What Is a Logarithm?

The biggest mistake people make about logarithms is thinking of them as some new kind of number or operation separate from exponents. They aren't. A logarithm is just the exponent you need. When a student asks "what is a logarithm", the answer is short: log_b(x) = y means b^y = x. That equation is the entire definition. Everything else, the rules, the graphs, the real-world uses, follows from that one swap between exponential form and logarithmic form.

If you can read 10² = 100, you can read log₁₀(100) = 2. The logarithm tells you the exponent (2) that the base (10) needs to produce the argument (100). That is the only thing a logarithm is. The rest is practice converting, running through the domain rules, and learning what the result means when it is below zero, zero, or a fraction.

Every logarithm is three numbers in a fixed relationship:

• b = base (must be above zero, not equal to 1)
• x = argument (must be greater than 0)
• y = logarithm (the power, can be any real number)

Written as log_b(x) = y, it says b^y = x. That is the definition you use to solve every problem. If you have log₂(8), you ask "2 to what power gives 8?" The answer is 3 because 2³ = 8. If you have log₁₀(0.01), you ask "10 to what power gives 0.01?" The answer is, 2 because 10⁻² = 0.01. The logarithm is the power.

The definition does not change when the numbers get messy. log₁₀(3) ≈ 0.4771 because 10^0.4771 ≈ 3. The power is just not an integer. This is where most confusion lives: people accept that 10^0 is 1 and 10^1 is 10, but they struggle with the idea that a power can be 0.4771. It works exactly the same way as an integer power. The power rule and the change of base formula are built on accepting that fact.

Converting between the two forms is the single skill that unlocks every logarithm problem. Given log_b(x) = y, write b^y = x. Given b^y = x, write log_b(x) = y. That is it.

Try these conversions in your head, then check using the inverse property:

log₁₀(1,000) = 3 → 10³ = 1,000. Correct because 10 × 10 × 10 = 1,000.

log₂(1/8) =, 3 → 2⁻³ = 1/8. Correct because 2⁻³ = 1/2³ = 1/8.

log₅(25) = 2 → 5² = 25. Correct.

The failure case happens when the argument is not an exact power of the base. For log₁₀(50), the exponential form is 10^y = 50. There is no integer y that works. The answer is about 1.6990 because 10^1.6990 ≈ 50. The conversion still works, the power just is not a round number. Use the change of base formula on a calculator or look up the mantissa in a four-figure log table.

The logarithm definition has two hard rules you cannot break in real numbers.

Argument Must Be Above Zero

There is no real power that makes a positive base produce zero or a number below zero. 10^y = 0 has no solution. 10^y =, 5 has no solution. Trying log(0) or log(, 3) on a calculator returns an error. The same applies to any positive base. If your input is zero or below zero, you have made a mistake, check the problem.

Base Restrictions

Base must be above zero and not equal to 1 (b > 0, b ≠ 1). A base below zero creates problems because fractional powers can produce imaginary numbers. Base 1 is useless because 1^y = 1 for every y, so log₁(x) is undefined for any x that is not 1, and even then it is ambiguous. Stick to bases like 10, e, and 2. Those three cover nearly every real-world use.

The domain restriction is not a theoretical curiosity. It matters when you work with pH (the argument is hydrogen ion concentration, which is always above zero) and when you compute decibels (the power ratio is always above zero). If your data contains a zero or value below zero, the logarithm, and any scale built on it, cannot process it.

A logarithm below zero tells you the argument is between 0 and 1. That is the only thing it means. log₁₀(0.5) ≈, 0.3010 because 10⁻⁰·³⁰¹⁰ ≈ 0.5. The power is below zero, so the result in exponential form is a fraction.

This appears in three common places:

pH

pH =, log₁₀([H⁺]). A hydrogen ion concentration of 1.0 × 10⁻³ mol/L gives pH 3. The minus sign in the definition means a high concentration (large argument) produces a low pH. The logarithm itself is below zero before you apply the minus sign. The IUPAC definition of pH is pH =, log₁₀(a_H⁺), where a_H⁺ is hydrogen ion activity.

Decibels and Decay Models

Decibels. Sound pressure level in dB uses a logarithm of a power ratio. A ratio less than 1 produces a dB value below zero, meaning the sound is quieter than the reference level.

Decay models. Radioactive half-life calculations take the log of a fraction (the remaining amount divided by the original amount). The result is below zero, and multiplying by the half-life gives a positive time.

The failure case is forgetting the minus sign in pH. If you compute, log(0.001) and get 3, you have done it correctly. If you forget the minus sign, you get, 3, which is not a valid pH. Check your sign before you write the answer.

ln(1) = 0. log(1) = 0. log₂(1) = 0. Every logarithm of 1, in any valid base, equals 0. The reason is built into the definition: b⁰ = 1 for any b > 0, b ≠ 1. The power that produces 1 is always 0, so the logarithm is always 0.

This is not a coincidence or a special case. It is the direct consequence of the definition. When you see a logarithm result of 0, you know the argument is exactly 1. That is the neutral point on every logarithmic scale. On the Richter scale, a magnitude 0 earthquake corresponds to a specific reference amplitude. In decibels, 0 dB means the measured power equals the reference power. On the pH scale, pH 7 (log₁₀ of 10⁻⁷) is neutral, but the logarithm of 1 itself gives 0, not 7, pH uses the log of the concentration below zero, so the zero point of the log function does not correspond to neutral pH.

If you calculate a logarithm and get 0, check that your argument really equals 1. If it does not, something is wrong with the base or the computation.

Logarithms are not just a classroom exercise. Three real-world scales use them directly, and each one depends on the definition.

pH Scale

pH =, log₁₀([H⁺]). A hydrogen ion concentration of 1.0 × 10⁻⁷ mol/L gives pH 7.00. The IUPAC Green Book defines pH as an operational measure based on the log of hydrogen ion activity below zero in base 10. Practical accuracy is about ±0.01. The logarithm converts a range of concentrations spanning 14 orders of magnitude into a readable 0-14 scale.

Decibels and Earthquake Magnitude

Decibels (dB). Sound intensity level = 10 log₁₀(P / P₀), where P is the measured power and P₀ is the reference power. A factor of 10 in power is exactly 10 dB. A factor of 100 in power is 20 dB. The logarithm compresses the vast range of human hearing into a manageable scale.

Earthquake magnitude (Richter scale). Magnitude = log₁₀(A / A₀), where A is the wave amplitude and A₀ is a reference amplitude. Each whole-number step multiplies amplitude by 10. The USGS notes that the energy release per step is about 31.6× (commonly approximated as 32×), not 10×. This is the most common mistake people make about logarithms in the news. A magnitude 6 earthquake releases about 32 times the energy of a magnitude 5, not 10 times.

The failure case is using the wrong base. pH and the Richter scale use base 10. If you use ln (natural log) on these scales, every number shifts by a factor of 2.3026. On a calculator, the log key is base 10 and the ln key is base e. Use the right one.

The single most practical action you can take is this: when you see a logarithm, immediately write it in exponential form. That one step eliminates most mistakes.If you cannot write the exponential form, you do not understand the problem yet.

Check the domain first. Is x > 0? Is b > 0 and b ≠ 1? If either is violated, the logarithm is undefined in real numbers. Fix the input before proceeding.

Then ask: is the result an integer, a fraction, or below zero? An integer means x is an exact power of the base. A fraction means x is between 1 and b (if above zero) or between 0 and 1 (if below zero). This tells you immediately whether the scale is indicating growth, decay, or a reference point.

The thing that most often goes wrong here is skipping the domain check and trying to compute the log of an argument below zero or an argument of zero. That returns a value that is not a real number every time. Verify the input before you touch the calculator.