High frequency resolution in waterfalls forces poor time resolution that misrepresents room decay.

Evidence

In Figure 9.11(b) the frequency response at the back, near time = 0, is closely similar to the curve in (a), because a narrow 7 Hz bandwidth was chosen. However, the successive curves comprising the waterfall all look very much like it because in achieving high resolution in the frequency domain, the time-domain resolution has been sacrificed; it is 142 ms. Within the total 500 ms decay that is shown, nothing much changes because every successive curve moving forward is an average of events within a 142 ms window. The first several curves after t = 0 are almost identical, and those following simply drop smoothly in amplitude. It looks as though this room has uncontrolled ringing at many frequencies, but this is wrong. The data are misleading. This kind of pattern is frequently seen in audio discussions. — Toole, Floyd E, Sound reproduction- the acoustics and…, 953

Categorization

• DEFINE: Establishes the fundamental time-frequency resolution trade-off and how it manifests in waterfall measurements.
• SEED: Angle: common audio-forum waterfall plots that look alarming are often artifacts of measurement resolution, not real room problems.