Exponential sine sweep Farina · room impulse response measurement
f1 = 600 Hz, f2 = 4.4 kHz, T = 0.85 s. Equal time per octave — which is why it bends upward on a linear frequency axis.
f1 = 600 Hz, f2 = 4.4 kHz, T = 0.85 s. Equal time per octave — which is why it bends upward on a linear frequency axis.
fc = 2.9 kHz, fm = 190 Hz, β(t) = 5.5 sin²(πt/T), T = 0.95 s. Sidebands every fm; the dark gaps inside the diamond are zeros of the Bessel functions Jn(β).
f0(t) = 380 Hz + 160 Hz·t/T + 14 Hz·sin(2π·5.5 Hz·t)·t/T — a glide with growing vibrato. The fundamental is left out so the stack stays clear of the labels.
A tone whose frequency traces the sinc: main lobe, decaying side lobes, zero crossings at every integer x. The kernel behind every resampler.
mk(t) is row k of the rendered word “LimitlessGreen”. Tones sit on every third FFT bin centre: a Hann window spreads a bin-centred tone over exactly ±1 bin, so neighbours never beat.
Exponentially decaying noise as impulse response, applied to everything except the lettering, which stays dry and crisp.
Hann window w, N = 1024, hop R = 168, fs = 24 kHz, 78 dB display range.
ρ = 0.52 afterglow (0.66 in light mode), F = 1.25 flare, τ = 0.55 s; tx is when the beam passes column x.
Source: make_banner.py · make_video.py