take no data into consideration when dezooming
currently, pixel with some part covered with no data are excluded. why not calculate average over present data? maybe a threshold of coverage?
currently, pixel with some part covered with no data are excluded. why not calculate average over present data? maybe a threshold of coverage?