AcoustiTrack / Features
Version 0.1.0 · pre-release
What AcoustiTrack does
Every feature in the build that exists today. Nothing planned is listed.
In development — not released yet.
Measurement
The four views
- Spectrum — a real-time analyserRTA Real-time analyser: a live display of level in each frequency band. More in the dictionary: fractional-octaveOctave and fractional-octave bands Frequency bands of equal musical width — an octave or a fraction of one — as the ear hears them. More in the dictionary bands (1/1 to 1/48 octave) and the narrowband spectrumNarrowband spectrum The spectrum at full FFT resolution, one point per bin, without grouping into bands. More in the dictionary, with peak holdPeak hold A trace holding the highest level each band has reached since the last reset. More in the dictionary, and a spectrographSpectrograph Frequency against time, with level as colour. Shows how the spectrum changes. More in the dictionary pane underneath (history length, automatic or fixed colour range, colour or greyscale).
- Transfer Function — dual-channel transfer functionTransfer function What the system does to the signal at each frequency: output compared with input, as gain and phase. More in the dictionary: magnitudeMagnitude Gain per frequency of the transfer function, in dB. 0 dB means unity gain relative to the reference. More in the dictionary, phasePhase Timing of each frequency relative to the reference, in degrees. Used to align loudspeakers. More in the dictionary and coherenceCoherence How consistently the mic signal follows the reference, 0 to 1. High supports the reading; it is not proof of accuracy. More in the dictionary, with the live impulse responseImpulse response The system's response to a single click: direct sound, reflections and decay, against time. More in the dictionary beside it. Coherence fades the trace where the measurement is not trustworthy.
- Compare — two stored measurements compared, one subtracted from the other.
- Impulse Response — a swept-sine impulse response with arrival, decay and clarity figures.
Analyser
- FFTFFT Fast Fourier transform: splits a block of audio into its frequencies. Larger blocks separate low frequencies more finely. More in the dictionary sizes from 256 to 131,072 points; HannWindow function A taper applied to each block before the FFT to limit leakage. Hann is the usual choice. More in the dictionary, Hamming, Blackman-Harris or rectangular window.
- Multi-time-windowMulti-time window Different FFT lengths for low and high frequencies, spliced into one even-resolution trace. More in the dictionary (MTW) mode: three window lengths at once, for even resolution across octaves.
- Averaging by count (none to 32, or infinite) or by time (1, 2, 5 or 10 seconds); RMS (power) or polar (complex)Complex averaging Averaging real and imaginary parts, so noise with varying phase partly cancels and the steady response remains. More in the dictionary averaging for the transfer functionTransfer function What the system does to the signal at each frequency: output compared with input, as gain and phase. More in the dictionary.
- An "instant" trace — the last frame, faint, behind the average.
- Bands too narrow for the FFT size to resolve are marked as such, not drawn as if they were real.
Microphones
- Up to 8 measurement microphones, each with its own name, colour and visibility, and each with its own reference channel.
- Microphone correction curves: load the maker's response file per microphone (.txt, .csv, .frd, .cal, .dat).
- Average mics: the main trace becomes the energy mean of the microphones shown (spatial averagingSpatial averaging Several microphone positions combined into one response for the listening area, not one seat. More in the dictionary).
- An internal reference, so no Y-splitter is needed in the measurement rig.
- Loopback out: a spare output carries an exact copy of the test signal — muted until armed — so a plain cable into the reference input does the splitter's job.
Transfer function and delay
- Find delay per microphone, and delay tracking that follows drift; nudge a delayDelay Time offset between reference and microphone, in ms or as an equivalent distance. Used to time-align loudspeakers. More in the dictionary by hand. A delay is refused, not guessed, when the correlation peak is unclear.
- Fractional-octave smoothingSmoothing Averaging a response over 1/N octave around each frequency, to show its overall shape. More in the dictionary (1/48 to 1 octave).
- PhasePhase Timing of each frequency relative to the reference, in degrees. Used to align loudspeakers. More in the dictionary shown wrapped, unwrapped or as group delayGroup delay How late each frequency arrives, in ms — the slope of the phase. More in the dictionary; coherenceCoherence How consistently the mic signal follows the reference, 0 to 1. High supports the reading; it is not proof of accuracy. More in the dictionary plain or squared; blanking below a coherence threshold.
- Target curves: flat, X-curveTarget curve The response shape a tuned system aims for, such as flat, a house curve or the cinema X-curve. More in the dictionary (cinema), two house curves, or your own from a file.
- Display weighting of the traces (A, C or Z).
Impulse response and room acoustics
- Swept-sine measurement, with several sweeps averaged on request.
- Decay time per octave band (T20T20 Reverberation time measured on the decay from −5 to −25 dB and scaled up to 60 dB. More in the dictionary, EDTEDT Reverberation from a fit to the first 10 dB of decay, scaled to 60 dB. Tracks perceived reverberance. More in the dictionary) with the Schroeder curveSchroeder curve Smooth decay line from the backward-integrated impulse response; the decay times are read off it. More in the dictionary, and an All Bands table of decay and clarity.
- Clarity and definition: C50C50 Early (first 50 ms) to late sound energy, in dB. The clarity figure for speech; higher is clearer. More in the dictionary, C80C80 Early (first 80 ms) to late sound energy, in dB. The clarity figure for music. More in the dictionary, D50D50 Share of the total sound energy arriving in the first 50 ms, as a percentage. Higher favours speech. More in the dictionary and centre timeCentre time When, on average, the sound energy arrives at a seat, in ms. Lower values mean a clearer sound. More in the dictionary.
- Energy-time curveEnergy-time curve The impulse response as energy in dB against time. Shows individual reflections clearly. More in the dictionary (ETC), the impulse in dB, an IRImpulse response The system's response to a single click: direct sound, reflections and decay, against time. More in the dictionary spectrographSpectrograph Frequency against time, with level as colour. Shows how the spectrum changes. More in the dictionary, and the IR's frequency responseTransfer function What the system does to the signal at each frequency: output compared with input, as gain and phase. More in the dictionary.
- Broadband filters on the impulse before the findings (high-pass 100 or 200 Hz, low-pass 5 kHz, or a 100 Hz – 5 kHz band).
- Speech intelligibility: STISTI Speech Transmission Index: 0 to 1 score for how intelligible speech is at a position (IEC 60268-16). More in the dictionary from the impulse response and %ALcons%ALcons Estimated percentage of consonants likely to be misheard. Lower is better; usually converted from the STI. More in the dictionary from it; and STIPA, with AcoustiTrack's own STIPASTIPA STI measured with a modulated test signal played through the sound system (IEC 60268-16). More in the dictionary test signal built to IEC 60268-16.
Signal generator
- Pink noisePink noise Noise with equal power in every octave. The standard test signal for tuning a system. More in the dictionary, white noiseWhite noise Noise with equal power per hertz, so most of its energy is at high frequencies. More in the dictionary, sine, pseudorandom (periodic pink) noise, speech-weighted noise, the STIPASTIPA STI measured with a modulated test signal played through the sound system (IEC 60268-16). More in the dictionary test signal, and playback of your own WAV or AIFF file; the swept sineSwept sine A test tone gliding from low to high frequency, deconvolved into the impulse response. More in the dictionary for impulse measurements.
- Level steps and a slider with a ceiling that follows the signal type; a noise fade-in; presets; the same signal on extra outputs, each trimmed.
- Safety by design: nothing plays until the output is armed; the output disarms itself on any device, sample-rate, channel or output change; an armed-but-silent state is shown as such.
Spectrum tools
- Track peak: marks the most prominent peak, with its musical note — for feedback hunting.
- Cursor readouts in frequency, level, wavelength, musical note, time and distance.
- Captures: store a measurement, overlay it, shift it up or down to compare shapes.
- Total harmonic distortion (THDTHD Total harmonic distortion: the overtones a device adds to a pure tone, relative to the tone itself. More in the dictionary and THD+N) with a sine.
Venues, reports and data
- Venues: each job in its own dated folder — details, pictures, files and saved readouts.
- Templates: save and load your global settings under a name.
- Reports: choose the sections; export as HTML, PDF or both, with your company's logo and footer.
- Export traces as CSV — bands, bins, transfer functionTransfer function What the system does to the signal at each frequency: output compared with input, as gain and phase. More in the dictionary or impulse — with a choice of decimal mark.
Working with it
- Remote viewing over your local network: a built-in web page shows the charts on a phone or tablet; off by default and read-only unless you set a password, which allows a few settings to be changed and the output to be switched off — never armed: arming is only at the Mac.
- A Control Panel for a Sidecar iPad, and a second Chart Window with its own view and mic.
- Illustrate Routing: a live drawing of how signals flow through your rig.
- Quick Help on every control; light and dark appearance.
- Save Support Report: one text file to paste into an e-mail to support — nothing is sent by the app.
Platform
- A standalone macOS application.
Sound level
- Calibration to dB SPL per microphone from an acoustic calibratorAcoustic calibrator Device that plays a known level (94 or 114 dB at 1 kHz) into a microphone, so levels read in dB SPL. More in the dictionary (94 dB, 114 dB or another stated level). A capture that is clipped, unsteady, too short or implausible is refused.
- Level readout: Fast, Slow, Peak, LeqLeq Energy-average level over a period: the steady level with the same total sound energy. More in the dictionary (since reset, 1, 10 or 15 minutes, or your own period) and C − A, with A, C or Z weighting.
- Meter grid: one meter per microphone in a floating window.
- Level history: logging over a show with Max, L10L10, L50, L90 Levels exceeded for 10, 50 and 90 % of the time: louder parts, typical level and background estimate. More in the dictionary, L50 and L90, notes, and a report.
- Venue limit: e.g. 100 dB(A)Frequency weighting A, C and Z filters for level readings. A follows the ear and is used for limits; C keeps the bass; Z is flat. More in the dictionary LAeqLAeq A-weighted Leq over a stated period, e.g. 15 minutes — the usual form of a venue noise limit. More in the dictionary over 15 minutes, with how much room is left.
- Alarms per microphone: a level exceeded for a set time.
- Noise exposure: NIOSHNIOSH Noise exposure rule: 85 dB(A) for 8 hours, permitted time halves every 3 dB. More in the dictionary and OSHAOSHA US workplace noise rule: 90 dB(A) for 8 hours, 5 dB exchange rate; separate PEL and hearing-conservation thresholds. More in the dictionary daily dose.
- Maximum SPL: an AES75Maximum linear SPL The loudest a system plays while still meeting stated linearity criteria, found by stepping the level up. More in the dictionary-style stepping procedure that finds the loudest linear level of a system, with AcoustiTrack's own test noise.