Help: Why use a 20k Sweep for a Sub, and not just 20-80Hz?

[stero01]

Hi everyone,

I need a quick sanity check on my 2.1 DSP integration. The "correct" method I've been given seems counter-intuitive and physically dangerous for my sub.

My Hardware & Goal:

• Amp: Arylic B50 SE (DSP)
• Speakers: 2x KEF IQ3
• Sub: 1x KEF SW 2100 (Active)
• Goal: Correctly set Crossover, Level, Time Delay, and PEQ in the DSP (ACP Workbench). Measurement with REW.

---

My Problem: Is this method (as taught me) really the right way?​

Method A: The "Pro" Method (as taught to me)

1. Level: Level KEFs AND Sub to 69 dB using 20-20k Pink Noise.
2. Problem: This forces my sub's physical gain knob to 75% (from its usual 25%).
3. Measure: Use a full 20-20k Sweep.
4. Violent Result: The sweep is dangerously loud (100+ dB) in the sub's 20-140 Hz range, rattling my room. I'm afraid it will damage the driver.
5. Fix: Apply a massive -21 dB digital cut in the DSP, to finally arrive the 69 dB.

Method B: My "Logical/Safe" Method (is this the better way?)

1. Level/Measure: Use only 20-80 Hz Pink Noise and Sweep for the sub.
2. Result: I keep the physical gain at a safe 25%. Levels match, requiring 0 dB digital cut.

---

The Core Question​

I was told my "Logical Method B" is wrong and will fail, specifically because:

1. Time Alignment: REW needs the full 20k-Sweep data from all speakers and the subwoofer to compare their Impulse Response and find the correct Time Delay/Phase.
2. Digital Headroom: Method A (-20 dB cut) is the correct "Gain Staging" to prevent digital clipping from future PEQ boosts, while Method B (0 dB cut) is dangerous. The subwoofer has to stay on 75% gain.

Are these two points true? Is the seemingly dangerous "Method A" really the only correct way to get this right?

Thanks!

 

[Keith_W]

Something has gone awry with your volume matching. If you use an SPL meter to volume match your sub to your main speakers with pink noise from the listening position, AND you use the same SPL meter to calibrate REW, then both the sweep and the pink noise will be at 69dB. You should certainly NOT be measuring speakers at SPL's so loud that you might damage your speaker. Not to mention, it won't be representative of reality.

 

[Lid9497]

My method:
Use pink periodic noise for subwoofer(s) - Sub cal (30-80hz) to target dB (I use 75dB), use Speaker Cal (500-2000hz) for mains to target dB.

It is better to measure and equalize each speaker separately (for mains it's for preserving stereo imaging). I discovered measuring subwoofer from 20-20kHz produced bad results, 20-500Hhz is enough, especially if you want to use alignment tool later for mains - subwoofer time alignment. I measure mains 5Hz below their -+3db advertised response (I have -+3db at 45Hz, I measure from 40-20kHz).

I use RTA for mains, and sweep at 4 positions around listening spot for subwoofer. I use dual subs, so I use MSO after. For single sub, you can average the 4 positions measured.

Once you equalize all speakers, remeasure (mains separately first to verify EQ, then together to verify response of them both playing together) with new EQ filters applied with acoustic time reference (again, you need to measure subwoofer and speakers (mains together now) separately), then use alignment tool to find ideal response.

I checked your DSP Arylic B50 SE, quickly skimmed the video demonstration and I couldn't see any sort of subwoofer control. If you want to time align subwoofer and main speakers, you will need to control each channel individually, because you only want to delay mains.

Also check my thread, I just finished tweaking my 2.2 setup, and was facing similar challenges like you, people gave me really good advice there.

 

[FargateOne]

To level match sub and speaker I use the second method here

Speaker reference level calibration