Archived posting to the Leica Users Group, 2001/07/25
[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]> > I believe your premise is flawed. You need to > > sample at slightly more than 2x the maximum frequency > > you want to acquire. A scanner IS a data acquisition > > system. 53 lp/mm is 106 lines/mm, which, as > > you say, is ~2700 SPI. In order to assure that > > the scanner can actually "acquire" all the lines, > > both black and white (which is what a line pair is), > > you would need to sample at > 5400 SPI. > > No. 2700 dpi is sufficient. 2700 samples per inch = 1350 cycles > per inch = > just over 53 cycles (i.e., line pairs) per millimetre. As I said, that premise is flawed, whether you believe it or not. It shows an ignorance of digital signal processing basics. 2700 SPI will not reliably sample 53 lp/mm, period. You will see "something", but not lines. I have shown why, and if you aren't able to understand that, then, if you have the ability, pick up some books on digital signal processing and attempt to understand them. By your understanding, you could just use one pixel to sample the entire scene! Just saying "no" does not make you right. You have not, and can't, because it's simply wrong, provide any substantiation to this flawed "understanding" (really lack thereof).