Multispectral Orthomosaic export

[Geomet]

I am working with multispectral imagery from the DJI Mavic 3 Multispectral (M3M). I have created a reflectance-calibrated orthomosaic in Metashape using a calibration panel.

When I export the orthomosaic, the output is very dark, and some bands (particularly Red and Red Edge) appear almost completely black when viewed outside of Metashape.

Could you please clarify:

1.
What is the correct expression to use in Set Raster Transform so that each band (G, R, RE, NIR) is exported with appropriate brightness and contrast for visualization?
----For example, is this the recommended form for calibrated data in the 0–1 range?
----(Bx * BRIGHTNESS) * CONTRAST + 0.5 * (1 - CONTRAST)

2.
How should I determine the appropriate brightness and contrast factors for each band?
– Should they be based on histogram percentiles (e.g., 2–98%), or are there default/recommended values for multispectral exports?

My goal is not to change the radiometric values (I need those preserved for analysis), but to create exports that are also easier to visualize in external software.

thanks.

[RoryG]

I work with M3M datasets pretty much all the time. To export the individual bands set the raster transform to the band you want then export as Index Value in the raster transform dialog. I don't change any other settings, once you have the raw values you can manipulate them in GIS for visualisation or raster calculations. In Metashape the individual bands may look like a single colour with little or no detail, this is what you get if you export as Index Color rather than value.

[Alexey Pasumansky]

Hello Geomet,

By default (is Raster Transform is not used in the Export Dialog) each output band will contain integer values in 0 - 32767 range, providing that reflectance calibration has been completed.

So if you do not want to change these values on export, you should adjust the visualization in the external software that you are using for post-processing analysis and visualization. For example, in QGIS you can use stretch feature for visualization based on selected interval of band values.