okay this has been bugging me for a while and i feel like theres probably a standard way people handle it that im not doing. we're on a feature doc shoot next month and the director wants the option to go from the X7 to a thermal Zenmuse depending on the scene, and doing this on location with a full crew watching you fumble around with the quick release mount is not a great look.

like technically the swaps are easy enough on the bench but in the field with wind and people around and the pressure of turnaround time its a different thing. do you guys carry a second aircraft, do you do the swap on a ground pad, any tips for keeping the gimbal calibration from being a pain after you remount. last time i did a payload swap mid-day the gimbal horizon was slightly off and i didnt catch it until we were reviewing cards that night which was not fun

two aircraft is the only real answer if budget allows, the time cost of doing swaps plus the risk of something going wrong with the mount under pressure just isnt worth it on a proper production. we run two I2s on bigger jobs, one stays configured for the X7 all day and the other is whatever the secondary need is. yeah its more gear to transport but youre not standing there torquing a mount in 20mph wind while a gaffer is sighing at you

for the gimbal horizon thing, make sure youre doing the IMU calibration on a totally flat surface before first flight of the day and not just relying on whatever state it was in from the last job. i think a lot of horizon drift issues come from people skipping that step when theyre rushing setup

we do payload swaps semi regularly and honestly the gimbal calibration thing is just something you gotta budget time for. like do the swap, do a quick hover test at eye level, check the horizon on the monitor before you commit to a take. its maybe 4-5 minutes but it saves you the exact situation you described with the cards that night. also make sure the mount contacts are clean, we had an issue once where a dirty contact was causing intermittent gimbal behavior and we thought it was a calibration thing for like half a day before someone actually looked at the port