ARCHETYPE: Beat-Frequency & Tuning Problems

You already know the headline of the previous article: two tones of nearly equal frequency beat at the plain difference, $f_{\text{beat}} = |f_1 - f_2|$ . That fact is beautiful and it is almost useless in an exam — because the questions run the formula backwards. They tell you the beat count and one known frequency, and ask for the unknown. And there the trouble starts, because the difference $|f_1 - f_2|$ has forgotten its sign. A $256\ \mathrm{Hz}$ fork beating four times a second with a wire tells you the wire is $4\ \mathrm{Hz}$ away — but $4\ \mathrm{Hz}$ above or $4\ \mathrm{Hz}$ below? The beat count alone cannot say, and every year students write one answer where the marking scheme demands two, then throw away the follow-up mark by guessing the wrong one. This article is about the single skill that resolves it: deliberately nudging one source and watching which way the beats move. Load a tuning fork with a blob of wax and its pitch drops; file metal off its prongs and its pitch rises. Do one of those, watch whether the throbbing quickens or slows, and the ambiguity collapses to a certainty. That "watch the beats respond" logic is the whole archetype, and once you own it, an entire family of CBSE, ICSE, JEE and NEET problems — forks against forks, forks against wires, three forks in a chain, a wire whose tension you tweak — becomes one reflex applied over and over. So we treat the setup as the concept. The master recipe first, then the loading and filing rules that break the tie, then four worked examples escalating from a two-line board question to a three-fork triangulation, and finally the traps examiners reuse every single year.

Key takeaways

A beat count $n$ against a known frequency $f_k$ gives the unknown as $f_u = f_k \pm n$ — two candidate answers, never one.
The $\pm$ is resolved by changing one source and watching the beats: loading (adding mass) lowers a frequency; filing (removing mass) raises it.
If a change that lowers the unknown makes the beats speed up, the unknown was already below the known source; if they slow down, it was above.
Sources chain: a wire's frequency obeys $f \propto \sqrt{T}$ , so a small tension change $\tfrac{\Delta T}{T}$ shifts the frequency by about half as much, $\tfrac{\Delta f}{f} \approx \tfrac12\tfrac{\Delta T}{T}$ — the beat follows.
Three forks in a line are pinned by two beat counts read consistently from both ends.
A "no-beat" reading means the two sources are equal in frequency (the tuner's goal), not that one is silent.
Key idea: turn a beat count into $f_k \pm n$ , then break the $\pm$ by nudging one source with wax or a file and reading which way the beats move — the direction of the change, not the count, is the answer.

The core insight

Every beat-tuning problem is $f_u = f_k \pm n$ followed by a tie-breaker: because loading a source lowers its frequency and filing raises it, watching whether the beat rate rises or falls after a known nudge tells you which side of $f_k$ the unknown lies on, converting an ambiguous count into a single frequency.

Where this lives today

The loading-and-filing trick is a museum piece, but the logic it teaches — change something you control and watch how the beats respond — runs modern measurement. A piano technician still tunes entirely by beats, counting the throb between a string and a fork and against other strings, driving each to the target beat rate rather than to zero. Beyond music, the same "nudge and watch" idea is how a laser is locked to an atomic transition: the light is deliberately dithered and the beat against the atoms tells the servo which way to push. Radio and radar work the difference frequency directly — a Doppler speed gun reads a car's speed from the beat between emitted and reflected microwaves, and shifting the emitted frequency resolves any ambiguity exactly as filing a fork does. Even the comparison of the world's atomic clocks is a beat measurement: one clock is beaten against another and the slow difference is counted. Every one of these is this article's reflex — a known nudge, a beat that moves, a sign resolved. The next chapter opens the nature of the sound wave these forks and wires launch into the air, and later the Doppler effect turns the moving-source beat into its own topic.

The setup

In an archetype the bookkeeping is the physics, so start with exactly what a beat measurement gives you and what it withholds. Sound a source of unknown frequency $f_u$ together with a reference of known frequency $f_k$ — a calibrated tuning fork, a piano's concert-A string, an electronic oscillator. You hear the loudness swell and fade, and you count the swells: $n$ per second. From the previous article, that count is the difference of the two frequencies, and nothing else:

f_{\text{beat}} = |f_u - f_k| = n. \qquad (1)

where: $f_k$ is the frequency you trust; $f_u$ is the one you want; $n$ is the number of beats per second, always a positive count. The absolute-value signs are the entire difficulty. They mean the measurement is blind to order: a source $n$ hertz sharp and a source $n$ hertz flat produce identically many beats a second against the same reference. Equation (1) therefore has two solutions,

f_u = f_k + n \qquad \text{or} \qquad f_u = f_k - n,

and a beat count on its own can never choose between them. This is not a gap in your technique; it is a genuine limit of the information a beat carries. To break it you must feed in one more fact — and the cheapest fact available is what happens when you deliberately disturb one of the sources.

Figure

A known fork beats with an unknown source four times a second and loading the source lowers its frequency, so whether the beats then speed up or slow down reveals which side of the fork the unknown sits on

The two answers, and the tie-breaker

The disturbance you use is mechanical and its direction is fixed by physics, not choice. A tuning fork sings at a frequency set by the mass and stiffness of its prongs. Add mass — a dab of wax, a small rider, a clamped bead — and the heavier prongs swing more sluggishly: the frequency drops. Remove mass — file metal off the prong tips — and the lighter prongs swing faster: the frequency rises. These two moves, loading and filing, are the levers of the whole subject:

\text{load (add mass)} \Rightarrow f \downarrow, \qquad \text{file tips (remove mass)} \Rightarrow f \uparrow. \qquad (2)

A stretched wire has the same two levers wearing different clothes. Its frequency is $f = \tfrac{1}{2L}\sqrt{T/\mu}$ , so increasing the tension raises the pitch and adding to the wire's linear mass density lowers it — the same up/down logic, since $f \propto \sqrt{T}$ and $f \propto 1/\sqrt{\mu}$ . Whatever the source, you always have one knob that you know moves the frequency in a known direction. Now put the lever to work. Suppose you have narrowed the unknown to $f_k + n$ or $f_k - n$ , and you apply a change that you know lowers the unknown. Two things can happen. If the beats speed up, the gap between unknown and reference grew as the unknown fell — which is only possible if the unknown was already below the reference, running away downward. If the beats slow down, the gap shrank as the unknown fell — so the unknown was above the reference, moving down toward it. One observation, one bit of information, and the $\pm$ is gone.

Figure

The beat count alone leaves two candidates straddling the reference, one sharp and one flat, and only a deliberate change to one source picks between them

The master recipe

Every problem in this family is the same five moves. Step 1. Write the beat equation $f_u = f_k \pm n$ and commit to both answers on paper — losing the second candidate is the commonest way to drop a mark. Step 2. Identify the change the question makes to a source (wax, file, extra tension, a second wire) and use equation (2) to state, with certainty, whether that change raises or lowers the affected frequency. Step 3. Read the reported effect on the beats — faster or slower — and match it to the direction: a change that moves the unknown away from the reference speeds the beats, a change that moves it toward the reference slows them. Step 4. Pick the surviving candidate. Step 5. If the source is a wire or a pipe, chain in its own formula to answer a further part — for a wire that means $f \propto \sqrt{T}$ . That last link is worth making quantitative, because it powers a whole sub-genre of "change the tension, find the beat" questions. Since $f \propto \sqrt{T}$ at fixed length and mass density, a small fractional change in tension produces about half that fractional change in frequency:

\frac{\Delta f}{f} \approx \frac{1}{2}\,\frac{\Delta T}{T}. \qquad (3)

where: $\Delta T/T$ is the fractional change in tension and $\Delta f/f$ the resulting fractional change in frequency. The one-half is the fingerprint of the square root, and it is the single fact examiners test hardest: raise a wire's tension by $4\%$ and its pitch rises by only about $2\%$ , not $4\%$ . Equation (3) is an approximation for small changes; when the change is not small you use the exact $f_2 = f_1\sqrt{T_2/T_1}$ , and we will do exactly that below.

Derivation AResolving the sign — why the direction of the beats decides

The goal is to prove, cleanly, that watching the beats after a known nudge always resolves the $\pm$ in $f_u = f_k \pm n$ — and to derive the working rule "beats faster means the unknown moved away, beats slower means it moved toward." The strategy is to write the beat count as a distance on the frequency axis and ask what a small, known displacement of the unknown does to that distance. Everything follows from one fact: the beat count is $|f_u - f_k|$ , the gap between the two frequencies, so anything that changes the gap changes the beats.

Step 1. Fix the reference at $f_k$ and place the unknown at one of its two candidates, $f_u = f_k + n$ (sharp) or $f_u = f_k - n$ (flat). In both cases the current beat count is the gap, $|f_u - f_k| = n$ .

Step 2. Apply a nudge whose direction you know — say one that lowers $f_u$ (loading). If the unknown was sharp, $f_u = f_k + n$ , lowering it moves it toward $f_k$ , shrinking the gap: the beats slow. If the unknown was flat, $f_u = f_k - n$ , lowering it moves it away from $f_k$ , growing the gap: the beats quicken.

Step 3. Read the implication both ways. A lowering nudge that quickens the beats can only have come from a flat unknown, so $f_u = f_k - n$ ; a lowering nudge that slows them came from a sharp unknown, so $f_u = f_k + n$ . For a raising nudge (filing, tightening) every arrow reverses: quicker beats now mean the unknown was already sharp.

Step 4. Compress it to one sentence you can carry into an exam: a change that pushes the unknown away from the reference speeds the beats; a change that pushes it toward the reference slows them. The beat count gave the size of the gap; this single observation gives its sign, and the ambiguity is gone. Verified: the resolutions 252 Hz (loading lowers, beats 4 to 6, gap grows, unknown flat) and 345 Hz (filing raises, beats 5 to 7, gap grows, unknown sharp) are recomputed in verify_hs2-beats-problems.py.

Go deeper · Derivation B · 1 minChaining a tension change into a beat

Derivation BChaining a tension change into a beat

Here we turn "the tension changed by a few per cent — how many beats?" into a two-line calculation, and derive the small-change rule $\Delta f/f \approx \tfrac12\,\Delta T/T$ that lets you predict the answer before touching a calculator. The goal is the beat between a wire at tension $T$ and an identical wire at $T(1+x)$ ; the strategy is to get the shifted frequency from $f \propto \sqrt{T}$ and then take the plain difference, checking the exact route against the linear one.

Step 1. Start from the wire law $f = \tfrac{1}{2L}\sqrt{T/\mu}$ . At fixed length and mass density, only the tension changes, so $f \propto \sqrt{T}$ and the ratio of the two frequencies is $f_2/f_1 = \sqrt{T_2/T_1}$ .

Step 2. Put in the numbers for a $4\%$ rise, $T_2/T_1 = 1.04$ : $f_2 = f_1\sqrt{1.04} = 256 \times 1.0198 = 261.07\ \mathrm{Hz}$ . The beat against the untouched wire is $f_2 - f_1 = 261.07 - 256 = 5.07\ \mathrm{Hz}$ .

Step 3. Now the shortcut. For a small fractional change $x = \Delta T/T$ , the square root satisfies $\sqrt{1+x} \approx 1 + \tfrac{x}{2}$ , so $\Delta f/f \approx \tfrac12 x$ . A $4\%$ tension rise therefore lifts the frequency by about $2\%$ , and $2\%$ of $256\ \mathrm{Hz}$ is about $5.1\ \mathrm{Hz}$ — matching the exact $5.07\ \mathrm{Hz}$ .

Step 4. Note the trap the algebra exposes. Scaling the frequency by the full $4\%$ would predict a $\sim 10\ \mathrm{Hz}$ beat, double the truth. The square root is not decoration; it halves the fractional change, and the whole difficulty of these problems is remembering that it is there. Verified: $f_2 = 256\sqrt{1.04} = 261.07\ \mathrm{Hz}$ , beat $5.07\ \mathrm{Hz}$ , the one-half rule giving $2\%$ and $\approx 5.1\ \mathrm{Hz}$ , and the wrong full- $4\%$ route giving $\approx 10\ \mathrm{Hz}$ , are recomputed in verify_hs2-beats-problems.py.

Limiting cases

case	what you observe	what it means
beats vanish, $n = 0$	steady, un-throbbing tone	the two sources are equal — the tuner's target, not silence
load the unknown, beats speed up	gap grows as $f_u$ falls	$f_u$ was below $f_k$ : take $f_u = f_k - n$
load the unknown, beats slow down	gap shrinks as $f_u$ falls	$f_u$ was above $f_k$ : take $f_u = f_k + n$
file the unknown, beats speed up	gap grows as $f_u$ rises	$f_u$ was above $f_k$ : take $f_u = f_k + n$
tension up by small $\Delta T/T$	pitch up by about $\tfrac12\Delta T/T$	the square-root law halves the fractional change
beat count large, tens of Hz	no countable throb	too far apart to beat — two separate tones, formula does not apply

Common trapA beat count is one number with two answers — never stop at the count

The single most expensive habit in this topic is writing $f_u = f_k \pm n$ and then quietly keeping only the value the question "seems" to want. Both $f_k + n$ and $f_k - n$ are legitimate until you break the tie, and a marking scheme awards the tie-breaking reasoning, not the lucky guess. The reasoning is always the same shape: name the change, use loading-lowers or filing-raises to fix its direction, then read whether the beats grew or shrank. A subtler cousin of the trap: students memorise "loading lowers frequency" and then forget it applies to the source they loaded, not the other one — if you wax the reference fork instead of the unknown, every conclusion flips. Always be explicit about which source you disturbed. And remember the count is a magnitude: " $4$ beats" is never $-4$ , so the sign lives entirely in the direction the beats move, never in the number itself.

Worked examples

1 · Board level — a fork against a sonometer wire (CBSE and ICSE pattern, 3 marks)

A tuning fork of frequency $256\ \mathrm{Hz}$ is sounded with a sonometer wire and $4$ beats per second are heard. A small mass is then loaded onto the vibrating wire, and the beat rate rises to $6$ per second. Find the frequency of the wire.

Step 1. Beat equation: the wire differs from the fork by $4\ \mathrm{Hz}$ , so it is $256 + 4 = 260\ \mathrm{Hz}$ or $256 - 4 = 252\ \mathrm{Hz}$ . Write both — the tie-breaker decides.

Step 2. Fix the direction of the change. Loading the wire adds to its mass per unit length, and since $f \propto 1/\sqrt{\mu}$ , its frequency falls.

Step 3. Read the effect. The beats rose from $4$ to $6\ \mathrm{Hz}$ : lowering the wire's frequency made the gap to the fork wider. A gap that widens as the wire drops is only possible if the wire was already below the fork, running further away downward.

Step 4. Conclude: the wire was the lower candidate, $\boxed{f_{\text{wire}} = 252\ \mathrm{Hz}}$ . Sanity check: at $252\ \mathrm{Hz}$ the wire is $4\ \mathrm{Hz}$ under $256$ ; drop it a couple of hertz and the gap becomes $6\ \mathrm{Hz}$ , exactly as heard. Had the wire been $260\ \mathrm{Hz}$ , loading it would have moved it toward $256$ and the beats would have slowed.

2 · Board level — two forks, resolved by filing (CBSE and NEET pattern)

Fork A has frequency $340\ \mathrm{Hz}$ . Sounded with an unknown fork B, it gives $5$ beats per second. When the prongs of B are filed slightly, the beat rate rises to $7$ per second. Find the frequency of B.

Step 1. Beat equation: B is $340 \pm 5$ , that is $345\ \mathrm{Hz}$ or $335\ \mathrm{Hz}$ .

Step 2. Fix the direction. Filing metal off B's prongs removes mass, so B's frequency rises (equation 2).

Step 3. Read the effect. The beats rose from $5$ to $7\ \mathrm{Hz}$ : raising B's frequency made the gap to A wider. A gap that widens as B climbs means B was already above A, moving still further away.

Step 4. Conclude: $\boxed{f_B = 345\ \mathrm{Hz}}$ . Check the logic against the other candidate: had B been $335\ \mathrm{Hz}$ (below A), filing it upward would have carried it toward $340$ and the beats would have slowed to about $3\ \mathrm{Hz}$ — the opposite of what was heard. Notice the mirror image of Example 1: there a downward nudge grew the beats and proved the source low; here an upward nudge grows the beats and proves the source high. Same rule, read in the direction of the actual change.

3 · Competitive level — a tension change turned into a beat (JEE-Main and NEET pattern)

Two identical sonometer wires are tuned to the same frequency, $256\ \mathrm{Hz}$ . The tension in one of them is then increased by $4\%$ while the other is left alone, and the two are sounded together. How many beats per second are heard?

Step 1. Recognise the chain: change tension, get a new frequency, then beat it against the untouched wire. The untouched wire stays at $f_1 = 256\ \mathrm{Hz}$ .

Step 2. New frequency of the tightened wire, using the exact square-root law $f_2 = f_1\sqrt{T_2/T_1}$ with $T_2/T_1 = 1.04$ : $f_2 = 256\sqrt{1.04} = 256 \times 1.0198 = 261.07\ \mathrm{Hz}$ , which we quote as $\boxed{261.1\ \mathrm{Hz}}$ .

Step 3. Beat frequency is the difference: $f_{\text{beat}} = |261.07 - 256| = \boxed{5.07\ \mathrm{Hz}}$ , about five beats a second.

Step 4. Sense-check with equation (3), the fast route worth carrying into an exam: a $4\%$ tension rise gives roughly a $2\%$ frequency rise, and $2\%$ of $256$ is about $5.1\ \mathrm{Hz}$ — matching the exact answer and confirming you have not accidentally used the full $4\%$ (which would wrongly give about $10\ \mathrm{Hz}$ ). The square root has quietly halved the change. Timing note: about 90 seconds. The instant you see "tension changed by a few per cent, find the beats," reach for $\Delta f/f \approx \tfrac12\Delta T/T$ to predict the answer, then confirm with the exact $\sqrt{T_2/T_1}$ . The trap that eats the mark is forgetting the square root and reporting a $10\ \mathrm{Hz}$ beat.

4 · Competitive level — three forks in a chain (JEE and NEET pattern)

Three tuning forks A, B, C are arranged so that A has frequency $300\ \mathrm{Hz}$ and C has frequency $305\ \mathrm{Hz}$ . A sounded with B gives $3$ beats per second; B sounded with C gives $2$ beats per second. Find the frequency of B, given that B lies between A and C.

Step 1. Write both beat equations. From A and B: $f_B = 300 \pm 3$ , so $303$ or $297\ \mathrm{Hz}$ . From B and C: $f_B = 305 \pm 2$ , so $307$ or $303\ \mathrm{Hz}$ .

Step 2. Find the value consistent with both counts. The two lists are $\{297, 303\}$ and $\{303, 307\}$ ; the only frequency appearing in both is $303\ \mathrm{Hz}$ .

Step 3. Confirm against the stated ordering. $f_B = 303\ \mathrm{Hz}$ sits between A ( $300$ ) and C ( $305$ ), reads as $300 + 3$ from below and $305 - 2$ from above, and reproduces both beat counts exactly. So $\boxed{f_B = 303\ \mathrm{Hz}}$ .

Step 4. See why triangulation replaces loading here: two references, read from both sides, pin the middle fork without ever waxing or filing anything. The consistency requirement — one number must satisfy every beat count at once — is itself the tie-breaker. Timing note: under two minutes. List the candidates from each pair and look for the overlap; do not try to reason out signs one fork at a time. When a problem hands you two beat counts against two known frequencies, the intersection of the two $\pm$ lists is the answer, and any extra clue (like "B lies between A and C") is there only to confirm it.

Exam traps

Stopping at $f_k \pm n$ . The beat count gives two frequencies and a board answer must show both, then resolve. Writing a single number — even the correct one — throws away the reasoning mark, and guessing throws away the answer mark half the time. Always carry both candidates to the tie-breaker.

Loading versus filing, and which source you touched. Loading (adding mass) lowers a frequency; filing the prong tips (removing mass) raises it. The conclusion depends on applying the rule to the source you actually disturbed. Wax the wrong fork and every "above/below" flips. State plainly: "I loaded the unknown, which lowers its frequency."

Reading the beat direction backwards. A change that pushes the unknown away from the reference speeds the beats up; a change that pushes it toward the reference slows them down. Memorise it as "away is faster, toward is slower," and check it against Example 1 every time you doubt it.

Forgetting the square root in tension problems. $f \propto \sqrt{T}$ , so a $4\%$ tension rise is only a $\sim 2\%$ frequency rise (equation 3). Using the full $\Delta T/T$ doubles the beat and is a favourite distractor: for the $256\ \mathrm{Hz}$ wire the wrong route gives about $10\ \mathrm{Hz}$ , the right route about $5\ \mathrm{Hz}$ .

Applying the beat idea to far-apart tones. The whole picture needs the two frequencies within a few hertz. If a problem's numbers differ by tens or hundreds of hertz there is no perceptible beat, and "the beat frequency is $60\ \mathrm{Hz}$ " is not a physical answer — the ear hears two distinct tones instead.

Board scope. These beat-and-tuning drills are India-facing: they are standard on CBSE, ICSE, JEE-Main, JEE-Advanced and NEET, where the sonometer and the tuning fork are lab staples. They are not an assessed objective on IB or AP Physics 2, and IGCSE 0625 does not treat beats at all. If you sit one of those boards, read this for the physics of superposition, not as exam drill.

Origins · 1 minNorbert Wiener and how this was discovered

Norbert Wiener

1894–1964 · Cambridge, Mass.

Norbert Wiener was an American mathematician — a child prodigy who took his Harvard doctorate at eighteen and spent his career at MIT — and his deepest work is, in effect, the theory of what "an unknown frequency, found by beats" really means. In Generalized Harmonic Analysis (1930) he built the rigorous mathematics of the spectrum of a signal that never repeats, the framework that says how much of a sound sits at each frequency, including the slow difference-frequency two close tones make. His wartime prediction-and-filtering theory (1942, published 1949) then showed how to pull a wanted frequency out of a noisy, beating mixture — the ancestor of every spectrum analyser and lock-in amplifier. Wiener never tuned a piano; what he gave us is the theory behind the tuner's ear, and the name for the field he founded along the way: cybernetics.

History

1636

— Marin Mersenne, in Harmonie universelle, describes the slow throbbing of two nearly-tuned organ pipes and links a string's pitch to its tension and length.

1701

— Joseph Sauveur counts the beats between two pipes to fix their absolute frequencies by ear and a clock, the first use of beats as a measuring tool.

1802

— Thomas Young explains beats as the interference of two wave trains, embedding the tuning trick in a wave theory of sound.

1930

— Norbert Wiener publishes Generalized Harmonic Analysis, the rigorous spectral theory of non-repeating signals and their difference frequencies.

1949

— Wiener's Extrapolation, Interpolation, and Smoothing of Stationary Time Series sets out how to extract a wanted frequency from a noisy, beating signal.

Going further

Texts. NCERT Physics Class 11, Ch. 15 ("Waves"), the beats section and its tuning-fork problems · H. C. Verma, Concepts of Physics Vol. 1, "Sound Waves," worked beat-and-tuning exercises · D. C. Pandey / Irodov-style problem sets on beats for JEE and NEET drill.

Provenance. Standard results. Recompute: candidates $256 \pm 4 = 252, 260$ and $340 \pm 5 = 335, 345$ ; the loading resolution $252\ \mathrm{Hz}$ (beats $4 \to 6$ ) and the filing resolution $345\ \mathrm{Hz}$ (beats $5 \to 7$ ); the tension case $f_2 = 256\sqrt{1.04} = 261.07 \approx 261.1\ \mathrm{Hz}$ , beat $5.07\ \mathrm{Hz}$ , with the small-change check $\tfrac12 \times 4\% = 2\%$ of $256 \approx 5.1\ \mathrm{Hz}$ ; the three-fork intersection $\{297,303\}\cap\{303,307\} = 303\ \mathrm{Hz}$ ; and the quiz beat $256$ against a $4\%$ -tightened wire, about $5\ \mathrm{Hz}$ . All are checked in verify_hs2-beats-problems.py.

Prerequisites. Beats — the superposition and sum-to-product identity that give $f_{\text{beat}} = |f_1 - f_2|$ , on which every step here stands. Leads to. the university-track article Beats, which treats the beat as a wave group and connects the tuning null to phase and group velocity.

Go deeper (UG). The spectral view of "find the unknown frequency" — a beat as the low end of a signal's spectrum, and how you extract a buried frequency from a noisy mixture — is Wiener's territory, developed in Beats in the university mechanics track. Enrichment only; nothing here depends on it.

Why does a beat count always give two answers instead of one?

Because a beat frequency is a magnitude, $|f_u - f_k|$ , and a magnitude has thrown away the sign. Four beats a second means the two frequencies differ by four hertz, but "differ by four" is satisfied equally by a source four hertz sharp and one four hertz flat. The measurement genuinely does not contain the information you need to choose; no amount of careful listening to a steady beat will reveal which tone is higher. That is why the archetype always has a second move — loading, filing, or a second reference — to inject the missing bit. If you ever find yourself certain of the sign from the beat count alone, you have assumed something the problem did not give you.

How exactly does loading a fork lower its pitch and filing raise it?

A tuning fork is a little oscillator whose frequency rises with the stiffness of its prongs and falls with their mass, roughly as $f \propto \sqrt{\text{stiffness}/\text{mass}}$ . Loading the prongs — a blob of wax, a clamped rider — adds mass without adding stiffness, so the denominator grows and $f$ drops. Filing metal off the tips, where the prongs move most, removes mass and lifts $f$ . (Filing near the base, by contrast, removes stiffness and lowers $f$ — which is why the rule is specifically "file the tips.") For a stretched wire the same up/down control comes from tension: $f \propto \sqrt{T}$ , so tightening sharpens the note and loosening flattens it. Whatever the source, you always have one knob whose direction you know for certain, and that certainty is what breaks the ambiguity.

In the wire example, why does a 4 percent tension change give only about a 2 percent frequency change?

Because frequency depends on the square root of tension: $f \propto \sqrt{T}$ . When a quantity varies as the square root of another, a small fractional change in the input produces about half that fractional change in the output — mathematically, $\Delta f/f \approx \tfrac12\,\Delta T/T$ . So a $4\%$ rise in tension lifts the pitch by roughly $2\%$ , not $4\%$ . The exact figure comes from $f_2 = f_1\sqrt{1.04} = f_1 \times 1.0198$ , a $1.98\%$ rise, which the linear one-half rule approximates beautifully for small changes. Examiners love this because the "obvious" answer — scale the frequency by the full $4\%$ — doubles the beat, turning a correct $5\ \mathrm{Hz}$ into a wrong $10\ \mathrm{Hz}$ .

If the beats disappear completely, are the two sources definitely equal?

Almost always, yes — a zero beat rate means the two frequencies match to within the fraction of a hertz your ear can resolve, and that is precisely the tuner's goal. There is one classic exception worth knowing: if the two sources are a whole octave apart, one at exactly twice the other, they do not beat either, because there is no slow difference frequency in the audible sense — the higher tone completes exactly two cycles for each one of the lower. So "no beats" ordinarily means "in unison," but a good examiner may ask you to notice that "no beats" could also mean "an exact octave apart." In a tuning context, where you have already brought the sources close, unison is the safe reading.

Can I resolve the sign by loading the reference fork instead of the unknown?

Yes, but you must then read the logic the other way round, and this is exactly where students trip. Every "above/below" conclusion is stated relative to the source you disturbed. If you load the reference (lowering its frequency) and the beats speed up, it is the reference that has run away from the unknown downward — meaning the reference is now below the unknown, i.e. the unknown was above the original reference. It is cleaner, and far less error-prone in an exam, to always disturb the unknown and apply the rule directly: load the unknown, watch its own frequency fall, and let "beats faster means it was already lower" finish the job. Pick one habit and keep it.

What if two beat counts are given, like the three-fork problem — do I still need loading or filing?

No. Two beat counts against two known frequencies over-determine the unknown, and the tie-breaker becomes consistency rather than a physical nudge. Each pairing gives its own $\pm$ pair of candidates; the true frequency is the one value that appears in both lists at once. In the three-fork example, A gives $\{297, 303\}$ and C gives $\{303, 307\}$ , and only $303\ \mathrm{Hz}$ satisfies both. This is why examiners hand you two beat measurements when they do not want to talk about wax and files: the arithmetic intersection does the work. Any extra sentence, such as "B lies between A and C," is a confirmation, not an independent requirement.

Does the amplitude or loudness of the two sources affect the beat frequency?

Not at all. The beat frequency is set purely by the difference of the two frequencies, $|f_u - f_k|$ , and is completely indifferent to how loud each source is. Unequal loudness changes only the depth of the throb — with a quiet unknown against a loud reference the silences between swells are no longer total, because the loud tone is never fully cancelled — but the swells still arrive at the same rate. This is genuinely useful: a barely audible mistuned string still beats at a countable rate against a strong fork, which is why you can tune a weak source against a strong reference. The count you time is trustworthy even when the two volumes are wildly different.