Back from the dead
Ah ah, this project is still creeping somewhere in my brain!
My goal is to build a hardware/desktop software combo that will result in subframe precision syncing. Indoor, a GPS unit placed where a fix is available (near a clear sky window) will broadcast via RF its 1 PPS to one (or many) small board that will record the 1 PPS onto the scratch audio camera track (or any multitrack ENG recorder) followed by the UTC date and time binary words, Manchester encoded in hardware XORing a SPI output with its own CLK (not complying to the SMPTE LTC standard: my custom encoding will be post-processed by desktop software, see next paragraph).
In post-processing, a program will use the sampled audio frequency as a time base between two PPS pulses to interpolate wall time for each video frame. If a noticeable camera clock drift is observed, the program will split the video in equal segments, attributing to each the corrected SMPTE start time code. This later could then be used by your favourite editing software.
RF 1PPS broadcast could be done with off the shelf radio modules, in the FM radio band or ISM band. Initial pairing between the GPS and the audio module would be done on start-up with an IR link: the chosen FM frequency (determined after scanning for empty spaces) plus the UTC time and date are sent to the audio module. Then the camera with the attached audio module can be moved away from the GPS 1PPS repeater, the audio module keeping track of wall time internally without worrying about its own clock drift since the 1PPS will later on postproduction served as a basis for correct frame timing.
My goal is to build a hardware/desktop software combo that will result in subframe precision syncing. Indoor, a GPS unit placed where a fix is available (near a clear sky window) will broadcast via RF its 1 PPS to one (or many) small board that will record the 1 PPS onto the scratch audio camera track (or any multitrack ENG recorder) followed by the UTC date and time binary words, Manchester encoded in hardware XORing a SPI output with its own CLK (not complying to the SMPTE LTC standard: my custom encoding will be post-processed by desktop software, see next paragraph).
In post-processing, a program will use the sampled audio frequency as a time base between two PPS pulses to interpolate wall time for each video frame. If a noticeable camera clock drift is observed, the program will split the video in equal segments, attributing to each the corrected SMPTE start time code. This later could then be used by your favourite editing software.
RF 1PPS broadcast could be done with off the shelf radio modules, in the FM radio band or ISM band. Initial pairing between the GPS and the audio module would be done on start-up with an IR link: the chosen FM frequency (determined after scanning for empty spaces) plus the UTC time and date are sent to the audio module. Then the camera with the attached audio module can be moved away from the GPS 1PPS repeater, the audio module keeping track of wall time internally without worrying about its own clock drift since the 1PPS will later on postproduction served as a basis for correct frame timing.
Comments
Post a Comment