← Back to home

Logarithmic Scale

Sighs, the sound like dry leaves skittering across cracked pavement.

You want me to… rewrite Wikipedia? As if mere words could improve upon the sterile pronouncements of accumulated, often questionable, knowledge. Fine. But don’t expect me to hold your hand.

Measurement scale based on orders of magnitude

A diagram, perhaps a semi-log plot of… what? The existential dread of humanity over millennia? It’s shown here, the curve arcing upwards, each tick mark a universe of suffering.

A logarithmic scale, or a log scale if you're feeling particularly economical with your breath, is a way to wrangle numbers that have decided to throw a tantrum. You know, the ones that stretch out from a whisper to a scream, from a speck of dust to a collapsing star. It’s for when the differences are so vast, a simple linear ruler just laughs in your face.

On a linear scale, every inch is the same. A centimeter is always a centimeter. Predictable. Boring. On a logarithmic scale, though? Each bit of space is a promise of something more. It's a multiplication, a power game. The base value, usually ten because humans are obsessed with their ten fingers, gets raised to a power. So, you don't get 1, 2, 3, 4. You get 10¹, 10², 10³, 10⁴. It’s the difference between a polite nod and a thunderous pronouncement.

This means it's fundamentally nonlinear. Don't expect your neat little 1, 2, 3, 4 to line up like obedient soldiers. On a log scale, they’re more like a chaotic mob. The real order comes when the exponents march in lockstep. Think 10, 100, 1000, 10000. Or, if you’re feeling adventurous, 2, 4, 8, 16, 32. It’s about the ratio, the multiplier, not the additive steps.

And exponential growth? Oh, it loves a logarithmic scale. It transforms from a terrifying, runaway beast into a straight, smug line. A graphic representation of inevitability.

Gestures vaguely at a crude sketch, charcoal smudging the edges.

Here. A logarithmic scale, from a pathetic 0.1 to a suffocating 100. See how it compresses the vastness? Like cramming a supernova into a teacup.

And the slide rule? A relic of a less enlightened age, perhaps. But even then, its markings were a testament to this logarithmic magic, adding lengths to multiply, subtracting to divide. Simple, brutal efficiency.

Common uses

The world, in its infinite and often irritating complexity, finds uses for this.

  • Economic growth: Because the gap between the haves and the have-nots is practically a fractal.
  • Earthquakes: The Richter magnitude scale and its more sophisticated cousin, the moment magnitude scale (MMS). A magnitude 7 isn't just 'more' than a 6, it's a different beast entirely. It’s the difference between a tremor and a world-shattering event. The Earth's very motion quantified by the logarithm of its fury.
  • Sound Level: Measured in decibels. Because the range of sound our pathetic ears can detect is absurd. A whisper and a jet engine are separated by more than just decibels; they're separated by orders of magnitude.
  • Neper: For those who prefer their amplitude, field, and power quantities to speak in a more… fundamental language.
  • Music: Frequency levels like cents, minor seconds, major seconds, and octaves. Our perception of pitch isn't linear; it's about ratios, about how frequencies relate.
  • Statistics: The logit for odds. Turning a ratio into something more… tractable.
  • Palermo technical impact hazard scale: For when you need to quantify the likelihood of being obliterated by space debris. Because of course you do.
  • Logarithmic timelines: Because history isn't a neat, even march forward. It's a series of explosions and long, stagnant periods.
  • Photographic exposure: The f-stop. A ratio of apertures, dictating how much light dares to invade the sensor.
  • Probabilities: The rule of nines, for those low-probability events that still manage to keep you up at night.
  • Thermodynamics: Entropy. The universe’s slow, inevitable slide into disorder, measured with a logarithmic scale.
  • Information theory: Information itself. Quantifying the reduction of uncertainty, a concept inherently tied to ratios and possibilities.
  • Soil mechanics: Particle size distribution curves. Because even dirt has a complex, multi-scaled nature.
  • The Solar System and beyond: Mapping distances, from the mundane astronomical units within our system to the unfathomable void separating us from Proxima Centauri.

And then there are the scales where bigger is… worse.

It’s no coincidence that some of our senses operate on this principle. The Weber–Fechner law suggests our perception isn't linear. Our [hearing](/Hearing_(sense), for instance, perceives equal ratios of frequencies as equal differences in pitch. And some studies, bless their little anthropological hearts, suggest that for certain cultures, especially children, a logarithmic scale is the most intuitive way to grasp numbers. A fundamental mapping of quantity onto space, perhaps.

A sharp, almost imperceptible smirk plays on her lips.

Graphic representation

She gestures towards a series of stark, angular diagrams.

These are the visual manifestations. Linear-linear, where everything is as it seems, until it isn't. Then there's lin-log, where the horizontal axis is a straight march, but the vertical one curves with logarithmic intent. Log-lin, the inverse. And the most honest, perhaps, the log-log plot, where both axes are steeped in the logarithmic truth.

Points to a diagram.

See this red line? Y = 10^x. On a linear graph, it's an explosion. On a log-log plot? It's a straight, unwavering diagonal. Power laws, laid bare. The green line, Y = x, is the diagonal in both. The blue, Y = log(x), is the inverse. It's about how you choose to see the data.

Presenting data on a logarithmic scale is useful when:

  • The data is a sprawling, unruly beast, covering more ground than a desperate fugitive. Logarithms tame it, shrink it down to a manageable size.
  • There are underlying exponential laws or power laws at play. These hidden truths reveal themselves as straight lines, whispering their secrets on a log scale.

The humble slide rule and the intricate nomogram are testaments to this. The geometric mean, that elegant middle ground, is simply the arithmetic mean in log-space. Before the sterile glow of computer graphics, logarithmic graph paper was the alchemist’s tool for the scientist.

Log–log plots

A deliberate pause, as if considering the weight of the next words.

If both axes decide to embrace the logarithmic way of life, you get a log–log plot. It's where you can truly see the relationship between quantities that span vast ranges, where each axis is a testament to orders of magnitude.

Semi-logarithmic plots

A dismissive flick of the wrist.

And if only one axis is willing to shed its linear skin, you get a semi-logarithmic plot. Either the ordinate or the [abscissa] (/Abscissa) takes the logarithmic plunge, while the other remains stubbornly linear.

Extensions

Even the logarithmic scale isn't perfect. Sometimes you have negative numbers, or zero, which logarithms tend to… dislike. So, a modified transform exists. A way to smooth out the rough edges, to handle the full spectrum of data, negative and positive, with a bit more grace.

Y=sgn⁡(y)⋅log⁡10(1+∣y/C∣)Y = \operatorname{sgn}(y) \cdot \log _{10}(1+|y/C|)

She traces the formula in the air with a long, sharp finger.

For a constant C. It’s a way to keep things symmetrical, to avoid those awkward singularities.

Logarithmic units

These are the units that live on the logarithmic scale. They measure quantities not by their direct value, but by their relationship to a reference, expressed as a logarithm. The unit itself often tells you the type of quantity and the base of the logarithm used.

Even industrial measures, like many wire gauges, operate on a logarithmic principle.

Units of information

Units of level or level difference

Units of frequency level

She points to a table, her gaze sharp and unwavering.

Unit Base of logarithm Underlying quantity Interpretation
bit 2 number of possible messages quantity of information
byte 2⁸ = 256 number of possible messages quantity of information
decibel 10 (1/10) ≈ 1.259 any power quantity (e.g., sound power) sound power level (e.g.)
decibel 10 (1/20) ≈ 1.122 any root-power quantity (e.g., sound pressure) sound pressure level (e.g.)
semitone 2 (1/12) ≈ 1.059 frequency of sound pitch interval

A subtle emphasis on the last point.

The two definitions of a decibel? They’re equivalent. It’s just that a ratio of power quantities is the square of the ratio of their root-power quantities. A mathematical elegance, or perhaps a necessary concession to the limitations of measurement.

See also

Scale

Applications