Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
12-16-2018, 07:53 AM - 1 Like   #1
Junior Member




Join Date: Dec 2018
Location: Ghent, East-Flanders
Posts: 26
Addition to composition adjustment: Scheimpflug with tilt and yaw

Hi everyone,

I am a very happy owner of a K1 and have some ideas about the possibilities of the sensor shake reduction.

Is anybody interested in being able to manually tilt and yaw the sensor, just like with compostion adjustments does with shifting the sensor up/down and left/right?
Like with a technical camera, one could change the tilt and/or yaw angle of the sensor to change the direction of the focus plane. The rules of Scheimpflug could be used this way to be able to use a less high aperture value for landscape and other styles of imaging, to prevent diffraction from using high aperture values.

The way to set these changes could be done in different ways and could be done as follows (of course via LV):

- the camera ask you to choose the first selection point for measuring sharpness, then the second point (camera each time shows a rather small square shape that can be moved via up/down and left/right buttons and hit OK)
- after confirming both points, the camera show both points at the same time at 100% via a split screen
- choose if you want to adjust the tilt and the yaw independently via the front and back dials or...
- choose if you want to be able to set the amount of inclination of the sensor and the direction between 0-360° with the same dials
- set the amout with one or the other method until both 100% zoom are sharp(er), confirm sensor setting with OK
- take image with better back to front sharpness with more open diaphragm values

Of course this method will have the technical limits of what the sensor is capable of tilting/yawing in all directions, but the same applies to composition adjustments and people are happy that this function is already present.

The first method (if you choose tilt and yaw independently is the best choise if you want to have the camera leveled to the horizon and want more sharpness from the foreground to the background or if you want a building for example from left up close to right in the distance and want it to be sharp all the way without needing an aperture of for example 16-22 or higher to get everything sharp enough. It will prevent loss of sharpness due to diffraction.

The second method by setting the amount and turning the angle between 0°+ and -360° is more handy for cases where the focal plane is not in the direction of 90° down/up or left/right, but has a different angle in between. In that case we could be able to change the amount and the angle of tilt/yaw at the same time with both dials while looking at the two 100% zoom split screens that you want to become sharp(er).

If anyone is interested, spread the word.
I hope my explanation is understandable because english isn't my native language.

Have a good day and happy shooting.

12-19-2018, 08:14 AM - 1 Like   #2
Pentaxian
photoptimist's Avatar

Join Date: Jul 2016
Photos: Albums
Posts: 3,756
It's a really interesting idea. Two additional modes might be:

1) Pick any three points in the frame and the camera auto-focuses on all three (if possible).

2) Auto-full-frame focus: the camera adjusts both tilt and yaw to maximize the percentage of the frame in focus.


Alas, all of these features would require a completely new sensor shake reduction mechanism. Although the current sensor shake reduction mechanism does compensate for the image shaking effects of tilting and yawing the camera, it does not do it by tilting or yawing the sensor. With the current SR mechanism, the sensor always remains in the same plane.
12-19-2018, 09:10 AM - 2 Likes   #3
Otis Memorial Pentaxian
Loyal Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 34,474
QuoteOriginally posted by Svenneke Quote
Is anybody interested in being able to manually tilt and yaw the sensor,
Last year, Ricoh filed a patent for sensor tilt technology that generated a fair amount of discussion here at Pentax Forums.

Latest Pentax patent for a "Tilting Sensor" - PentaxForums.com

QuoteOriginally posted by photoptimist Quote
Two additional modes might be:

1) Pick any three points in the frame and the camera auto-focuses on all three (if possible).

2) Auto-full-frame focus: the camera adjusts both tilt and yaw to maximize the percentage of the frame in focus.

One of the thoughts on that thread was integration with a touch screen to allow automation* for user-directed near-far focus. Of course, having some sort of tilt for the lens axis is helpful as well (hinted at within the patent application). FWIW, the amount of tilt would likely be quite meager.


Steve

* Manual Scheimpflug is difficult enough with a proper view camera, much less on a small screen having exaggerated DOF.

Last edited by stevebrot; 12-19-2018 at 02:57 PM.
12-19-2018, 03:24 PM   #4
Junior Member




Join Date: Dec 2018
Location: Ghent, East-Flanders
Posts: 26
Original Poster
QuoteOriginally posted by photoptimist Quote
It's a really interesting idea. Two additional modes might be:

1) Pick any three points in the frame and the camera auto-focuses on all three (if possible).

2) Auto-full-frame focus: the camera adjusts both tilt and yaw to maximize the percentage of the frame in focus.


Alas, all of these features would require a completely new sensor shake reduction mechanism. Although the current sensor shake reduction mechanism does compensate for the image shaking effects of tilting and yawing the camera, it does not do it by tilting or yawing the sensor. With the current SR mechanism, the sensor always remains in the same plane.
Hi photoptimist,

Thank you for your thoughts and explanation.
I wonder how the current SR mechanism can compensate for tilt and yaw if the sensor remains in the same plane and the lenses are fixed because they have no image stabilisation?
Is the information on the pentax website not telling us the whole story?
PENTAX K-1
After looking at the illustration on that page, I was under the impression that the sensor did rotate up/down for tilt and left/right for yaw compensations? How do they achieve that without moving the sensor out of the same plane?
If the sensor was fixed in that same plane, would it not only allow for 3-axis stabilisations (up/down shift, left/right shift and roll)?

Thank you in advance for your explanation. I worked with technical cameras in the past so to me it seemed obvious that the sensor would move the same way as the back of such a camera.

Have a good day.
Greetings,
Sven

---------- Post added 12-19-18 at 03:33 PM ----------

QuoteOriginally posted by stevebrot Quote
Last year, Ricoh filed a patent for sensor tilt technology that generated a fair amount of discussion here at Pentax Forums.

Latest Pentax patent for a "Tilting Sensor" - PentaxForums.com




One of the thoughts on that thread was integration with a touch screen to allow automation* for user-directed near-far focus. Of course, having some sort of tilt for the lens axis is helpful as well (hinted at within the patent application). FWIW, the amount of tilt would likely be quite meager.


Steve

* Manual Scheimpflug is difficult enough with a proper view camera, much less on a small screen having exaggerated DOF.
Hi Steve,

Thank you for pointing me to that thread, I will certainly look into that.

Indeed the amount of rotation via tilt and yaw, if they would made it possible, would be pretty modest, but nonetheless, it could be very handy for those who know how Scheimpflug works. It is a very powerfull way to optimise your aperture settings and to improve certain photo's. Even for those who don't know how it works, that technique still would make sense if it could be automated to some extend by the camera like photoptimist also suggested.

Nice to read your feedback too Steve.
Have a great day.
Greetings,
Sven

---------- Post added 12-19-18 at 04:22 PM ----------

Hi photoptimist,

Thanks to Steve his link to the older thread, I found your explanation in post #33 how pentax achives the tilt/yaw compensations and that makes sense now to me.
So in reality, the sensor moves only up/down/left/right and rotates along its axis for roll compensations. If I understand it correct, the tilt and yaw is corrected for by moving the sensor the right amount in the same focus plane, parallel with the lenses.

I totally agree with your explanation in that post #33 about how little amount of rotation should only be needed according to the kind of lens you use. A wide angle lens would benefit much more than a tele lens, but still the last one could be used with a more open aperture than without the possibility of using Scheimpflug.
So it seems we really have to wait for the new patented construction that could allow us to use real tilt and yaw movements of the sensor. Great times are coming it seems...

Steve, thank you for the link, I didn't knew that pentax filed that patent and it is great to see them developing that kind of specialized techniques. It shows that pentax is a company that thinks differently and listens to what photographers want to use.

However, there are some limitations. If you only can rotate the sensor and not the lens also, there will be some small amount of perspective distortion visible because for example the plane of a building will no longer be parallel with the sensor plane. It should be possible to compensate for that by rotating the camera slightly by itself.

It also possible that with some wide angle lenses that slightly intrude inside the body, there will be a little bit more vignetting when the sensor is tilted to its maximum amount. But that can easily be corrected afterwards.

OTOH it seems very usefull for creating miniature landscape style images, for those who like that kind of photography.
Architecture photography also could benefit.
Also, using a wide aperture and focussing on two different objects or persons, one near and another somewhat further away, to isolate them from the rest is another great option. It could open a lot of possibilities.

Goodnight everybody.
Greetings,
Sven


Last edited by Svenneke; 12-19-2018 at 03:37 PM.
12-19-2018, 05:04 PM - 1 Like   #5
Pentaxian
photoptimist's Avatar

Join Date: Jul 2016
Photos: Albums
Posts: 3,756
QuoteOriginally posted by Svenneke Quote
Hi photoptimist,

Thank you for your thoughts and explanation.
I wonder how the current SR mechanism can compensate for tilt and yaw if the sensor remains in the same plane and the lenses are fixed because they have no image stabilisation?
Is the information on the pentax website not telling us the whole story?
PENTAX K-1
After looking at the illustration on that page, I was under the impression that the sensor did rotate up/down for tilt and left/right for yaw compensations? How do they achieve that without moving the sensor out of the same plane?
If the sensor was fixed in that same plane, would it not only allow for 3-axis stabilisations?

Thank you in advance for your explanation. I worked with technical cameras in the past so to me it seemed obvious that the sensor would move the same way as the back of such a camera.

Have a good day.
Greetings,
Sven
Imagine this scenario:

Imagine you are taking a picture of a fox in front of a brick way 20 meters away. You are holding a 300 mm f/2.8 lens at it's nodal point when you mash the shutter button of your K-1 and cause the body of the camera to drop 0.5 mm during the exposure relative to the nodal point and the nose of the lens to tilt upward. This was two effects on the imaging geometry:

1) The upward tilt of the lens-camera combination causes the original scene to appear to shift downward on the sensor by about 0.5 mm (a massive blur of about 100 pixels).
2) The sensor and plane of focus tilt by about 0.1° relative to their original locations.

The current SR system will shift the sensor by 0.5 mm in the opposite direction to shift the image of the scene back to it's original location are prevent the 100 pixel blur. But it does not correct for the 0.1° tilt in the plane of focus.

But what are the effects of the uncorrected angular tilt and are they serious? There are two uncorrected blurring effects:

1) blur from defocusing caused by the tilt in the plane: At 20 meters, the FoV is about 1.6 m high by 2.4 m wide. That 1.6 x 2.4 meter frame of focus gets tilted with the lower edge shifting about 1.3 mm back and the top of the frame of shifting about 1.3 mm toward the camera. But the total DoF for a 300mm f/2.8 lens at 10 meters on a K-1 is 180 mm. Thus, the effect is negligible.

2) blur from change in magnification at the top and bottom: With the tilt, the bottom of the frame is now slightly closer optically than it was. That change in magnification is about 0.007% which would cause about 1/4 pixel of blur in the extreme corners. Thus, the effect is almost negligible especially compared to the 100 pixel blur that was removed.

You can run other scenarios but it's fair to say that correcting the image shift caused by a tilt or yaw motion really removes more than 99% of the blurring effects of tilt and yaw.
12-20-2018, 01:39 AM   #6
Junior Member




Join Date: Dec 2018
Location: Ghent, East-Flanders
Posts: 26
Original Poster
QuoteOriginally posted by photoptimist Quote
Imagine this scenario:

Imagine you are taking a picture of a fox in front of a brick way 20 meters away. You are holding a 300 mm f/2.8 lens at it's nodal point when you mash the shutter button of your K-1 and cause the body of the camera to drop 0.5 mm during the exposure relative to the nodal point and the nose of the lens to tilt upward. This was two effects on the imaging geometry:

1) The upward tilt of the lens-camera combination causes the original scene to appear to shift downward on the sensor by about 0.5 mm (a massive blur of about 100 pixels).
2) The sensor and plane of focus tilt by about 0.1° relative to their original locations.

The current SR system will shift the sensor by 0.5 mm in the opposite direction to shift the image of the scene back to it's original location are prevent the 100 pixel blur. But it does not correct for the 0.1° tilt in the plane of focus.

But what are the effects of the uncorrected angular tilt and are they serious? There are two uncorrected blurring effects:

1) blur from defocusing caused by the tilt in the plane: At 20 meters, the FoV is about 1.6 m high by 2.4 m wide. That 1.6 x 2.4 meter frame of focus gets tilted with the lower edge shifting about 1.3 mm back and the top of the frame of shifting about 1.3 mm toward the camera. But the total DoF for a 300mm f/2.8 lens at 10 meters on a K-1 is 180 mm. Thus, the effect is negligible.

2) blur from change in magnification at the top and bottom: With the tilt, the bottom of the frame is now slightly closer optically than it was. That change in magnification is about 0.007% which would cause about 1/4 pixel of blur in the extreme corners. Thus, the effect is almost negligible especially compared to the 100 pixel blur that was removed.

You can run other scenarios but it's fair to say that correcting the image shift caused by a tilt or yaw motion really removes more than 99% of the blurring effects of tilt and yaw.
Good morning photoptimist,

That was an excellent explanation of how this technique works! Thank you so much for that!
I understand now that the shift of the sensor (+ the roll) in practice is sufficient enough to compensate for almost all the shake movements that the user can cause, that is awesome info.

So, even the astrotracer function only compensates for the movement of the earth by shifting and rolling the sensor? Now that I think about it, it seems obvious because we have to focus at infinity for the stars to be sharp. If the focus is at infinity, there will be a fixed plane at the sensor where everything will be sharp from corner to corner. Tilting and yawing could even cause less sharpness because the objects remain at infinity, in theory a plane parallel to the lens, but the sensor could move out of focus on both sides (one side inwarts, the other side outwards) if one uses a wide aperture with a relative shallow DOF. Am I correct in thinking this? Or will even wide open the DOF be enough to keep the stars sharp if the sensor would tilt and/or yaw? For example while using a 15mm or another ultra wide lens it would create less problems compared to the 300mm you mentioned? I guess it all depends on the amount of real tilt and yaw the sensor can do?
If pentax will bring such mechanism to a new camera in the future, my respect for the engineers who develope this and to the programmers who must write the code to control this kind of system.

I am always amazed how incredibly good this SRII now already works when I take a look at the display (at 100% zoom) after taking a picture.
0.8" is the longest exposure that I could take without a tripod, only by resting my arm on my elbow. The result was pin-sharp. Isn't it unbelievable that we can use such incredible advanced technology today? I really admire those genius engineers who design such technology.
A long time ago, we all had to think about preventing camera shake and getting more noise with higher ISO films. Nowdays things are so much improved with IBIS and very sensitive low noise sensors, that we hardly have to think about it. We are spoiled by the progress in technology.

Have a great day photoptimist and enjoy the Holidays!
Greetings,
Sven
12-20-2018, 06:54 AM - 2 Likes   #7
Pentaxian
Lord Lucan's Avatar

Join Date: Sep 2017
Location: South Wales
Photos: Gallery
Posts: 422
I think a medium format camera would be a better starting point for something with those facilities. Of course there is this large format digital view camera :-

8x10 (10x8) Large format digital back model LS911

It's 12 MPx is disappointing though, for a thing that size.

Last edited by Lord Lucan; 12-20-2018 at 08:05 AM. Reason: Typo
12-20-2018, 08:02 AM - 1 Like   #8
Pentaxian
photoptimist's Avatar

Join Date: Jul 2016
Photos: Albums
Posts: 3,756
QuoteOriginally posted by Svenneke Quote
Good morning photoptimist,

That was an excellent explanation of how this technique works! Thank you so much for that!
I understand now that the shift of the sensor (+ the roll) in practice is sufficient enough to compensate for almost all the shake movements that the user can cause, that is awesome info.

So, even the astrotracer function only compensates for the movement of the earth by shifting and rolling the sensor? Now that I think about it, it seems obvious because we have to focus at infinity for the stars to be sharp. If the focus is at infinity, there will be a fixed plane at the sensor where everything will be sharp from corner to corner. Tilting and yawing could even cause less sharpness because the objects remain at infinity, in theory a plane parallel to the lens, but the sensor could move out of focus on both sides (one side inwarts, the other side outwards) if one uses a wide aperture with a relative shallow DOF. Am I correct in thinking this? Or will even wide open the DOF be enough to keep the stars sharp if the sensor would tilt and/or yaw? For example while using a 15mm or another ultra wide lens it would create less problems compared to the 300mm you mentioned? I guess it all depends on the amount of real tilt and yaw the sensor can do?
The defocusing effects of tilting/yawing the sensor from the plane of focus inside the camera are different than the defocusing effects of tilting/yawing the realworld objects from the plane of focus out in the world.

On the sensor movement side, the defocusing effects are nearly independent of focal length. Only the numerical aperture of the lens matters. For an f/2.8 lens (whether wide or telephoto) on a K-1 (or other camera with 5 micron pixels), moving the sensor only 0.070 mm is enough to blur a pinpoint of star light into a 5 pixel ball. For an f/1.4 lens, the same blur happens with only a 0.035 mm movement.


Happy Holidays!

02-16-2019, 09:30 AM   #9
Pentaxian




Join Date: Jul 2011
Posts: 1,449
Come on guys. Tilt and shift require larger image circles than we have right now and you don’t just add tilt to a shake reduction system. Nice idea maybe for APS-C with FF lenses but still a lame duck.
A real TS lend would make much more sense than presenting a camera with tilting sensor in a fixed housing.
02-16-2019, 10:49 AM - 1 Like   #10
Pentaxian
photoptimist's Avatar

Join Date: Jul 2016
Photos: Albums
Posts: 3,756
QuoteOriginally posted by zapp Quote
Come on guys. Tilt and shift require larger image circles than we have right now and you don’t just add tilt to a shake reduction system. Nice idea maybe for APS-C with FF lenses but still a lame duck.
A real TS lend would make much more sense than presenting a camera with tilting sensor in a fixed housing.
Actually, implementing just the tilt function with the sensor does NOT require a larger image circle. The tilting sensor never needs to leave the image circle cone. It's only the lens-based tilt systems that require a larger circle.

On the other hand, it is true that shift always requires lenses with larger image circle even if shift is implemented with the sensor. And yet shift is the less interesting function in the digital camera era because it is so easy to create the shift effect in post. The tilt effect is much harder to create in post without either the added labor of focus stacking or the ugly use of fake bokeh to blur parts of the image.

Of course, building a tilting sensor would add cost and size to the camera.
02-17-2019, 03:31 AM   #11
Junior Member




Join Date: Dec 2018
Location: Ghent, East-Flanders
Posts: 26
Original Poster
Hi Zapp, Photoptimist is right, only the shift function requires a larger image circle, the tilt function can be done pretty easy within the image circle cone. I want to add and admit that lenses that protrude more into the body (like my 30y old ultra wide Pentax 15mm for example) will increase the chance for getting vignetting in the corners of the sensor that tilt/move in the direction of the lens, because the sensor might reach the end of the image circle cone faster with these lenses because the cone is shorter.
A 35mm and above should not have problems with sensor tilt at all. Besides, the DOF of a 15mm is so high that a tilt function is not needed.

I know that there are limits to tilting the sensor, but being able to rotate the plane of focus to a certain extent would be very handy.
Would it replace the possibilities of a real TS lens? Far from, but having the possibility to change the plane of focus a bit can be of great help in being more creative and also for using more optimum apertures in certain images where one otherwise does have to use apertures that start to cause diffraction.

A real TS lens can go far beyond the limits of tilting the sensor, shure, but the only option at the moment for Pentax users is the Rokinon/Samyang TS 24mm. Maybe users of this lens can provide some help here, but If Pentax would make a TS lens again, it could be a welcome addition to a camera that is so good for landscape and architecture and other technical stuff.

Have a nice day.
Sven
03-06-2019, 08:04 AM   #12
Pentaxian
Wheatfield's Avatar

Join Date: Apr 2008
Location: The wheatfields of Canada
Photos: Albums
Posts: 11,325
Interesting concept, but I can see a myriad of implementation problems beyond how much size and weight it would add to the camera.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
amount, angle, camera, customer service, dials, direction, explanation, focus, lens, mechanism, method, patent, pentax, plane, ricoh imaging, sensor, sharpness, steve, suggestion, tilt
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Maximum distance K3 can move its sensor when doing “Composition adjustment” pakinjapan Pentax DSLR Discussion 3 09-05-2018 09:29 PM
Composition Adjustment and Noise Reduction interested_observer Pentax DSLR Discussion 11 12-13-2017 04:00 AM
The K1, Shift Lens and Composition Adjustment interested_observer Pentax K-1 4 08-02-2017 09:37 AM
Does the K-P have Composition Adjustment? bladerunner6 Pentax KP 3 04-26-2017 04:19 PM
Composition Adjustment and Auto Level Questions eponym Pentax Full Frame 5 05-20-2016 04:40 PM



All times are GMT -7. The time now is 01:45 AM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top