investigate sentinelhub retrieval of float32 vs int16

[bossie]

Context: mail titled "Re: Problematic process graph - too many credits used"

Specifically a suggestion by D. Thiex:

don't load data as 32bit tiff but only as 16bit tiff (divides costs by 2)

expensive_graph.json.txt

[bossie]

This is already the case for this collection (note the sampleType of UINT16):

{
      "input": {
        "bounds": {
          "bbox": [676800.0, 5172070.0, 679360.0, 5174630.0],
          "properties": {
            "crs": "http://www.opengis.net/def/crs/EPSG/0/32632"
          }
        },
        "data": [
          {
            "type": "sentinel-2-l2a",
            "dataFilter": {"timeRange":{"from":"2023-09-25T00:00:00Z","to":"2023-09-26T00:00:00Z"}},
            "processing": {}
          }
        ]
      },
      "output": {
        "width": 256,
        "height": 256,
        "responses": [
          {
            "identifier": "default",
            "format": {
              "type": "image/tiff"
            }
          }
        ]
      },
      "evalscript": "//VERSION=3\nfunction setup() {\n  return {\n    input: [{\n      \"bands\": [\"B03\", \"B11\", \"SCL\"]\n    }],\n    output: {\n      bands: 3,\n      sampleType: \"UINT16\",\n    }\n  };\n}\n\nfunction evaluatePixel(sample) {\n  return [sample.B03 * 10000, sample.B11 * 10000, sample.SCL];\n}"
    }

[jdries]

investigation done and can be closed?