All guides

Help Center · SpectraSense

Tune — magnitude, phase, coherence, polarity and the delay finder

Measure a system's transfer function against a source — the master or the test tone — match the levels, check the polarity, find and compensate the delay, save and compare snapshots.

Open SpectraSense

On this page
  1. The layout
  2. The source
  3. Levels, Match and polarity
  4. Measuring
  5. The three panes
  6. Tune options
  7. Snapshots
  8. Limits of a curve

Tune compares two signals: a source (what goes into the system) and the measurement mic (what comes out at the listening position). The difference between them is the system's transfer function — the speaker, its processing and the room — shown as magnitude, phase and coherence. It works on a phone, a tablet and a computer; an audio interface with a master is best.

The layout

On a computer the Tune view has two parts:

  • Left column, top to bottom: Source (Master or Test tone), the mic and master levels stacked on one scale with Δ · Match and the polarity, Measure, the Delay card (value, Apply, Track; From the distance folded underneath) and Save snapshot.
  • Right: the plot, with Transfer | Impulse above it, Freeze (holds the drawn curve while the measurement goes on) and Options — the Tune options panel.

With less room (a tablet, or an options panel open beside it) the plot comes first and the controls follow under it.

On a phone Tune is one column: the plot first, the controls under it. Measure sits in the action bar at the bottom, next to Start. A phone's own mic is a single channel, so the source is the Test tone: the phone plays pink noise into the PA and measures what comes back. The distance → delay calculator is in the Delay card's From the distance fold.

The source

Pick the Source at the top of the left column:

  • Master (best): patch the console or processor output that feeds the system into a second input of your interface. The curve is then the system and the room only, and any program works — pink noise, but also the band's music during a show.
  • Test tone: one input is enough. The test tone's own signal, as it leaves SpectraSense, is the source — play pink noise with Play test tone. The curve then also includes the computer's output, the system and the room. An input with one channel only always uses the test tone.

Until you pick one yourself, the source is Master once a master is set in Settings → Input, otherwise Test tone. Once you pick one, Tune keeps it.

The channels are chosen in one place only: Settings → Input — the mic channel and the master. Tune uses exactly those (the mic curve from Settings is applied to the mic channel). With Master as the source but no master set yet, Tune says No master yet — pick it in Settings with a button that opens it.

Levels, Match and polarity

With Master as the source, the left column shows the mic level above the master level, on the same dB FS scale — the same bars as the status bar. The difference between them reads Δ +3.1 dB · Match: tap it and the master's trim is set so both read alike (✓ Matched; Undo is in Settings → Input).

Next to it, Polarity says whether the master and the mic have the same polarity, from the sign of the impulse response's peak while you measure:

  • Polarity ✓ normal — the peak points up.
  • Polarity flipped — check the cable or the desk — the peak points down: a miswired cable, a polarity switch on a channel or the desk's output, or a speaker wired the other way round.
  • — until a measurement has run.

Measuring

  1. Tap Start in the status bar — on an upright phone, in the action bar (Measure never opens the input itself — while stopped it says Start first), then Measure.
  2. Watch the Delay to the speaker card. After about three seconds of both signals, SpectraSense estimates how much later the mic hears the source — about once a second, up to 1 s (and up to 0.1 s the other way) — and shows it next to the delay that is applied: Estimate 12.43 ms (+0.31). Nothing is applied by itself.
  3. Tap Apply (or press ⌘E / Ctrl+E in Tune) to take the estimate. The delay is then compensated — the phase is shown as if both arrived together — and the average starts over.
  4. Read the curve. Restart average after moving the mic or changing an EQ.

Off · Track · Follow:

  • Off — the estimate is only shown; Apply takes it.
  • Track — applies each clear estimate and stops by itself once three in a row agree within ±0.02 ms (Locked at 12.43 ms). Use it after moving the mic.
  • Follow — keeps applying every clear estimate: for wind, walking the mic, or a processor whose latency changes. Setting the delay by hand (−, + or a typed value) pauses Track and Follow — they never fight your hand; tap the mode again to continue.

A weak estimate is shown grayed and is never applied — not by Apply, not by Track or Follow; a wrong delay would bend the whole phase curve. Raise the level, check that the source is really the signal that feeds the system, or move the mic closer. Polarity inverted means the correlation peak is negative.

Fine-tuning: − and + nudge the delay by 0.01 ms (Shift = 0.1 ms; hold to repeat); the arrow keys in the field do the same, or type a value.

Distance: once a delay is applied, the card shows the distance sound travels in that time, in m and ft, at the air temperature — 20 °C is 343 m/s; sound is about 0.6 m/s faster per degree warmer.

From the distance: the other way round — type how far the speaker is, in m or ft, and the card gives the milliseconds at the air temperature (°C or °F): 20 m is 58.24 ms at 20 °C, about 2.9 ms per meter or 0.89 ms per foot. Handy for a delay tower or a front fill measured with a tape; Use as delay puts that time into the measurement as a starting point (with the test tone as the source, plus the computer's known latency), then Track or Apply fine-tune it.

System latency (test tone as the source): the delay then also contains the computer's own output and input latency, so that distance is not the distance to the mic — until you measure the latency once per output and input. In Tune options → System latency, patch a cable from the output into the mic input (the accurate way; holding the mic against the computer's speaker is rough) and tap Measure latency: the test tone plays, three agreeing reads give the latency, and it is remembered on this device for that output and input. From then on the card says Latency compensated and takes it off the distance.

Align two speakers (A / B) (in Tune options): apply the delay of one speaker (say the main) and tap Mark as A, then the other (the sub) and Mark as B. B − A is the delay to dial into the processor for the earlier one — then both arrive together. Works the same for front fills and delay towers. The marks stay on this device.

Levels: if either channel clips, the plot says which one (Mic (ch 1), Master (ch 2)) with Check level — Settings → Input level. There, Match sets the Master trim (dB) so both read alike; a trim moves both curves' levels, never the measurement.

The three panes

  • Magnitude (dB): how much louder or softer each frequency comes out than it went in. Normalize to 0 dB puts the average level (100 Hz – 10 kHz, where the coherence is good) at 0 dB.
  • Phase (−180° to +180°, after the delay compensation): a flat line around 0° means the frequencies arrive together. A steady slope means the delay is a little off — nudge it.
  • Coherence (0 to 1): how much of what the mic hears is the source. Close to 1 = trust the curve. Low = noise, reflections that arrive too late for the window, a wrong delay, or two clocks drifting apart. Where coherence is below the threshold you set (Fade below coherence), magnitude and phase are drawn faded (0 = off). Or tick Fade weak data: every part of the curve is as solid as its coherence — no threshold to choose.
  • Phase: Unwrapped draws it as one line instead of folding it back at ±180° (Wrapped) — a delay that is off then reads as a straight fall, steeper the further off.

Impulse response

Switch the plot from Transfer to Impulse to see the same measurement over time: the system's response to a click, from shortly before its peak to 80 ms after (Time range in Tune options: 20 to 500 ms; Scale Linear, or Log (dB) to see the decay and late reflections). Peak at 12.43 ms (say) is when the direct sound arrives — the delay to align to — and the reflections after it are the room. Polarity inverted means the peak points down. The time is always when the sound arrived, whether or not a delay is applied.

Point at (or tap) the plot for an exact readout: frequency, dB, degrees and coherence. Arrow keys move it when the plot has focus.

Tune options

Options above the plot opens Tune options — a side panel beside the plot, like Ring out's Spectrum options; the measurement keeps running while you change things. It holds every Transfer and Impulse setting:

  • FFT size 8k – 64k: longer windows resolve low frequencies better (64k at 48 kHz: 0.73 Hz bins, 1.4 s) but react more slowly. 32k is the default.
  • Averaging (N) and Averaging type: Exponential keeps a running average of about the last N blocks (new blocks count more); Linear (last N) is the plain mean of the last N blocks. Blocks overlap by half.
  • Smoothing 1/3 – 1/24 octave: magnitude is power-averaged across the band, phase is averaged as complex values, coherence is averaged.
  • Phase Wrapped / Unwrapped, Overlay (a saved snapshot over the live curve, or None), Fade below coherence, Normalize to 0 dB, Fade weak data — see The three panes.
  • Impulse: Scale and Time range.
  • System latency and Align two speakers — see Measuring.
  • Limits of a curve links to the section below.
  • Reset to defaults puts the analysis back; the delay, the latency and the A / B marks stay.

The channels are not in Tune options: they are chosen in Settings → Input.

How it is computed: both signals are cut into Hann-windowed blocks; the transfer function is the H1 estimate (cross-spectrum ÷ source auto-spectrum, averaged as complex values — noise at the mic averages out of it); coherence is |cross-spectrum|² ÷ (both auto-spectra). The delay finder is a phase-transform cross-correlation (GCC-PHAT), refined to a fraction of a sample.

Snapshots

Save snapshot (bottom of the left column) keeps the curve as it is drawn — magnitude, phase, coherence and the delay — with the input, the calibration and the master's name (Settings → Input) in Saved, the one panel every view opens (up to 50 snapshots, grouped by show). Draw or hide each one, rename it, add a note, delete it, or Average shown snapshots (two or more: magnitude power-averaged, phase as a weighted complex mean — useful for several mic positions). Difference draws A − B of two snapshots. Each snapshot exports as .txt / .frd / .csv for other analyzers. Signed in, snapshots are synced to your account; see Snapshots for everything Saved does.

Limits of a curve

  • Separate clocks drift. A USB mic plus a separate output, or two interfaces, each run on their own clock: the delay walks a few milliseconds a minute. With the master on another input (Settings), Tune follows it: every delay estimate feeds a straight line through the last 40 seconds, and each block of the master is read along that line, between the samples. The delay card says Two clocks: following the drift (+23 ppm); if the browser skips samples, the delay jumps — the card counts it and the average starts over. A hand-set delay stays as an offset to the line. With the test tone as the source and a USB mic, nothing follows the drift: keep those measurements short. One interface for the mic and the master (and the output) is still the most precise.
  • With the test tone as the source the delay includes the computer's own output and input latency — measure it once (above).
  • Bluetooth is unusable here: long, changing delay and its own processing.
  • The browser may resample the input or output to its own rate, which adds a little delay and can soften the top octave.
  • The browser decides how many input channels arrive: Chrome and Edge pass all inputs of an interface; Safari and Firefox often only one or two (see Settings).

Was this page helpful?