The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Calculating Max Shift

rga

Member
I could use some help from the resident gurus. I am trying to calculate how much shift I can use for my lenses. Here is what I know:

Lenses:
80mm APO Digitar with image circle of 80mm @ f11
35mm XL with image circle of 90mm
Sensor Dimension:
P45+ back with portrait sensor dimensions of 49.1mmHx36.8mmW

Maximum shift available:
18mm Left/Right

So my question is what is the furthest left or right shift I can use without any part of the sensor, in portrait mode, extending beyond the image circle? I have the same question for each lens with the sensor in landscape mode.

Is there a formula I can use (that hopefully doesn't involve pi)?

Thanks to any and all who can help,
Bob
 

Bob

Administrator
Staff member
From Pythagoras' Theorem:
The Radius of your image circle R must be larger than the square root of ((half of your sensor width)+hshift squared plus (half of your sensor height)+vrise squared).
maxima can be calculated by an inflection point in the first derivative.
Since the function for one is dependent on the value of the other we can compute:
Solving for vrise we have
vrise= SQRT(R^2-(w/2+s)^2)-h/2
where R is the radius of the image circle, w is the sensor width, s is the amount of shift and h is the sensor height
of course YMMV
-bob
 
Last edited:

rga

Member
From Pythagoras' Theorem:
The Radius of your image circle R must be larger than the square root of ((half of your sensor width)+hshift squared plus (half of your sensor height)+vrise squared).
maxima can be calculated by an inflection point in the first derivative.
Since the function for one is dependent on the value of the other we can compute:
Solving for vrise we have
vrise= SQRT(R^2-(w/2+s)^2)-h/2
where R is the radius of the image circle, w is the sensor width, s is the amount of shift and h is the sensor height
of course YMMV
-bob
But I just want shift...
 

GrahamWelland

Subscriber & Workshop Member
I could use some help from the resident gurus. I am trying to calculate how much shift I can use for my lenses. Here is what I know:

Lenses:
80mm APO Digitar with image circle of 80mm @ f11
35mm XL with image circle of 90mm
Sensor Dimension:
P45+ back with portrait sensor dimensions of 49.1mmHx36.8mmW

Maximum shift available:
18mm Left/Right

So my question is what is the furthest left or right shift I can use without any part of the sensor, in portrait mode, extending beyond the image circle? I have the same question for each lens with the sensor in landscape mode.

Is there a formula I can use (that hopefully doesn't involve pi)?

Thanks to any and all who can help,
Bob
SK 80 APO with P45+ - max 17mm / 19mm movements @90mm image circle, 11/13mm at 80mm.
SK 35XL - the same.

Look at the Alpa shift doc I pointed you to and look up the lens and then across to the 37x49 sensor column. That'll give you the max rise/fall or shift L/R which will equate to your landscape/portrait movements.

The calculation table I pointed out in the other spreadsheet allows you to enter the desired movements, sensor and lens dimensions and then will give you the equivalent lens coverage and whether it will fit the sensor dimensions too.
 

rga

Member
SK 80 APO with P45+ - max 17mm / 19mm movements @90mm image circle, 11/13mm at 80mm.
SK 35XL - the same.

Look at the Alpa shift doc I pointed you to and look up the lens and then across to the 37x49 sensor column. That'll give you the max rise/fall or shift L/R which will equate to your landscape/portrait movements.

The calculation table I pointed out in the other spreadsheet allows you to enter the desired movements, sensor and lens dimensions and then will give you the equivalent lens coverage and whether it will fit the sensor dimensions too.
Perfect! Thanks both Graham and Bob.
 

Udo

Member
If you don't like the calculating approach, then get some graph or scale paper and draw concentric circles corresponding to the image circle(s) you are interested in. Cut out a piece of paper of the same dimensions like your digital back's sensor and move it within the appropriate circle's area. The millimeter scaling of the graph paper gives you immediately the shift movement in mm.

Regards,
Udo
 

gazwas

Active member
IMO looking up lens IC is a misleading path to follow and its not just that black and white. For newbies and how I learned this, the relationship between IC and movements is not something you fully understand until you've purchased your expensive, shiny new camera and lens. The most important thing and something that is never mentioned when discussing IC and max camera movements is their relationship is not equal and is different with every focal length.

Many things come into play when using movements and while a manufacturer may quote massive IC numbers, the actual useably sharp IC is quite a bit smaller. Sharpness does fall off quite a lot as you move into the outer reaches of the projected image and especially with wide angles, field curvature and distortion can play havoc with large shifts. Also, light fall off will also limit the useable area as the corrections needed to balance this result in massive amounts of noise.

I don't own the two lenses you mention so can't comment on their usable IC but I'd be surprised if its the actual quoted Schneider numbers as I know my SK 43XL is nowhere near its quoted number.
 

Bob

Administrator
Staff member
IMO looking up lens IC is a misleading path to follow and its not just that black and white. For newbies and how I learned this, the relationship between IC and movements is not something you fully understand until you've purchased your expensive, shiny new camera and lens. The most important thing and something that is never mentioned when discussing IC and max camera movements is their relationship is not equal and is different with every focal length.

Many things come into play when using movements and while a manufacturer may quote massive IC numbers, the actual useably sharp IC is quite a bit smaller. Sharpness does fall off quite a lot as you move into the outer reaches of the projected image and especially with wide angles, field curvature and distortion can play havoc with large shifts. Also, light fall off will also limit the useable area as the corrections needed to balance this result in massive amounts of noise.

I don't own the two lenses you mention so can't comment on their usable IC but I'd be surprised if its the actual quoted Schneider numbers as I know my SK 43XL is nowhere near its quoted number.
Good points all, in addition the usable size changes with aperture and focus distance. Of course the definition of "usable" is personal. The shape changes with tilts (think conic projection on a tilted plane).
-bob
 

Thierry

New member
In this respect, the usuable IC given by Rodenstock for their lenses is more conservative, in general. One can easily go to the limit given, in many situations even quite a few mm over it.

Thierry

Many things come into play when using movements and while a manufacturer may quote massive IC numbers, the actual useably sharp IC is quite a bit smaller. Sharpness does fall off quite a lot as you move into the outer reaches of the projected image and especially with wide angles, field curvature and distortion can play havoc with large shifts. Also, light fall off will also limit the useable area as the corrections needed to balance this result in massive amounts of noise.

I don't own the two lenses you mention so can't comment on their usable IC but I'd be surprised if its the actual quoted Schneider numbers as I know my SK 43XL is nowhere near its quoted number.
 

Thierry

New member
The IC is always given for a subject distance (lens setting) at infinity, the worse-case scenario, when the distance lens to image plane is the shortest, respectively the IC the smallest, allowing the less movements within this IC.

And in addition, this IC at infinity is always give at a certain f-stop, in general the sweet spot of the lens (at least that was the case with LF lenses = f 22).

Thierry

Good points all, in addition the usable size changes with aperture and focus distance. Of course the definition of "usable" is personal. The shape changes with tilts (think conic projection on a tilted plane).
-bob
 

gazwas

Active member
In this respect, the usuable IC given by Rodenstock for their lenses is more conservative, in general. One can easily go to the limit given, in many situations even quite a few mm over it.
You may have more experience with a wider range of Rodenstock lenses and have heard this stated before but in my experience, I've not fount that to be the case and the ones I've tried, the quoted IC specs are pretty spot on. For example, 90mm for the 40HR-W. On a 40mm x 54mm sensor it will allow about 15mm of movement which is pretty on spec but IMO the final few millimetres go a bit squiffy so best to stick to about 12mm. After 15mm the edges go dark pretty quickly as expected. 15mm and above is still usable as long as its sky or something similar lacking detail but not so good if its an interior shot etc.

Longer lenses are different and as Bob stated, the IC can actually grow as you stop down and most digital lens IC's are quoted at f8 or f11 but below this diffraction plays its hand.
 

Thierry

New member
That's what I was meaning to say, in certain situations, e.g. sky or uniform subject without details. In this case one can go further than the given IC. Appologies if I was not clear.

One can argue of course, but in general that is my experience with Rodenstock lenses (not so with Schneider) that one can go to the limit given without noticeable loss of quality.

Actually, the IC does not get bigger, when stopping dow, it is the usuable IC out of the given IC which gets bigger, which means that the light fall-off becomes less. That's why any IC is given for a certain f-stop.

The IC itself increases its size only when the distance lens to image plane is growing.

Thierry

You may have more experience with a wider range of Rodenstock lenses and have heard this stated before but in my experience, I've not fount that to be the case and the ones I've tried, the quoted IC specs are pretty spot on. For example, 90mm for the 40HR-W. On a 40mm x 54mm sensor it will allow about 15mm of movement which is pretty on spec but IMO the final few millimetres go a bit squiffy so best to stick to about 12mm. After 15mm the edges go dark pretty quickly as expected. 15mm and above is still usable as long as its sky or something similar lacking detail but not so good if its an interior shot etc.

Longer lenses are different and as Bob stated, the IC can actually grow as you stop down and most digital lens IC's are quoted at f8 or f11 but below this diffraction plays its hand.
 

RodK

Active member
Because the projected image is curved like a bowl, there is a soft area before total vignetting, that is not very good called the penumbra. You see image, but how wide the penumbra that is usable is subject to interpretation. The wider the lens the more acute is the curvature, and the longer the lens the flatter or less acute, the image projection. Stopping down in either case does increase the usable image. But you also, as you stop down, introduce some focus shift. This is usually covered by depth of focus at the film plane when at approximately F8.5 or F11 and below on digital lenses, Testing at shooting apertures would yield possible focus adjustments if you do use a variety of apertures when shooting. The wide angle lenses often introduce this shift more so than normal to longer lenses. The best is to go out and shoot as some said above and test with notes then look on your computer to determine your usable movements. As we are dealing with a circular projection, any rise/fall in addition to shift will also change the results as we take rectangular captures of sections of the projected image. And as someone said we all have our own 'GE' factor. (Good Enough). Just some more food for thought.
 

dougpeterson

Workshop Member
Working with a dealer that specializing in just this kind of gear can take much of the work out of the equation...

What photographer wants numbers?? I want visuals!

Prepared just for you:





Doug Peterson (e-mail Me)
__________________

Head of Technical Services, Capture Integration
Phase One Partner of the Year
Leaf, Leica, Cambo, Arca Swiss, Canon, Apple, Profoto, Broncolor, Eizo & More

National: 877.217.9870 *| *Cell: 740.707.2183
Newsletter | RSS Feed
Buy Capture One 6 at 10% off

Masters Series Workshop:
New England Landscape - Fall Color (Oct 5-8)
 

rga

Member
Working with a dealer that specializing in just this kind of gear can take much of the work out of the equation...

What photographer wants numbers?? I want visuals!

Prepared just for you:





Doug Peterson (e-mail Me)
__________________

Head of Technical Services, Capture Integration
Phase One Partner of the Year
Leaf, Leica, Cambo, Arca Swiss, Canon, Apple, Profoto, Broncolor, Eizo & More

National: 877.217.9870 *| *Cell: 740.707.2183
Newsletter | RSS Feed
Buy Capture One 6 at 10% off

Masters Series Workshop:
New England Landscape - Fall Color (Oct 5-8)
I can see you're just not going to let go of that bone!! :cry:

Thanks Doug. Nice charts! The first (bottom one) raises a few questions. The yello square that represents "2 Vertical Frame Stitch - Left/Right 15mm overlap 6.8mm)": Is that for the sensor in portrait mode? It looks like it must be.
And the "15mm [Overlap: 6.8mm]", for the P45+ in portrait mode the horizontal is 37mm. Moving it over 15mm leaves 22mm overlap from the center position.
So I must be missing something.

That said, wonderful graphics.
Thanks,
Bob
 
Top