How to set data scales for non-8-bit images and choose the output bands (multiband)
This article will show you how to set custom data scale values and choose the output bands order. MapTiler Engine is capable of reading images in non-8-bit formats but sometimes default settings are not right. You might need to tweak the data scale values for output color bands to achieve desired visual results.
Setting custom data scale values is available only for non-8-bit images and requires at least MapTiler Engine Plus.
Output generated by MapTiler Engine contains tiles in RGB(A) color model. However, if your dataset contains more bands that encode additional information, MapTiler Engine provides you with the ability to map the bands from the input image to corresponding bands in the output RGB(A) model.
Band scale values
By default, MapTiler Engine reads the band data values and maps them to the full range provided by its format. However, your input image might contain band data values in 16, 32, or 64-bit format that do not cover the whole range of the format. In such cases you might see something like this:
This particular image is a piece of aerial imagery dataset that has all band data scaled to the 0 - 4000 range (instead of 0 - 65535 provided by the 16 bits). MapTiler Engine allows setting the min and max values for each of the bands of the input image. To do so, after opening a file in MapTiler Engine, you need to go to the Bands menu on the right-hand side of the window and select the Change bands… option. The effect will look like this:
Selecting bands order
Some data sources, such as Sentinel 2 aerial imagery, contain more information than the classic RGB(A) model. They might be additional spectral bands like Infra-Red, information for cloud detection, or plants vegetation data, each stored as a separate band in the input image. MapTiler Engine allows mapping these bands to corresponding bands of the output RGB(A) model. This setting is changed in the Bands → Change bands… dialog:
The example image used in this article, apart from the RGB channels, has a 4th band that contains the data from the NIR (Near Infra-Red) spectral band. To generate a monochrome image visualizing just this band, set Red, Green, and Blue bands to band 4. and leave the Ignore alpha channel setting. The result should look like this:
Conclusion
MapTiler Engine is capable of processing more complex input data than just classic RGB(A) images. In this article, we have shown you how to set custom data scale values. Also, we have presented how to choose the bands which you would like to end up in the output map when your input data contains more of them.