Range-gated viewing systems, which were at first used for object observation under poor visibility conditions, are used now also to determine distances to objects and obtain 3D images. A number of methods were suggested to determine distance. For instruments operating in real time, it is important to use fast methods. Range-intensity correlation methods are such methods because they use only two images. The methods work with rectangular-shaped illumination pulses and rectangular gate pulse shape. However, real pulses may differ from the rectangular ones that may introduce significant error into the distance measurement. To avoid the error one has to use a calibration dependence of the distance on the signal. The calibration curve depends not only on time profiles of the pulse and gate, but also on the gate duration and the difference between delay times of two images. Experimental search for optimal measurement condition through trial and error is a rather cumbersome problem. Therefore, we used numerical simulation and analytical estimates. The pulse time profile could be taken arbitrary. The gate time profile was assumed to be rectangular. The ratio of the energies received by the same pixel in two images depending on distance is used to build calibration curve. With given pulse shape, by varying gate time and delay time difference, we obtain calibration curves with different slope, different position on the distance scale and with different range of the determined distances. It appeared that the greater the slope of the curve, the narrower the range of measured distances and the lower the pixel energy. The conditions were found to get monotonous calibration curve with the highest possible return signal from all points on the distance determination interval.