Hi Alan, On Sun, 22 Dec 2013 07:14:36 +0200 Alan C wrote: > >Having read this thread, I can only conclude that either I am dof or there is >some some trick to being successful with this procedure. > >Last week I decided to fine tune my identifiable lenses on the K7. I used Bob >Atkins grid, set the lenses at their widest aperture & worked as close as I >could get. With each lens, I took shots at -10, -5, -2, 0 +2, +5 &+10. Well, with all lenses, there is absolutely no difference between the shots. > >OK, so I figured it was because they are slow consumer lenses and decided to >try it with my Super Tak 50/1.4 - same result.
Hmm, not sure what Bob's grid looks like, but I have had good success with a PDF I found on the net somewhere, that is printed out with a laserprinter, and has a high-contrast horizontal dark line in the center as a focus target. It is only a few millimeters wide, and important, there is NOTHING above and below it to make sure it is really focussing on this line and nothing else. The line is perhaps 5 centimeters or 2 inches wide. To each side of this focussing target, there is a sort of 'ruler' that has various patterns but more importantly has a series of numbers with 0 in the middle, en going down to -20 or so and +20 on the other side. What I do to adjust focus for a lens is: - Put camera with lens on a tripod, and position it such that it 'sees' the piece of paper withy the target at an angle between 30 and 45 degrees, at a distance that is suitable for the lens, usually between 50 cm and perhaps 3 meters, depending on focal length. It is important to make sure the focusing target and a significant part of the 'ruler' area are in the frame, and large enough ... - Set the camera to use the 2-second delay (which also does a mirror-up on K7/K5/K3) - Set the lens to it maximum aperture, wide open (for minimum depth of field) - set the lens to infinity, then let if autofocus on the target line (using center, spot focus) - Take the shot, then use the review/display to judge focussing. (You COULD do this after downloading to the computer, but that takes too much time for me) - Look at the ruler, ENLARGED at least 8x, and pan over it to see if the focusing/fuzziness is distributed evenly around the ZERO mark (usually it is NOT) - When the area of best focus id BEFORE the target (front focusing), dial in a small negative value (like -3 or -4) for the correction (and I always use APPLY ONE, because I want to have specific correction per lens) - When best focus is BEHIND the target (BACK FOCUSSING), dial in a small postive value. - Repeat the test, and keep adjusting until the area that is the sharpest is centered around the ZERO marker. So I do NOT take a lot of shots in one go, then evaluate, I take one shot, evaluate on the back display, zooming the magnification and panning around the image/ruler, then decide on the correction to apply. A bit trial and error, but this is what has worked best/fastest for me on the K7, K5 and now the K3 ... For zoom lenses, I tend to use the longest focal length (has lowest depth of field) but depending on the lens, you could also use your 'most used' focal length. For macro I try to use a smaller subject distance, but that may require a smaller target and ruler, so a different, scaled, print-out. For extreme tele, you could even use an outside target, like a row of telephone poles, as long as you make sure the camera has a single edge/line like thing to focus on, with no distracting other things that are near that same distance. The important thing is, you HAVE to make sure the camera focuses on a small/narrow hight contrast edge or line, and nothing else, AND you need something WELL AWAY from the focusing point that will clearly show you where the actual plane of maximum sharpness is, so you can see if there is a front- or back-focusing problem to correct. Regards, JvW ------------------------------------------------------------------ Jan van Wijk; http://www.dfsee.com/gallery -- PDML Pentax-Discuss Mail List [email protected] http://pdml.net/mailman/listinfo/pdml_pdml.net to UNSUBSCRIBE from the PDML, please visit the link directly above and follow the directions.
