On 2 Nov 2004 at 16:58, Jens Bladt wrote: > Hello Mishka > I admit to not having printed the taget in 600ppi, just 300ppi on an office > printer - so the target is a bit "rusty". (I am consiodering odering an > "official" test sheet (19 USD), showing all 5 or 9 USAF tagets on the same > poster).
Doubling the distance from the target doubles the effective target accuracy. I print mine in photo quality paper using a well maintained 1200dpi LASER printer and it got more than sufficient resolution for any camera lens. > The print quality of the tagets is's on source of error. Annother one is the > fact, that I don't know how to calculate the results from a "55 times Focal > Length" distance, since I'm actually looking at a factor 1.5 crop (a 50mm lens > gives a FOV equal to a 75mm on film). I guess the 55 times rule don't apply > here. The calculations are relative to magnification and hold true regardless of format so you need to use the FL printed on the lens for all calculations. > Nevertheless, if I change the calculations, I still get very poor > resultion compared to the 98 lpm that this is know to achieve on film. It's simple enough to determine the absolute limits of the system, the active pixels are 7.80�m square so there will only be 128 pixels (if you are considering luminance only) per mm of sensor surface therefore there can only theoretically be 64 line pairs recorded per mm (luminance only). Couple this with the physical optical anti-aliasing filter (to keep the high frequency components very slightly below the Nyquist frequency of the sensor) plus the errors and approximations introduced by the demosaic algorithms and the cumulative effect of the actual lens resolution and I'm surprised I even measured 46lppm. Rob Studdert HURSTVILLE AUSTRALIA Tel +61-2-9554-4110 UTC(GMT) +10 Hours [EMAIL PROTECTED] http://members.ozemail.com.au/~distudio/publications/ Pentax user since 1986, PDMLer since 1998
