Why do you need to export a kml from the small area?
The geocoding module allows you to download a background image from google earth.
You can look at your data in many ways and display all what you want, even in 3d.
If for some reasons you need to export the results in kml, you can still do it from the site processing, using the sparse geocoding module…
Handling of the data is much easier in google earth thank Matlab (zooming in/out, panning…). Plus, the background image in my case is not the google earth high resolution image. It is low resolution image.
I used the sparse geocoding module but it uses the full image for the export and not the small area.
In small area processing the resolution of your background Google Earth Image goes with your area. If your area is really small then SARPROZ downloads the highest resolution possible. If your area is very large then SARPROZ will download the image with a much coarse resolution. The size of the Google Earth image downloaded is roughly fixed (at around 2048*2048 pixels), so with smaller area comes better resolution and with larger area comes lower resolution.
The reason is that the background image is just to give an idea of the corresponding ground feature in a quick and fast way. And usually it will work quite well if your area is indeed small. If you browse a very large area in “small area processing” then it will take both time and space to download all the high resolution data and stitch them together.
If zooming and panning is an issue, it means that the area is not small.
The small area module allows you to do lots of low-level operations for data investigation.
If your aim is just browsing the data (and the area is not small), then you do not need the small area module…
If you just want to export part of your results: load a starting mask using the “load mask” function in site processing.
The use the scatter-plots function for applying additional conditions (eg samples/lines > this < that).
Save your final sub-selection, that you can then load using again the "load mask" function.
In this way you can use all site-processing functions on your sub-selection of points.
At that point, you can use e.g. the time series module to choose the options of your output and export it in kml...
Since you encountered this issue, we will consider including in the future also the kml option, but you'll need to wait a bit.
I would consider KML export from Small Area quite handy as well. The best part of it are possible fast changes of Ref. Point since exporting through Sparse Points Processing requires APS processing until Ref. Point is selected and playing with Load Mask/Scatter plots.