Monitoring · Room Setup
Getting a 10-inch subwoofer to disappear behind a pair of 5-inch speakers — placement, level, polarity, and how to measure it with gear you already own.
Two workable topologies. You're running B — the one where switching the sub off for the neighbours leaves the A5+ playing on their own, full-range.
The A5+ output is variable — it follows the volume knob, so the sub tracks level instead of sitting at a fixed output. This is Audioengine's own recommended way to add a sub. It also gives you somewhere to put a passive RCA attenuator (−6 or −12 dB) if the 10s can't get quiet enough on its own knob.
The trade is that the A5+ run full-range with no high-pass relief — which is exactly what you want on the nights the sub is off.
Put the sub on a switched strip or a smart plug so "sub off" is one tap instead of reaching behind the cabinet. Power it down after the speakers and up before them to avoid a thump.
Kept here for reference. The sub's crossover relieves the A5+ of everything below 80 Hz, but the signal passes through the sub — so powering it down silences the whole rig, which rules it out for you.
If the feed went into the sub's LFE input, that's your whole problem. It runs about 10 dB hotter than the L/R inputs and bypasses the crossover entirely. Move to L/R.
Outputs 1–2 are the pair the Scarlett's front knob controls. Outputs 3–4 are electrically identical but sit outside the knob unless you press Alt, so the monitor path has to leave from 1–2.
Three fixes that cost nothing and routinely solve "the sub is way too powerful" before a single number gets taken.
The corner is where the 10s is going, so plan around it rather than fighting it. A corner couples the sub to every mode in the room at once: you get 6–12 dB of free output and no null caused by the sub's own position — but everything below about 100 Hz comes up together, and the modes get excited hard.
Four things follow:
−12 dB attenuator stops being optional.On chain B the A5+ knob moves the mains and the sub together, since the sub feed is taken after it. So it changes loudness, never balance. Same for the Scarlett.
That leaves two relative controls: the 10s level knob, and a passive RCA attenuator between the A5+ output and the sub input. If the sub is still too loud at its own minimum, a −12 dB pad (Harrison Labs, ~$20) is the fix — not a different knob.
Start at 80 Hz. If it still feels heavy, try 60 — most of the "overwhelming" sensation lives in the 80–120 Hz band, and handing that back to the A5+ tightens things considerably.
The H6 with the XYH-6 capsule is a far better bass mic than a phone, which is high-passed in hardware precisely where you need to look. It isn't calibrated, and that's fine: every number that matters here is a comparison between two takes from the same mic in the same spot.
The PAD 0/−20 switch belongs to inputs 1–4, not the capsule — ignore it.
Touching the H6 gain or the Scarlett's monitor knob between takes. The absolute level is unknown and doesn't matter; the differences are the measurement, and both knobs destroy them.
One file, one continuous recording, three segments. Only the sub changes between them — you never pick a different file, so there's nothing to mix up and no labelling to get right.
Play 00_reference_pink_fullband every time. Leave about four seconds of silence between segments; the gaps are how the segments get found again.
| Segment | Sub | What it answers |
|---|---|---|
| 1 | powered off | What the A5+ do alone — the reference. |
| 2 | on, at its working level, polarity 0° | How many dB the sub adds, and where. |
| 3 | on, polarity 180° | Which polarity adds at the crossover instead of cancelling. |
Because all three segments play identical material at one fixed gain, the comparison is direct — no correction, no assumptions. Segment 2 minus segment 1 is the subwoofer.
Set it on segment 2 — the loudest — for peaks around −18 dBFS. You're recording 24-bit, so conservative gain costs nothing and clipping ruins the take outright.
Send the single file, plus a photo of the sub's panel showing where the level knob sits — the correction comes back in dB, and the knob position is what turns dB into an amount of rotation.
An SPL meter app gets you most of the way in one sitting. Set it to C-weighting — A-weighting deliberately discards bass and makes the whole exercise meaningless — and Slow response. Phone at the listening position, ear height, propped rather than held.
Sub powered off. Play 00_reference_pink_fullband and use the Scarlett knob to land near 80 dB C. Mark that knob position — it's your reference listening level from now on.
Scarlett knob untouched. Play 01_level_pink_35-70Hz and bring the sub's level up until the app reads the same number as step 1. All four files sit at identical digital level, so matching readings approximates a flat handover.
Play 02_phase_pink_60-120Hz, note the reading, flip the polarity switch, note it again. Keep the louder position. This is the step a phone does well, because it's a before-and-after with the same mic.
Play 03_sweep_20-200Hz and listen for a lump or a dropout as it climbs.
Phone mics roll off in the deep bass, so step 2 under-reports and leaves the sub set slightly hot. Treat the result as a starting point, then trim down by ear.
Flip the sub in and out while a track you know cold is playing. Correct level is when the bass extends lower but the tonal balance doesn't change. If bringing the sub in makes the track sound more exciting, it's too loud — that's the trap, and it's how mixes come back bass-light everywhere else.
| What you hear | What it is | What to do |
|---|---|---|
| Dropout right around 80 Hz on the sweep | Sub and mains cancelling at the handover | Flip polarity. If neither position helps, you're sitting in a null — move the chair a foot and retry. |
| One note booms, its neighbours vanish | A room mode, not a level problem | Move the sub or the chair. Turning the sub down just makes everything quiet and uneven. EQ can cut the peak; it can't fill the null. |
| Loud everywhere below 100 Hz | Corner loading — expected, given the placement | Pad the feed and drop the sub's knob. This is a level problem, and it has a level answer. |
| Sub still too loud at its minimum | Too much signal arriving at the sub | Inline −12 dB RCA pad between the A5+ output and the sub input. Neither volume knob can fix this on chain B. |
| Character changes when the sub comes in | Crossover overlap or too much level | Drop the sub a notch, then try the crossover at 60 Hz. |
| Hum with nothing playing | The unbalanced RCA leg | Get everything on one outlet first. Ground-loop isolator only if that fails. |
A 10-inch sub is still oversized for near-field work with 5-inch speakers. If placement and rebalancing still leave it fighting you, its better home is the DJ rig — that's what a big sub is for — and the A5+ plus the M50x or the N5005 cover mixing on their own.
Two takes, a day apart, same method: one continuous recording, three segments, gain locked. Midband held within 0.6 dB across segments both times, which is what makes the comparison trustworthy — nothing moved but the bass.
| Band | Sub added, before | Sub adds now |
|---|---|---|
| 30–50 Hz | +37 dB | +22 dB |
| 50–63 Hz | +24 dB | +11 dB |
| 63–80 Hz | +16 dB | +3 dB |
| 80–100 Hz | +14 dB | +5 dB |
| 100–125 Hz | +11 dB | −3 dB |
The 40–80 Hz region now sits at +13 dB above midband against +11.5 dB for the 100–200 Hz region the room produces on its own. That gap was 16 dB a day ago; it's now under 2. The low end is balanced — the remaining lift across 120–200 Hz belongs to the room, and no amount of sub level will fix or cause it.
Polarity settled: 180°. At 0° the sub cancels against the mains around 110 Hz and digs a hole where the speakers already have one; at 180° it fills it. Measured across 55–175 Hz, 180° is both louder and 4 dB smoother end to end. That's where the switch already sits — leave it.
The pad is now optional. The volume knob at 9 o'clock did the whole job. Order the −12 dB attenuator only if you later want the knob back in the middle of its range for finer resolution.
Caveats: the H6's own bass rolloff means the true depth below 40 Hz is somewhat better than plotted, and a stray thump 14 seconds into the second segment was excised before analysis. Every comparison is between takes from the same mic in the same spot, so the differences hold regardless.
All four are 60 seconds, 48 kHz, generated at an identical −20 dBFS RMS — which is what makes comparing readings between them meaningful.
The H6 also runs as a USB interface, so it can feed REW on the EVO-X2 directly. That turns the record-and-send loop into a live plot that updates as you move the sub a foot — more setup, much faster iteration.