okay so this has been bugging me for a while and i want to get some outside opinions. i do aerial scouting for a few medium sized operations, mostly wheat and some specialty stuff, and right now im flying a Mavic 3 Multispectral for my NDVI work. the farmers i work with are mostly interested in nitrogen stress and early disease detection, nothing super exotic.

here's my thing ? i keep reading that a decent RGB camera with the right processing pipeline can get you pretty close to what multispectral gives you for certain use cases, and obviously the cost and workflow difference is massive. like i have a buddy who swears he's catching stress patches just as early with a Phantom 4 Pro and some processing in Agisoft compared to my multispectral setup but i honestly dont know if that's real or if he's just seeing what he wants to see in the data.

the NDRE layers i get from the multispectral do seem to pick up stuff before it's visible to the eye but i also wonder how much of that actually translates into actionable decisions versus just being impressive looking maps that the farmer nods at and then does what he was going to do anyway. anyone here using both and actually comparing outputs side by side in real field conditions

your buddy is probably seeing real stress but the question is timing. RGB can absolutely show you chlorosis and canopy irregularities once things get bad enough to affect visible color, but multispectral is catching the physiological changes before that ? the plant is already struggling before you can see it in true color. for nitrogen stress specifically that early window is where the value is because if you're waiting until it's visible you've already lost some yield potential depending on growth stage.

that said i think you're right that a lot of the maps end up being more of a reassurance product than an actual decision tool unless the farmer is set up to do variable rate applications. if they're just going to blanket apply anyway then yeah the extra resolution you're getting from multispectral might not change outcomes even if the data is technically better.

tbh the workflow friction on multispectral is real and i think people underestimate it. calibration panel every flight, making sure your lighting conditions are consistent, the processing time ? adds up. for a one man operation doing contract work that overhead is not trivial. whats your turnaround time looking like from flight to deliverable right now?