CloudCompare forum

Hello,

I'd like to open a 3*16-bits bands GeoTiff file in CC, and to assign each band to a specific Scalar field (for instance, Band1, or R, being the luminance, Band2, or G, the height values, and Band3, or B, being a progressive mask.

I managed so far to load the Tiff file in CC and to convert the RGB components to each Scalar Field following this command (Edit > Colors > Convert to Scalar Fields), but each of them are now converted to 8bits (256 values) instead of the 65535 values expected (loosing therefore all the accuracy of my set)

Is there any workaround to encompass this limitation?

I already tried to open each band from separate GeoTiff files, and to pass the values among them from SF to RGB, then back to SF with the command "Edit > Colors > Interpolate from another entity", which works pretty fine with small or medium sized datasets.... but fails with huge geotiffs (10000*10000 pixels)?

Thanks in advance for any hint or tips...

Best Wishes,

Alban

Indeed, RGB colors are stored as 8 bit values (it's quite a standard in fact).

So your only solution would be load the color bands as scalar fields. I believe it wouldn't be difficult to load 16 bits bands as SF by default. Can you send me a sample file (with the 3 16 bits bands)? (cloudcompare [at] danielgm.net)

Hello. Any further steps here? Is there a way to overcome this 16bit to 8bit thing? When importing an rgb point cloud where the rgb information is stored as 16 bit values and exporting it, the values appear to be 8 bit values. It would be nice to keep them as 16 bit values. For point clouds produced by photogrammetric processes, 16 bit imagery is often used. Especially when different chanels are used as colors , for example the NIR. channel, it would essential to handle these at original precision.

Well, as I replied to the previous user, it would be possible to load this color fields as scalar fields (to avoid this issue). However I never got the sample dataset, therefore I couldn't do anything...