Depth of Field in the Digital World

Just want to start with some background first. I've been doing a lot of macro lens work over a few months getting high speed video footage as well as DSLR video and stills for lower speed sample testing at work. Have spent many hours reviewing and editing all of the images. In the process I thought that I noticed a correlation between the number of pixels on the various sensors and the depth of what seemed to be in focus. <br><br>Had a day at home over the weekend that was too windy, cold and rainy to want to be outside taking pictures so I decided it was time to put this observation to a more controlled test to see what would happen. I set up a macro lens with a collar on a tripod and pointed it at a ruler that was held on edge by a clamp. The ruler was set at an angle to allow using the graduations to measure the depth of what seemed to be in focus. I then took shots with three different bodies all manually focused as well as I could using 10x liveview on the same graduation mark of the ruler. The pictures were all taken with the exact same picture style settings, ISO, white balance and everything else that could be set to the same settings. They were imported using the same DPP settings with no sharpness or noise reduction applied. They were batch processed to 8x10 JPEGs so that all would be treated the same while being brought to a specific physical size. On reviewing the results, there was a progression in the apparent depth of field that matched up with the sensor resolution.<br><br>On further reflection this is making sense to me. With film, the physical properties would remain reasonably constant for one brand. Change the format size and you change the magnification factor required to enlarge it to an 8x10 print. With digital sensors the resolution changes with the number of pixels. The image capture and processing is different as well. We create a file of an image in our computers that is defined by pixels instead of millimeters. The only time that the file becomes defined by a physical dimension is when it is printed, or converted to a specific size in processing.<br><br>The reason I am posting this is that only two of my three bodies use the same physical size of sensor and I am hoping that someone who reads this will have three bodies of different resolution that use the same size sensor. Combined with curiousity and a willingness to repeat a similar setup of course. I am just wanting to remove the extra variable of sensor size from the test at this point. At this point, I don't wish to post pictures. The results are strongly suggestive with the two bodies, but not conclusive due to the sensor size change for the third.<br><br>Thanks,<br>Jeff