On Thu, Oct 03, 2013 at 03:14:24PM -0400, Igor Roshchin wrote: > > > Thu Oct 3 14:43:07 EDT 2013 > Godfrey DiGiorgi wrote: > > > On Oct 3, 2013, at 10:54 AM, Larry Colen <lrc at red4est.com> wrote: > > > > >> Wrong. Format ALWAYS influences apparent DOF. > > > > > > You are saying that if I take a photo with a D800, and in lightroom > > > crop it down so that I'm only using the area of a u4/3 sensor, then > > > the apparant DoF will change? > > > > No. By doing that you are changing the D800 format, making it the same > > as mFT format, and the resulting DoF will be the same if you use the > > same focal length lens on both cameras. You have changed the > > magnification by cropping the D800 format. > > > > G > > > Larry, the answer is "No", as Godfrey said, if you do not zoom in on > your screen (or in the print) after you did the crop. > If you zoom in, then yes, the apparent DOF will change, you'll see objects > less sharp.
That was pretty much my point. Depth of field is dependent on magnification aperture image resolution > > You are not surprised that somewhat out-of-focus photograph looks sharp > enough when you are looking at a small thumbnail (or preview on the back > of the camera), but when you blow it up on your 1900x1200 screen, you > clearly see it being OOF. > > Changing the crop factor is equivalent to blowing it up (to keep the > same overal size of the image [not of a particular object] on the > screen (print)). Exactly. If you blow up the image from a small sensor more, then DoF reduces. However, if you use a shorter lens to get the same AoV on a smaller sensor, then you get more DoF. The formula on the website is referring to the absolute DoF at the sensor, or film, level. When sensors outresolve lenses, then an increase in the sensor resolution will not have any effect on the absolute DoF. What most people don't think about is the difference between sensor DoF, and final image DoF. The final image DoF is dependent upon the effective resolution, which could be limited by the image resolution, the display(print resolution), or the human eye resolution depending upon how large the diplayed image is, how far away the viewer is, and how sharp the viewer's eyes are. The takeaway is that the DoF guides on lenses, are based on the assumption of the circle of confusion of a print of a certain size, viewed from a certain distance, and therefore, if you want to be sure of your zone focusing, or hyperfocal focusing, you should be conservative in your use of them. For example, I showed how in Cotty's case, the theoretical far focusing distance can vary significantly, when the theoretical hyperfocal distance is 2.06 feet, if you set your focus point to 2.00 feet or 1.9 feet, the max distance goes from infitity, to 72feet to IIRC 12 feet. In short, if you are using zone focusing, or hyperfocal points, you should probably give yourself half a stop, or a stop of margin. On the street, it rarely matters if the minimal focusing distance is 1 foot or 3 feet, but it does matter whether max focusing distance is 12', 72' or infinity. That is the opposite of what bit me with the picture of the storm trooper girl. I had set my lens for hyperfocal distance, and was just thinking, 8mm, stopped down, near infinite field of view, and then I had opened things up, and photographed something that was no longer in the zone of focus. Since I looked at it on the low res monitor on the camera, I didn't notice. I will still contend that to a first approximation that the same (length) lens, at the same aperture, shooting the same subject, either at the sensor level, or with the image cropped to the same AoV, or enlarged to the same final magnification (i.e. 12"x18" for a u4/3, 16" x24" for an APS or 24"x36" for FF) will have the same depth of field. In short, once again Godfrey and I are probably in violent agreement, but we're just feeling different parts of the elephant. > > HTH, > > Igor > > > -- > 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. -- Larry Colen [email protected] http://red4est.com/lrc -- 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.
