio body model.

Regards...... Guy

 Guy Parsons's gear list:Guy Parsons's gear list
Olympus PEN E-PL5 Olympus PEN E-P5 Panasonic Lumix G Vario 45-150mm F4-5.6 ASPH Mega OIS Olympus 12-40mm F2.8 Panasonic Lumix DMC-LX3 +17 more
Reply   Reply with quote   Complain
boxerman
Contributing MemberPosts: 708Gear list
Like?
Re: 5-axis IBIS in a P&S
In reply to tt321, Aug 9, 2013

By the way, what does this rotational compensation deal with any movements of the rotational axis? The camera holder is not obliged to rotate around the centre of the sensor, or even around a point within the sensor's boundaries. Is this axis position shift dealt with by the shifting parts of the IS?

It may help to know that there is a mathematical theorem that rotation (of a plane) about any axis can be accomplished by rotation about any OTHER axis, plus a translation. E.g., you can always rotate about the sensor's center and add a shift to cover rotation about non-sensor axis.

You probably didn't mean it, but "movements of the rotational axis" suggests actual movement of the axis, rather than rotation about an axis that is not aligned with the sensor center. I doubt it is necessary to do this, and it would take a series of sensor readings, not just instantaneous ones, to "predict" the movement of the axis. That's "second order correction." Again, I think dubious as implemented, and dubious it'd get you noticeably better correction.

-- hide signature --

The BoxerMan

 boxerman's gear list:boxerman's gear list
Olympus OM-D E-M5 Olympus M.Zuiko Digital ED 9-18mm 1:4.0-5.6 Panasonic Lumix G 20mm F1.7 ASPH Olympus M.Zuiko Digital ED 75-300mm 1:4.8-6.7 Olympus M.Zuiko Digital ED 40-150mm 1:4-5.6 R +1 more
Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads