In a recent May 2025 video, Dave Rat highlighted a puzzling metering mystery. Inside his digital mixing console, he summed two internally‑generated white noise signals — first fully correlated, then de‑correlated.

According to theory, the peak level should rise by +6 dB for correlated signals and only about +3 dB for uncorrelated ones. Yet in both cases, the console’s peak meter showed the same +6 dB increase.

Meanwhile, his external analog meters behaved as expected, clearly reflecting the 3 dB difference — and when that analog sum was routed back into the digital console, its meter tracked correctly too. So why did the initial test, conducted entirely within the digital domain, fail?

Commenters offered explanations ranging from metering type differences and DSP summing shortcuts to peak probability, meter ballistics, and filtering effects, yet none questioned how the white noise itself is generated.

TL;DR

Other than being a digital peak meter, there is nothing wrong with it.

The meters disagree because — in the digital domain — uniformly distributed white noises are summed. Whereas, after the signals leave the console and pass through D‑to‑A and/or A‑to‑D converters, they became more Gaussian‑like — and those distributions don’t sum the same way.

Don’t rely on peak meters to judge loudness — use them to watch for clipping.

Not All White Noises Are Equal

White noise can be generated in different ways — and those methods matter. For example, you might create it from a sequence of coin flips (a Bernoulli process), by rolling dice (a Uniform distribution), or by sampling from a Gaussian (Normal) distribution.

Each approach produces a signal with different statistical characteristics — specifically, their crest factor — even though they all sound equally “white” to both you and me, as you’ll hear in the following video.





Bernoulli white noise has a crest factor of 1 (or 0 dB), uniform white noise has a crest factor of √3 (approximately 4,8 dB), and Gaussian white noise — typically — exhibits a crest factor around 4 (about 12 dB). The latter is explained shortly.

Click here to download the audio files.

Gaussian Noise

The crest factor of Gaussian random noise is theoretically infinite, as the Gaussian distribution is unbounded. In practice, however, extreme values are rare. For example, in white Gaussian noise generated at 48 kHz:

• a sample exceeding +12 dB above the RMS level is expected about once every 300 milliseconds;
  
  
• a +15 dB event may occur once every 18 minutes;
  
  
• and a +18 dB peak might appear only once every 335 years.

Because it’s impractical to design around these extremely rare events, the audio industry has reached a general consensus: treating Gaussian noise as having a 12 dB crest factor. While there are some exceptions and differing practices, most follow this guideline — a convention I’ll adopt for the remainder of this article.

Figure 1Figure 1 shows the probability distribution functions (PDFs) for Gaussian noise. Note the characteristic bell-shaped curve.

When you sum two Gaussian noise signals, the result is Gaussian as well — and no crest factors are harmed in the making of this summation.

When summing — correlated — Gaussians, both RMS and peak levels rise by +6 dB; for — uncorrelated — Gaussians, the increase in both RMS and peak is only +3 dB. In either case, the 12 dB crest factor remains functionaly the same.

What about the — white noise — generators in our consoles and DAWs. Which distribution to they use?

Uniform Noise

The easiest way to generate white noise is to use a random number generator that produces values uniformly distributed between two equal but opposite extremes, where each value is equally likely.

It’s reminiscent of rolling a die, where each number from 1 to 6 has an equal chance of appearing — except that our “die” has significantly more faces.

Unlike Gaussian noise, which has a high crest factor due to its rare outliers, this signal maintains a much lower crest factor — just under 5 dB. So what happens if we sum these together?

Figure 2When summing — correlated — uniform distributions (Figure 2), both RMS and peak levels once again rise by +6 dB — and, once again, no crest factors were harmed in the making of this summation.

However, when summing — uncorrelated — uniform distributions, the resulting distribution changes from uniform to triangular.

Figure 3Reminiscent of the board game Catan (Settlers of Catan), where you roll two dice and take the sum (Figure 3). While each die on its own follows a uniform distribution, their combined total does not.

It's much rarer to roll a 2 or a 12 than it is to roll a 7, simply because there are fewer possible combinations that produce the extremes — and many more that add up to values near the center.

Triangular Noise

The crest factor of a triangular distribution is 3 dB higher than that of a uniform distribution — increasing from just under 5 dB to just under 8 dB. This is because — just as with uncorrelated Gaussians — the sum's RMS level increases by +3 dB. However, unlike the Gaussian case, there is a small but non‑zero chance that the peak level approaches +6 dB instead of just +3 dB.

Triangular noise exhibits more frequent near‑maximum peaks. At a 48 kHz sampling rate:

• a sample typically falls within 0,5 dB of the distribution’s extremes once every 7 milliseconds, and;
  
  
• within 0,1 dB approximately every 160 milliseconds.

As a result, the peak‑hold feature of a meter refreshing at, for example, 50 Hz, is very likely to capture a near +6 dB peak during each 20‑millisecond interval.

Therefore, when summing uniform noises, peak levels tend toward +6 dB, regardless of whether the signals are correlated or not. This is evident from the PDFs for the sums in Figure 2, whose bounds remain unchanged.

So why can’t this be reconciled with analog meters?

Get Ready To Be Converted

Figure 4Figure 4 shows the signal flow in Dave's setup.

When sending noise — generated internally — from a digital console to an analog console, the signal must pass through a Digital-to-Analog Converter (DAC).

Contemporary DACs typically employ oversampling to reduce the burden on the reconstruction filter. This involves an interpolation process that inserts new samples at a higher rate, effectively smoothing the signal.

As a result, the noise distribution changes: what began as a uniform or triangular distribution — the latter already somewhat bell‑shaped — becomes increasingly Gaussian‑like, due to the low‑pass filtering inherent in the oversampling process.

White noise is more susceptible to this transformation than pink noise, since its top octave carries half the signal’s power — compared to just one‑tenth in pink noise.

So, while one was summing uniform distributions in the digital domain, one was — perhaps unknowingly — summing Gaussian‑like distributions in the analog domain. And for Gaussians, as pointed out earlier, both RMS and peak levels rise in lockstep.

In contrast, in the digital domain with unaltered uniform distributions, only the RMS level increases when switching from uncorrelated to correlated signals — a change that the peak meter won't reflect, since the peak levels remain functionally constant.

The meter on Channel 1 of the digital console, connected to the analog console’s mix bus, will mimic the analog meter — as the signal has already undergone the transformation that affects the noise's distribution and subsequent peak behavior.

Counterexample

Repeat the entire experiment — but this time, instead of using the digital console’s internal (uniform) white noise generator, load two uncorrelated Gaussian white noise signals. I’m quite confident that even the digital console’s peak meter — as well as the external analog meters — will now reflect whether one is summing correlated or uncorrelated signals.

Acknowledgements

I’d like to thank Dave Rat for indulging me when I shared this observation with him. True to form, Dave raised plenty of follow‑up questions and concerns — given the prevalence of peak meters on digital consoles — about the practical implications of using them to monitor signal levels in a way that aligns with how humans perceive loudness.

My answer was simple: don’t rely on peak meters to judge loudness — use them to watch for clipping.

Dave’s video, which inspired this article, can be found below:





Files