wlaidlaw Posted March 25, 2007 Share #1 Â Posted March 25, 2007 Advertisement (gone after registration) I have been reading various posts over the last few days on the subject of focus change on rotation of say an M8. Some of the arguments strike me a specious at best and some of the mathematics just do not stand up, unless maths have changed a lot since I did a science degree at Cambridge some 40+ years ago. I agree that given that the height to width ratio now becomes the width to height ratio, there may be tiny changes due to DOF but other than that I cannot see a reason for the focus to change. The object to lens and lens to sensor distance are not changing so why should the focus? In other words if the long edge is sharper than the short edge of the image that will remain the case on rotation but the sharper edges are now on the horizontal axis rather than the vertical, if you are using the camera in portrait mode instead of landscape. Am I missing something or is this another case of hokum in the mysterious world of the rangefinder? Â Wilson Link to post Share on other sites More sharing options...
Advertisement Posted March 25, 2007 Posted March 25, 2007 Hi wlaidlaw, Take a look here Focus Change on Rotation - A Load of old ........ I'm sure you'll find what you were looking for!
carstenw Posted March 25, 2007 Share #2 Â Posted March 25, 2007 You are missing something The discussed rotation is not about the optical axis, but around the vertical axis... Link to post Share on other sites More sharing options...
Englander Posted March 25, 2007 Share #3 Â Posted March 25, 2007 Since I wrote some of the hokum, I guess I will try to justify it to you. At the time a lot of people were concerned about evaluating the accuracy of their rangefinders, were coming to conclusions that I thought might be misleading them to feel they should make adjustments to fix the near focus of their cameras when such adjustments weren't needed. Â The math would hold, in my opinion, if you were focusing through the lens. With an M, we are focusing by substitution, and actually focusing through a window that is off-set from the lens, the viewfinder is not directly centered above the lens. I think that can be agreed. However, when focusing at a tilted angle to evaluate accuracy of the rangefinder, I think the orientation can affect the accuracy. If the camera is in landscape mode and tilted forward, the rangefinder is moved forward of the film plane, perhaps enough to cause some consternation among those who were doing the evaluation. If the camera is in portrait mode, then the rangefinder and the film plane are tilted forward by the same degree and probably produce a better evaluation of the focusing accuracy of the system. Â Joe Link to post Share on other sites More sharing options...
wlaidlaw Posted March 25, 2007 Author Share #4 Â Posted March 25, 2007 Joe/Carsten, Â Some of the discussions, unless the posters were being very unclear, definitely seemed to be discussing focus in landscape and portrait mode. One poster seemed to be saying if you focused in landscape mode and rotated the camera, the subject you focused on would no longer be in perfect focus. Sure, if you tilt the camera and you are changing the lens to object distance, you would need to re-focus but that is no different from re-framing. Â 1) I would agree that depending upon the distance to the subject and the lens fitted, there are varying degrees of error introduced by parallax because the distance measuring mechanism and lens are not coaxial. Â 2) I am still not convinced that tilting the camera in any plane or axis makes any difference, as long as the object to lens distance remains constant, since the entire viewfinder mechanism is tilted in exactly the same way as the camera. Apart from the parallax error, even if you are focussed on a plane at an acute angle to the camera, the point you focus on will always be in focus to the same degree for the same lens to focus point distance. DOF, particularly with a lens with a lot of aperture shift, such as the Summilux 35mm may become an issue at this point. Â .....or am I arguing at cross purposes? Â Wilson Link to post Share on other sites More sharing options...
carstenw Posted March 25, 2007 Share #5 Â Posted March 25, 2007 With respect to focusing distance for the case of the focusing/recompose method, I believe that modern lenses are designed to have a flat plane of focus, not spherical. This means that the distance in the corners is further than the centre, and if you focus something in the centre and then push it to the corner, then it is too far. Link to post Share on other sites More sharing options...
wlaidlaw Posted March 25, 2007 Author Share #6 Â Posted March 25, 2007 With respect to focusing distance for the case of the focusing/recompose method, I believe that modern lenses are designed to have a flat plane of focus, not spherical. This means that the distance in the corners is further than the centre, and if you focus something in the centre and then push it to the corner, then it is too far. Â Carsten, Â You are absolutely right and this was what I meant by refocusing, if you effectively re-frame. I too am sure that modern lens designers aim for a flat field to get maximum edge definition. However, I feel that unless you are demanding a super-small CoC for very big enlargements and/or using a lens with a very small DOF such as the 75 f1.4 Summilux, the change is likely to be within the DOF but perhaps Joe's maths would prove me wrong. Â On a different point, have you noticed that when you do re-frame and are also locking the exposure by holding the shutter release at half pressure, how easy it is to inadvertently take the image too early. I suppose it is better than a shutter release that was too stiff. I think this is one aspect that might be being worsened by using an Abrahamson Softie, like I do. Â Wilson Link to post Share on other sites More sharing options...
stunsworth Posted March 25, 2007 Share #7 Â Posted March 25, 2007 Advertisement (gone after registration) Wilson, as a fellow softie user, I think it's something you get used to :-). Link to post Share on other sites More sharing options...
guywalder Posted March 26, 2007 Share #8  Posted March 26, 2007 I too am sure that modern lens designers aim for a flat field to get maximum edge definition. However, I feel that unless you are demanding a super-small CoC for very big enlargements and/or using a lens with a very small DOF such as the 75 f1.4 Summilux, the change is likely to be within the DOF but perhaps Joe's maths would prove me wrong. Wilson  Wilson, it seems to me (admittedly without having tried it out) that the problem is more apparent with wide angle lenses when the camera to subject distance is small. After all, the longer the focal length, the narrower the field of view so the less you can rotate before the subject disappears from the FOV. But with a wide angle and a close subject rotating the camera can, theoretically at least, place your subject a long way off the plane of focus as you re-compose. Mitigating that is the greater DOF of a wide angle lens. I have to say that I have never picked this up in practice, but maybe I just wasnt looking... Guy Link to post Share on other sites More sharing options...
msr Posted March 26, 2007 Share #9  Posted March 26, 2007 You guys should look at this thread:http://www.leica-camera-user.com/digital-forum/20010-focus-recompose.html  Malcolm Link to post Share on other sites More sharing options...
wlaidlaw Posted March 26, 2007 Author Share #10  Posted March 26, 2007 You guys should look at this thread:http://www.leica-camera-user.com/digital-forum/20010-focus-recompose.html Malcolm  Malcolm,  It was after reading the thread including Martin Tai's spreadsheet on Rangefinder forum amongst other threads that I started this thread. Your own post puts my mind at rest that as long as it is a plane surface, normal to the lens axis that we are taking, it is not a factor that we need to concern ourselves with. However, since it will be rare that the image plane will be exactly at right angles to the lens axis, if you re-frame to the edge of the frame, it might well be best either to at least think of aperture/DOF or refocus. Thanks for running the maths through Martin's spreadsheet. BTW when I clicked on the link to get the spreadsheet at the end of last week, it was a dead link.  Wilson Link to post Share on other sites More sharing options...
msr Posted March 26, 2007 Share #11  Posted March 26, 2007 BTW when I clicked on the link to get the spreadsheet at the end of last week, it was a dead link. Wilson  Wilson,  I have printed a PDF of Martin Tai's original film-based "Leica M & R Recompose Focus Guide" and a small image of the Excel spreadsheet. If someone can tell me where & how to upload it to one of those file sharing sites I will post it. Just remember that it is for pre-M8 Leicas.  Hopefully Woody, with his new calculations, will be able to update the spreadsheet for us.  Malcolm Link to post Share on other sites More sharing options...
wlaidlaw Posted March 26, 2007 Author Share #12  Posted March 26, 2007 Wilson, I have printed a PDF of Martin Tai's original film-based "Leica M & R Recompose Focus Guide" and a small image of the Excel spreadsheet. If someone can tell me where & how to upload it to one of those file sharing sites I will post it. Just remember that it is for pre-M8 Leicas.  Hopefully Woody, with his new calculations, will be able to update the spreadsheet for us.  Malcolm  Malcolm,  I like many others, use YouSendIt - File Sharing Transfer Delivery - PC FTP Replacement which is free for the basic ftp service.  Wilson Link to post Share on other sites More sharing options...
lars_bergquist Posted March 26, 2007 Share #13  Posted March 26, 2007 Being an evil old cynic, I do suspect that lots of this ‘focus shift on rotation’ flap is caused by the fact that it is extremely difficult when rotating the camera to keep the sensor in *exactly* the same plane. And even if you use a tiltable pan head on a tripod, you will of necessity move the optical axis! Add to this the belief that you should be able to use superspeed lenses wide open at distances where d.o.f. is measured in millimeters ...  For decades, advanced users of rangefinder cameras have used the technique of focusing in landscape mode and then rotating the camera to portrait format. These primitives were of course in the habit of stopping down a bit when possible.  Actually, we might well introduce *focus bracketing* with the M8. Use the motorised shutter to get off three shots rapidly, the second one leaning a little bit forward and the thitd a wee bit more laid-back.  The filthy old man from the Age of Stopping Down Link to post Share on other sites More sharing options...
msr Posted March 26, 2007 Share #14 Â Posted March 26, 2007 A PDF of Martin Tai's "Leica M Recompose Focus Guide" can be downloaded, for the next 7 days, at: http://download.yousendit.com/B82B38F00A691E83 Â A TIFF file, suitable for printing, of the Excel chart (for pre-M8 Leicas) for Martin Tai's Focus Guide can be downloaded here: http://download.yousendit.com/DAC303882C16514B Â Malcolm Link to post Share on other sites More sharing options...
wlaidlaw Posted March 27, 2007 Author Share #15  Posted March 27, 2007 A PDF of Martin Tai's "Leica M Recompose Focus Guide" can be downloaded, for the next 7 days, at: YouSendIt - File Sharing Transfer Delivery - PC FTP Replacement A TIFF file, suitable for printing, of the Excel chart (for pre-M8 Leicas) for Martin Tai's Focus Guide can be downloaded here: YouSendIt - File Sharing Transfer Delivery - PC FTP Replacement  Malcolm  Malcolm,  Thanks for doing that. What I would really like to get hold of is the original Excel file so that I can "get under the hood" and play around with the formulas - any ideas. I suppose I could email Martin.  Wilson Link to post Share on other sites More sharing options...
msr Posted March 27, 2007 Share #16 Â Posted March 27, 2007 Sorry Wilson, I don't have the original Excel file, Martin Tai is probably your best bet. Â Malcolm Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.