Olympus 5-axis IBIS for tilt-skew-shift photography

[Fri13]

I have been wondering that when does Olympus release a camera/firmware (possible) that allows user to control IBIS axis?

It would allow very interesting photography when each sensor axis could be changed and so on the focus would get altered as well.

The IBIS would then work against its purpose, that is to keep sensor still to subject time that is required to get it sharp.

But in this case user would press a Fn-button and enable TSS mode that disables IBIS but gives user possibility to rotate and move sensor in each of its 4 axis (roll would not be needed).

We would get a ultimate miniature technical camera for interesting new possibilities. The sensor being smaller, even a small degree translates to larger effect. We would get every objective mounted to camera work as tilt-shift objective.

Olympus OM-D: See the amazing 5-axis image stabil…: http://youtu.be/pxgnE6xY0OA

The sensor after all moves quite a lot. But I didn't see rotation in that video, only 2-plane movements and rotation. This to work the sensor would be needed to actually rotate on X and Y axis, as I have came to believe as how otherwise all 5-axis is corrected, rotations included instead just shifts?

 

[Rudy]

The sensor always stays in the focal plane. There is no tilting ever as this would also tilt the object plane that is in focus.
5 axis refers to the movements of the camera that can be sensed and corrected for, i.e. when you tilt the camera it senses that and calculates the image offset the tilt would generate depending on the focal length of the lens and then shifts the sensor by that amount.
Rudy
p.s.I agree that having a camera where the sensor plane yaw and pitch could be controlled would be awesome, but I would not hold my breath to see this in a consumer camera anytime soon.

 

[Fri13]

The sensor always stays in the focal plane. There is no tilting ever as this would also tilt the object plane that is in focus.
5 axis refers to the movements of the camera that can be sensed and corrected for, i.e. when you tilt the camera it senses that and calculates the image offset the tilt would generate depending on the focal length of the lens and then shifts the sensor by that amount.
Rudy
p.s.I agree that having a camera where the sensor plane yaw and pitch could be controlled would be awesome, but I would not hold my breath to see this in a consumer camera anytime soon.

I am interested how does Olympus correct the yaw and pitch changes unless tilting the sensor? As three axis is required so we can correct focal plane movement (side to side, up to down movement and roll, or said otherwise the X, Z and Y axis) but we can't correct pitch or yaw change (rotate left/right/up/down) as focal plane direction does get changed, that is the two other axis (old tech), together five.

http://static.squarespace.com/static/531df49be4b0467fe7e839de/t/53232317e4b0182e183c2cd5/1394811672533/5-axis IBIS.jpg

Like if we look the old Olympus E-series cameras that had IBIS, like E-520. The sensor rotates (doesn't move) in X and Y axis. Meaning sensor center stay in axis origin all the time, while with new 5-axis IBIS it can be moved around it freely, and tilted.

As if camera pitch is changes, like recomposition after focusing then the focus plane has already changed. To correct that, we need to rotate sensor back to previous position to counter the focus plane change.

Only axis Olympus has not (yet?) added is longitudinal movement, what would allow with accurate enough inertia sensors to even keep plane of focus still when camera moves forward or backward. It still couldn't be used for focus stacking because magnification changes, so no joy for macro photographers.

 

[Rudy]

The sensor in the Olympus camera cannot tilt. It only moves in plane. However the camera can detect tilt (yaw, pitch and roll). The first order effect of yaw and pitch is a shift of the image (that is the property of a lens, it changes angles into displacement).
Tilting the sensor would tilt the object plane that is in focus it would not correct for a tilt of the camera. Look up Scheimpflug if you are not familiar with this.
Rudy

 

[MajorMagee]

The idea of sensor based Tilt/Shift sounds patent worthy...

 

[eteless]

It's probably already patented as the Pentax K-7/K-5 did it a long time ago, not that anyone really noticed...

 

[Fri13]

It's probably already patented as the Pentax K-7/K-5 did it a long time ago, not that anyone really noticed...

It is true many doesn't know that, I didn't. But I had this idea since first IBIS camera came out from Olympus.

Turn any Pentax lens into a shift lens: Compositi…: http://youtu.be/RKbnO8rlg44

And I am not the only one, clearly it is first idea for engineers when they start designing IBIS.

 

[b_rubenstein]

Camera shake is primarily composed of low displacement movement, so the amplitude of the sensor correction movement is very small, and probably too small for useable tilt shift effects. Not to mention that maybe .001% of camera buyers would be interested in this feature.

 

[fortwodriver]

The sensor in the Olympus camera cannot tilt. It only moves in plane. However the camera can detect tilt (yaw, pitch and roll). The first order effect of yaw and pitch is a shift of the image (that is the property of a lens, it changes angles into displacement).

Rudy

You can actually see this in the viewfinder of the E-5 and E-M1 after a photo is taken. The image will sometimes be slightly shifted after being taken as the camera corrects the sensor displacement.

 

[Fri13]

Camera shake is primarily composed of low displacement movement, so the amplitude of the sensor correction movement is very small, and probably too small for useable tilt shift effects. Not to mention that maybe .001% of camera buyers would be interested in this feature.

As on Olympus presentation, the sensor can shift about 1/3 of the sensor. It isn't "very small". But the rotation is small, but at those distances it isn't required to be huge as sensor is so small. Smaller sensor requires smaller angle.

 

[Fri13]

Okay... New rumor. That sounds again little weird by timing... And why always just one/few features instead all?


“Keystoning will allow you to fix the distortion you see in architectural images. So buildings won’t “lean”…but the cool thing is you fix it in LIVE VIEW! You can see it as you fix it! Who needs a tilt shift lens? Olympus innovation is incredible.”


http://www.43rumors.com/ft4-e-m1-firmware-update-will-add-keystoning-feature/

But I don't believe to that as so much the sensor can't tilt. As to get correction the sensor would be required to be parallel with the building, so if I tilt camera up to get photo from tall building by 15 degrees, sensor should tilt 15 degrees. And 15 degree is little.
With software we can crop view based level sensor, but quality will suffer.

 

[Ross the fiddler]

Okay... New rumor. That sounds again little weird by timing... And why always just one/few features instead all?


“Keystoning will allow you to fix the distortion you see in architectural images. So buildings won’t “lean”…but the cool thing is you fix it in LIVE VIEW! You can see it as you fix it! Who needs a tilt shift lens? Olympus innovation is incredible.”


http://www.43rumors.com/ft4-e-m1-firmware-update-will-add-keystoning-feature/

But I don't believe to that as so much the sensor can't tilt. As to get correction the sensor would be required to be parallel with the building, so if I tilt camera up to get photo from tall building by 15 degrees, sensor should tilt 15 degrees. And 15 degree is little.
With software we can crop view based level sensor, but quality will suffer.

Whether you like it or not, that is the way it is done in software & if this feature comes in the new firmware, it would be emulating a software procedure too. It will still be useful, even for RAW shooters, as it would allow the user to see the corrected, cropped end result that can be basically repeated in other RAW converter software as well, like Capture One (Pro 7).