Sound Insulation: Level Difference and Rating Displays

The module combines two-channel spectra, time averaging, reference-curve displays and report export. Its colour-coded comparison is a software indication. Confirm the measurement quantity and acceptance method before using any displayed rating in a contract or building assessment.

The problem it solves

The source describes contour fitting and spectrum adaptation terms. The available procedure does not establish all test inputs or demonstrate the current rating algorithms. A level difference must not be relabelled as transmission loss, and applying a reverberation correction to obtain DnT does not convert it into Rw. Numerical acceptance criteria and detailed conformity claims are withheld pending confirmation.

Functions

Screen layout

The left panel holds the measurement conditions: source and receiving channels, averaging time, T60 (13 typical room presets or manual entry), the product standard and the reference curve switches. The upper right table lists seven values — Rw, C, Ctr, Rw + C, Rw + Ctr, STC and the lowest TL dip — each with its standard and meaning, and shows the limit in red or green when a product standard is selected. Below are the 1/3-octave levels of source and receiving rooms and the stepped TL curve with optional Rw and STC reference contours.

The following describes the source workflow. The rating implementation and its compatibility with the required input quantities have not been validated for this website.

Step 1: 1/3-octave level difference

Both signals are windowed and transformed by FFT, and line energy is summed in the 1/3-octave bands of IEC 61260 (30 bands, 20 Hz to 16 kHz) to give sound pressure levels. The receiving room spectrum is subtracted from the source room spectrum band by band to give the insulation curve D. With energy averaging enabled, each band accumulates energy (10^(L/10)) and divides by the number of frames, on the equivalent continuous level basis, so source fluctuation does not make the reading jump.

Step 2: reverberation correction (DnT)

The more reverberant the receiving room, the more the reverberant sound raises the receiving level and the lower the measured insulation. GB/T 19889 uses the standardized level difference:

DnT = D + 10 · lg(T60 / 0.5 s)

At T60 = 0.5 s the correction is zero. Thirteen typical room presets are built in (for example recording studio 0.3 s, residential bedroom 0.4 s, office 0.6 s, classroom 0.9 s, reverberation room 2 s, workshop 2.5 s), or the measured value can be entered.

Step 3: weighted sound reduction Rw (ISO 717-1 shifting method)

The source describes fitting a reference contour over 100–3150 Hz and reading its shifted value at 500 Hz. The assigned rating name depends on the measured quantity: a level-difference curve cannot simply be reported as Rw. The detailed fitting rules and implementation require confirmation against the applicable edition.

Step 4: spectrum adaptation terms C / Ctr

Actual insulation depends on the noise spectrum acting on the insulation curve. ISO 717-1 Table 4 gives two normalized spectra (A-weighted total 0 dB): spectrum 1 for living noise (pink-noise character) and spectrum 2 for urban traffic noise (strong low frequencies).

Cj = −10 · lg Σ 10^((Lij − Ri) / 10) − Rw

Lij denotes the selected normalized source spectrum; Ri must represent the appropriate validated band quantity for the rating being calculated. Adaptation terms do not replace the underlying measurement or establish legal acceptance.

Step 5: sound transmission class STC (ASTM E413)

The source describes an STC contour display over 125–4000 Hz. Formal STC requires suitable transmission-loss data and the applicable ASTM test and rating methods. Field level differences can require different rating names. The source-specific rounding and single-band rule are withheld pending confirmation of the edition and implementation.

Dip diagnosis and display

The lowest TL point in the rating range is found and its frequency marked. Narrow, deep dips are usually related to panel coincidence (stiffness, damping), while broad, flat dips often indicate leaks (seals, holes, installation defects). Curves are drawn as steps at the 1/3-octave band edges (center frequency × or ÷ 2^(1/6)).

Preparation

For an exploratory level-difference measurement, select source and receiving channels and calibrate both. Formal assessment additionally requires the applicable source and microphone positions, spatial averaging, background correction, specimen and room information, and measured reverberation or absorption data. Follow the selected test method; two fixed microphones and an estimated room preset alone are insufficient.

Procedure

Typical settings

The four steps follow an engineer's line of reasoning: the overall number, what remains against real noise, which band is the weak point, and which frequency and cause.