Spectral bandwidth

The width of the wavelength interval actually received or transmitted by an instrument at a measurement wavelength.

Setting an instrument to 260 nm does not mean that its detector receives a mathematically single wavelength. The actual signal comes from a finite wavelength interval whose width changes peak height, peak shape and available light throughput.

Bandwidth describes the instrument passband

Spectral bandwidth is commonly characterized by the full width at half maximum of an instrument's response to a narrow spectral line. It reflects the finite spectral interval covered by a measurement channel and is determined jointly by slits, dispersion, optical aberrations, detector pixels and the instrument line shape.

Wavelength sampling interval is simply the spacing between recorded points. Even if software saves points at very small intervals, a broad instrument bandwidth can make neighboring points contain largely overlapping information. Denser sampling cannot recover details already averaged out by the passband.

The trade-off between bandwidth and peak shape

If bandwidth is too wide relative to the sample's absorption band, different absorption at neighboring wavelengths is combined. Narrow peaks may broaden and lose height, and peak-position estimates may change. The Beer–Lambert relationship requires incident spectral bandwidth to be sufficiently narrow relative to the measured spectral feature.

Reducing bandwidth usually reduces light at the detector, potentially lowering signal-to-noise ratio or requiring longer integration. The smallest bandwidth is not always the best choice; balance target band width, required peak-position information and signal level.

Why settings alone are insufficient for comparing methods

Routine nucleic acid or protein quantification often concerns relatively broad UV absorption features, while narrow structures and peak-position analysis are more bandwidth-sensitive. Method suitability should be verified with actual samples and standards. A nominal bandwidth alone cannot predict performance for every spectroscopy task.

Nominal settings, actual instrument line shapes and sampling methods can differ between instruments. When comparing results, record actual bandwidth, wavelength step, integration and processing where possible, and confirm that reference values were established at compatible bandwidths.

Spectral bandwidth determines how wide a spectral interval each reading sees. Spectral resolution addresses whether neighboring features can actually be distinguished.