Latest Pentax patent for a "Tilting Sensor"

halimj7 · 11-10-2017, 04:08 AM

Originally posted by MJKoski

This would allow rear tilt with a DSLR camera. Like I can do with a proper monorail or view camera with my 4x5" setup. Very very useful feature - one can create wide angle type near-far relations with longer lenses than wide angles.

Like so: Large Format Techniques

Check the bottom of the page. Absolutely killer feature for landscapers and other suitable cases.

Bismillah

This is exactly what Ricoh needs to do to distinguish themselves from the market. Hopefully we see this and other great ideas implemented into the K3II replacement beginning next year.

pentasonic49 · 11-12-2017, 01:23 PM

Wouldn't the tilt give very uneven lighting from side to side due to the limited angle of acceptance of the micro lenses in the sensor array? Or would firmware linked to the sensor angle take care of the differences?

**Digitalis** · 11-12-2017, 10:10 PM

Originally posted by pentasonic49

Wouldn't the tilt give very uneven lighting from side to side due to the limited angle of acceptance of the micro lenses in the sensor array?

This would only be problematic if you are asymmetrically tilting the sensor*. Symmetrical movements shouldn't provoke inhomogeneities in the vignetting characteristics of the lens. Shift movements would cause problems depending upon the lens and the proximity of the sensor relative to the extremities of its image circle. Wide angle lenses being the worst offenders in this regard.

*with the TS mechanism being discussed I don't think it would be possible to employ such movements.

stevebrot · 11-13-2017, 02:28 PM

Originally posted by Digitalis

*with the TS mechanism being discussed I don't think it would be possible to employ such movements.

I did the unthinkable a few days back and actually read the patent application. It has some interesting fine points. One of which indicates preferred integration with a tilting lens component. That sounds a lot more feasible than working with rear tilt alone.

Steve

photoptimist · 11-14-2017, 12:30 PM

Originally posted by Digitalis

Well using PDAF is out of the question with sensor based tilt, we would only be able to see the effect through live view - in which case CDAF or on-sensor PDAF will be in use. I seem to recall PDAF sensors having problems with detecting phase differential when light rays strike them at sharp angles of incidence.

Exactly! Unless Pentax replicates the sensor-tilter on the focus screen, too, the only way to see the effects of sensor tilt is in live view.

On-sensor PDAF might still work if the sensor tilt is fairly small angle. A bit of math shows that even a 2-5° tilt angle of the sensor delivers a huge amount of tilt-effect to the image without creating excessively larger incidence angles or too much light fall-off across the sensor.

**Digitalis** · 11-14-2017, 04:50 PM

Originally posted by photoptimist

A bit of math shows that even a 2-5° tilt angle of the sensor delivers a huge amount of tilt-effect to the image without creating excessively larger incidence angles or too much light fall-off across the sensor.

This of course is heavily dependent upon the lens being used.

photoptimist · 11-14-2017, 05:18 PM

Originally posted by Digitalis

This of course is heavily dependent upon the lens being used.

Good point! At a 5° tilt, on-sensor PDAF would suffer back-out in the center at f/11. Off-axis, the on-sensor PDAF black-out would be a function of the exit pupil distance used when designing the sensor versus the actual exit pupil distance of the specific lens.

GrantFS · 11-28-2017, 01:20 PM - **2 Likes**

Rear tilt isn't exactly the same as front tilt.

Yes you will get Scheimpflug (converging planes) but it also alters the 'geometry' of the image. Think of shining a torch against a piece of card that you are holding vertically. The card represents the film-plane (image plane). The torch is like the light (image) coming through the lens and is parallel to the card, so you get a perfect circle.

If you tilt the card, then the circle becomes distorted. So this will alter the focus plane, but in order to get back to having a 'round' circle of light, you would need to tilt the entire camera, on its tripod, so that it would again be parallel. Once you've done that, then you're no longer looking at (pointing the camera at) the same thing. So you need a shift on the front to compensate. So you're pretty much back to square one.

So, yes rear tilt gives you scheimpflug, but it's not as simple, nor the same as, front tilt.

Phew!

stevebrot · 11-28-2017, 02:19 PM - **1 Like**

Originally posted by GrantFS

Rear tilt isn't exactly the same as front tilt.

Yes you will get Scheimpflug (converging planes) but it also alters the 'geometry' of the image.

Agreed, and one reason why I posted the below earlier in the thread...

Originally posted by stevebrot

I did the unthinkable a few days back and actually read the patent application. It has some interesting fine points. One of which indicates preferred integration with a tilting lens component. That sounds a lot more feasible than working with rear tilt alone.

The rule of thumb is to adjust plane of focus at the front and geometry at the back. The patent application hints that something more than just a tilting sensor is intended to allow for a proper solution. If so, something very exciting might be coming.

Steve

MJKoski · 11-28-2017, 04:45 PM

Why not tilt the mount as well? That would give front tilt all lenses.

rawr · 11-28-2017, 05:12 PM

Originally posted by MJKoski

Why not tilt the mount as well?

And the tripod mount.

Why not add a motorized three-axis gimbal there for good measure, turning your Pentax into a DJI Osmo.

**Digitalis** · 11-28-2017, 07:06 PM

Originally posted by MJKoski

Why not tilt the mount as well? That would give front tilt all lenses.

But that would throw the weather sealing out the window.

GrantFS · 11-29-2017, 12:10 AM

Originally posted by MJKoski

Why not tilt the mount as well? That would give front tilt all lenses.

When you have two moving parts, it's extremely difficult to get them perfectly aligned. The closer the lens-plane and film-plane, the more significant any movement is, but beyond that, in order to be able to usefully tilt or shift any lens, that lens needs to have a large enough image circle to cover any movement. Lenses for 35mm style cameras (of any size) are not made with a large image circle.

photoptimist · 11-29-2017, 08:45 AM - **1 Like**

Originally posted by MJKoski

Why not tilt the mount as well? That would give front tilt all lenses.

Tricky! (But not impossible)

The key is to ensure that the center of rotation of the mount is at the focal plane. That will tilt the lens without shifting the image circle of the lens off the sensor.

Tilting the lens around any other center of rotation requires lenses with larger image circles (e.g., a medium format lens on an FF body or FF lens on an APS-C body).

MJKoski · 11-29-2017, 11:10 AM

Yes, image circles are always something to consider. But living without movements feels weird after using 4x5" setup for the whole summer.

11-12-2017, 10:10 PM	#48
Digitalis Site Supporter Join Date: Mar 2009 Location: Melbourne, Victoria Photos: Gallery Posts: 11,694	Originally posted by pentasonic49 Wouldn't the tilt give very uneven lighting from side to side due to the limited angle of acceptance of the micro lenses in the sensor array? This would only be problematic if you are asymmetrically tilting the sensor. Symmetrical movements shouldn't provoke inhomogeneities in the vignetting characteristics of the lens. Shift movements would cause problems depending upon the lens and the proximity of the sensor relative to the extremities of its image circle. Wide angle lenses being the worst offenders in this regard. with the TS mechanism being discussed I don't think it would be possible to employ such movements. Last edited by Digitalis; 11-12-2017 at 10:18 PM.

11-28-2017, 01:20 PM - 2 Likes	#53
GrantFS New Member Join Date: Jul 2016 Posts: 24	Rear tilt isn't exactly the same as front tilt. Yes you will get Scheimpflug (converging planes) but it also alters the 'geometry' of the image. Think of shining a torch against a piece of card that you are holding vertically. The card represents the film-plane (image plane). The torch is like the light (image) coming through the lens and is parallel to the card, so you get a perfect circle. If you tilt the card, then the circle becomes distorted. So this will alter the focus plane, but in order to get back to having a 'round' circle of light, you would need to tilt the entire camera, on its tripod, so that it would again be parallel. Once you've done that, then you're no longer looking at (pointing the camera at) the same thing. So you need a shift on the front to compensate. So you're pretty much back to square one. So, yes rear tilt gives you scheimpflug, but it's not as simple, nor the same as, front tilt. Phew! Last edited by GrantFS; 11-28-2017 at 01:38 PM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
HD Pentax-DA 645 90-180mm or 80-160mm f/4.5ED AW SR lens patent filed in Japan	sculptor666	Pentax News and Rumors	28	08-16-2018 01:09 PM
Ricoh files a patent for a Pentax 35-80mm f/4 full frame lens	OoKU	Pentax News and Rumors	147	07-04-2016 11:15 AM
Ricoh patent: avoiding movie AF-related image blur (?)	Kunzite	Pentax News and Rumors	3	03-02-2016 10:00 AM
A New 645 next yr.? Ricoh sensor patent	22codfish	Pentax News and Rumors	10	02-07-2016 02:59 PM
Ricoh patent: Sensor AF using subpixel pattern	JPT	Pentax News and Rumors	7	12-30-2015 10:53 AM

11-10-2017, 04:08 AM	#46
halimj7 Forum Member Join Date: Oct 2017 Posts: 76	Originally posted by MJKoski This would allow rear tilt with a DSLR camera. Like I can do with a proper monorail or view camera with my 4x5" setup. Very very useful feature - one can create wide angle type near-far relations with longer lenses than wide angles. Like so: Large Format Techniques Check the bottom of the page. Absolutely killer feature for landscapers and other suitable cases. Bismillah This is exactly what Ricoh needs to do to distinguish themselves from the market. Hopefully we see this and other great ideas implemented into the K3II replacement beginning next year.

11-12-2017, 01:23 PM	#47
pentasonic49 Veteran Member Join Date: Oct 2012 Location: weston-super-mare Posts: 395	Wouldn't the tilt give very uneven lighting from side to side due to the limited angle of acceptance of the micro lenses in the sensor array? Or would firmware linked to the sensor angle take care of the differences?

11-13-2017, 02:28 PM	#49
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by Digitalis *with the TS mechanism being discussed I don't think it would be possible to employ such movements. I did the unthinkable a few days back and actually read the patent application. It has some interesting fine points. One of which indicates preferred integration with a tilting lens component. That sounds a lot more feasible than working with rear tilt alone. Steve

11-14-2017, 12:30 PM	#50
photoptimist Pentaxian Join Date: Jul 2016 Photos: Albums Posts: 5,122	Originally posted by Digitalis Well using PDAF is out of the question with sensor based tilt, we would only be able to see the effect through live view - in which case CDAF or on-sensor PDAF will be in use. I seem to recall PDAF sensors having problems with detecting phase differential when light rays strike them at sharp angles of incidence. Exactly! Unless Pentax replicates the sensor-tilter on the focus screen, too, the only way to see the effects of sensor tilt is in live view. On-sensor PDAF might still work if the sensor tilt is fairly small angle. A bit of math shows that even a 2-5° tilt angle of the sensor delivers a huge amount of tilt-effect to the image without creating excessively larger incidence angles or too much light fall-off across the sensor.

11-14-2017, 04:50 PM	#51
Digitalis Site Supporter Join Date: Mar 2009 Location: Melbourne, Victoria Photos: Gallery Posts: 11,694	Originally posted by photoptimist A bit of math shows that even a 2-5° tilt angle of the sensor delivers a huge amount of tilt-effect to the image without creating excessively larger incidence angles or too much light fall-off across the sensor. This of course is heavily dependent upon the lens being used.

11-14-2017, 05:18 PM	#52
photoptimist Pentaxian Join Date: Jul 2016 Photos: Albums Posts: 5,122	Originally posted by Digitalis This of course is heavily dependent upon the lens being used. Good point! At a 5° tilt, on-sensor PDAF would suffer back-out in the center at f/11. Off-axis, the on-sensor PDAF black-out would be a function of the exit pupil distance used when designing the sensor versus the actual exit pupil distance of the specific lens.

11-28-2017, 02:19 PM - 1 Like	#54
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by GrantFS Rear tilt isn't exactly the same as front tilt. Yes you will get Scheimpflug (converging planes) but it also alters the 'geometry' of the image. Agreed, and one reason why I posted the below earlier in the thread... Originally posted by stevebrot I did the unthinkable a few days back and actually read the patent application. It has some interesting fine points. One of which indicates preferred integration with a tilting lens component. That sounds a lot more feasible than working with rear tilt alone. The rule of thumb is to adjust plane of focus at the front and geometry at the back. The patent application hints that something more than just a tilting sensor is intended to allow for a proper solution. If so, something very exciting might be coming. Steve

11-28-2017, 04:45 PM	#55
MJKoski Veteran Member Join Date: Aug 2016 Posts: 1,784	Why not tilt the mount as well? That would give front tilt all lenses.

11-28-2017, 05:12 PM	#56
rawr Veteran Member Join Date: Jun 2009 Posts: 11,913	Originally posted by MJKoski Why not tilt the mount as well? And the tripod mount. Why not add a motorized three-axis gimbal there for good measure, turning your Pentax into a DJI Osmo.

11-28-2017, 07:06 PM	#57
Digitalis Site Supporter Join Date: Mar 2009 Location: Melbourne, Victoria Photos: Gallery Posts: 11,694	Originally posted by MJKoski Why not tilt the mount as well? That would give front tilt all lenses. But that would throw the weather sealing out the window.

11-29-2017, 12:10 AM	#58
GrantFS New Member Join Date: Jul 2016 Posts: 24	Originally posted by MJKoski Why not tilt the mount as well? That would give front tilt all lenses. When you have two moving parts, it's extremely difficult to get them perfectly aligned. The closer the lens-plane and film-plane, the more significant any movement is, but beyond that, in order to be able to usefully tilt or shift any lens, that lens needs to have a large enough image circle to cover any movement. Lenses for 35mm style cameras (of any size) are not made with a large image circle.

11-29-2017, 08:45 AM - 1 Like	#59
photoptimist Pentaxian Join Date: Jul 2016 Photos: Albums Posts: 5,122	Originally posted by MJKoski Why not tilt the mount as well? That would give front tilt all lenses. Tricky! (But not impossible) The key is to ensure that the center of rotation of the mount is at the focal plane. That will tilt the lens without shifting the image circle of the lens off the sensor. Tilting the lens around any other center of rotation requires lenses with larger image circles (e.g., a medium format lens on an FF body or FF lens on an APS-C body).

11-29-2017, 11:10 AM	#60
MJKoski Veteran Member Join Date: Aug 2016 Posts: 1,784	Yes, image circles are always something to consider. But living without movements feels weird after using 4x5" setup for the whole summer.