SLOP3255 · Week 02 · Provenance

From telescope to speaker

A sonification is only as trustworthy as the chain of decisions behind it.

A telescope does not hand us meaning

countsdetector response before physical calibration
timeexposure, cadence, gaps and clock standard
contextweather, pointing, comparison source and quality flags

The “raw data” already belongs to an instrument, a place and a measurement protocol.

The complete route

DetectPhotons become charge, counts or voltage.
CalibrateCorrect known detector and observing effects.
ReduceDerive the quantity the scientific claim needs.
PresentSelect, scale and map the result to sound.

Each stage changes the object. Provenance is the record of those changes.

Calibration is not cosmetic

CorrectionProblem addressedIf omitted
Bias / darkelectronic offset and thermal signalbaseline drifts upward
Flat fielduneven pixel sensitivityposition looks like brightness
Reference starchanging atmospherecloud can resemble a transit
Time standardclock and location differencesevents cannot be aligned

Calibration makes comparison possible; it does not make the result neutral.

Preparation and presentation are different layers

Scientific preparation

  • calibration
  • quality flags
  • background subtraction
  • derived quantities

Auditory presentation

  • window selection
  • normalisation
  • mapping range
  • playback speed

Document both layers, because they answer different questions.

A minimum provenance record

FieldExample entry
Sourceinstitution + dataset identifier + retrieval date
Selectionobject, channel, time interval, exclusions
Operationscalibration, filtering, filling, normalisation
Mappinginput units and output range for every sound dimension
Softwaretool, version and reproducible parameters

Another person should be able to retrace the route without guessing.

The polished-output trap

Beautiful but opaque

A smooth melody, no units, no missing values and no description of filtering.

The listener cannot separate signal from design.

Inspectable

Audio beside values, plot, operations and a plain-language claim.

The experience survives scrutiny.

What single missing field would most change your interpretation?

Worked case: a small transit study

The sonification should receive the calibrated target/reference ratio, not arbitrary camera brightness. Its legend names that distinction.

Reference workflow: NASA Exoplanet Watch

Listening Lab 02 · Output: provenance diagram

Reassemble the chain

Task

Put twelve shuffled operations between telescope and speaker. Mark each as measurement, calibration, reduction or presentation.

Test

  • every arrow has an owner
  • units survive each stage
  • one uncertainty is visible

Reference: NASA Exoplanet Watch · science.nasa.gov/citizen-science/exoplanet-watch/