[Marcin LipiÅ„ski] “For example: sound from the voice recorder + Panasonic camera (50 fps) + Sony A7RIII (24 fps) + old Sony 29.97 fps camera … everything placed on the timeline agrees (Project for example 25 fps!) — the recording time from one device is exactly one hour from another device. “
Sound is different because one second does equal one second. There are no frames of sound, only samples.
For video and fractional, non-drop frame rates, one hour of timecode does not equal one hour of real time. For whole frame rates (25, 50) the timecode corresponds to real time. 23.98 is a non-drop format, so the timecode does not equal real time, that is to say 01:00:00:00 of 23.98 video does not represent an actual hour of real time. In real time, “one hour” of 23.98 timecode is actually 00:57:32:14 of real time. So that “one hour” of 23.98 video is actually is nearly 3 minutes short of the one hour of 25fps video.
Said another way, an hour of 23.98 video is 86314 frames. That same amount of frames in 25fps video is 57minutes, 32 seconds, and 14 frames. Hopefully that makes more sense.
In order to get these two to sync, you would have to slow down the 23.98 video to match real time (because an hour of 23.98 video does not equal an hour of real time), or you’d have to add 02:30:00 of video (so the 23.98 clip is 01:02:30:00) to match an hour of real time. It’s very confusing, I know. I have to deal with these conversions for 23.98 non-drop to 29.97 drop frame to hit commercial timings. It sucks.
Jeremy