Field of view, sensitivity and instrument tradeoffs

A useful instrument choice begins with a question

A large angular region, a faint compact source, two close features and a rapidly changing source are different observing tasks. An instrument can be excellent for one and unsuitable for another. State the measured quantity and necessary coverage, detail and time information before comparing specifications.

Field of view is the angular region accessible in one observation under the instrument’s definition. Pixel scale describes the angular sampling associated with an image pixel. Angular resolution describes distinguishable detail, while sensitivity concerns the strength of a signal that can be measured under stated conditions.

Spectral resolving power concerns wavelength structure, and cadence concerns when successive observations are made. Exposure duration can average changes occurring during an observation. These properties cannot be compressed into one universal “power” ranking.

The next original chapter develops resolution, sensitivity and contrast in more depth. Here two controlled examples show how changing a system property can help one task while limiting another.

A larger image scale can cover less sky

Use an invented rectilinear geometric model with a square sensor 20 mm on each side and pixel pitch 0.01 mm, or 10 µm. It has 2000 pixels per side. Compare effective focal lengths 1000 mm and 2000 mm while holding entrance aperture and response fixed.

For sensor width W and effective focal length f, the full geometric angular width is θ =2 atan(W/(2f)). The inverse tangent, atan, returns an angle from its tangent; the result is in radians here. The model assumes an ideal undistorted mapping and does not include vignetting, atmosphere or finite optical blur.

The 1000 mm case covers about 68.75 arcminutes across. The 2000 mm case covers about 34.38 arcminutes. A centred hypothetical 40 by 20 arcminute patch fits within the wider square field but extends beyond the narrower one horizontally.

Near the field centre, the small-angle pixel scale is approximately p/f radians per pixel, where p is the pixel pitch. Converting radians to arcseconds gives about 2.063 arcseconds per pixel at 1000 mm and 1.031 at 2000 mm.

The smaller angular pixels sample the field more finely; they do not by themselves improve the aperture’s diffraction scale or remove atmospheric blur. With the same entrance aperture, the longer focal length has not doubled collecting area. Per-pixel signal and extraction uncertainty require the delivered image and detector model.

Splitting time adds reads but can serve other goals

Reuse the detector lesson’s restricted noise model: independent Poisson source, sky and dark totals, independent per-pixel per-read errors, and negligible uncertainty in the known background mean. Let the extraction use four pixels and read noise 5 electrons per pixel per read.

For expected totals S =100 source, B =80 sky and D =20 dark electrons, one read gives variance 100 +80 +20 +4 ×25 =300. The signal-to-noise ratio is about 5.77. Split the same total integration into four reads, keeping those totals fixed, and the read variance becomes 4 ×4 ×25 =400. Total variance is 600 and the ratio about 4.08.

The comparison does not establish that one long exposure is always preferable. Shorter exposures can avoid clipping, retain time information, help reject transient contamination or manage changing conditions. Those benefits were excluded from the simple arithmetic and need their own evaluation.

If the integration is instead four times longer with one read and constant rates, the expected source, sky and dark totals become 400, 320 and 80. The variance is 400 +320 +80 +100 =900 and the ratio about 13.33. This is a conditional prediction, not unlimited precision or a real instrument’s performance guarantee.

Choose the compromise that preserves the needed evidence

A wide-field instrument can cover a region efficiently, while a narrower field may require several pointings. The resulting observing time, overlap, calibration and source variability affect whether those observations can answer the original question.

A high spectral-detail observation can divide available photons among narrower intervals. A timing observation may need short integrations even when each has a weaker signal. A bright target may be limited by dynamic range rather than by a shortage of photons.

In radio interferometry, fine spatial sampling through long baselines does not automatically provide every angular scale or a wide primary field. In optical observing, adaptive correction addresses some atmospheric effects without removing every source of blur or background. The following lessons explain those distinct limits.

A clear observing justification names the tradeoff and its consequences: which quantity will be measured, which structures or times will be covered, what calibration is needed, and which limitations remain. A successful choice preserves the evidence needed for the scientific question rather than maximising one isolated specification.

Worked examples

Try the reasoning

1. How many 0.01 mm pixels fit across a 20 mm sensor side?

   Show the answer and reasoning

   2000.

   20/0.01 =2000. This is physical sampling, not a resolution guarantee.

2. Does changing 1000 mm focal length to 2000 mm at fixed entrance aperture multiply collecting area by four?

   Show the answer and reasoning

   No.

   Entrance area stays fixed. The geometric field and image scale change.

3. Can shorter exposures still be useful when repeated reads add noise?

   Show the answer and reasoning

   Yes.

   Time information, clipping control or transient contamination can justify the tradeoff.

4. Does the largest value of one specification identify the best instrument for every observing task?

   Show the answer and reasoning

   No.

   Coverage, detail, sensitivity, timing, dynamic range and calibration must match the question.

What to carry forward

• Specify the measurement before selecting field, image scale, timing and spectral detail.
• A fixed sensor covers less sky at longer focal length; finer sampling is not automatically finer physical resolution.
• Exposure and instrument tradeoffs require explicit source, noise and observing assumptions.