In the sixties, many original photographic-records were processed this way: some traces were selected (in order to construct a final film-section with well distributed data) and digitized by hand. The filtering criteria were function of several parameters, such as total profile length, station spacing, data quality, etc.. For instance, the example below belongs to the DSS-profile Volgograd-Nakhicevan, shot-point N-2 to the South. Shot-point position on the profile is 1030 (in pickets; 1 picket=100 m). Each trace contains geophone-position (not offset!) and absolute time from the time break. To determine the offset it's necessary to calculate the difference between geophone and shot-point positions. Here: the 1st trace has an offset of 228.5-103.0=125.5 km; reduction velocity is 6.0 km/s, so the reduced time is 24.0-125.5/6.0=3.08s. It means that the first arrivals of Pg waves is about 3.0s, PmP phase at about +4.0s. Amplitudes are not normalized, we can consider only relative aplitudes. The photographic multichannel seismograms were obtained using band-pass filter centred at about 10 Hz. one trace every 60  were drawn (that means a spacing station-plot every 6 km on average), and properly positioned in time. Parameters: reduction velocity: 6.0 km/s; main vertical lines distance: 1 sec.; amplitude: not scaled; numerical labels: BIG: geophone position axis [hm: from 2285=228.5 km up to 365.0 about, that is 136.5 km length], small: time-reference [secx10].