The last identity family of the chapter, in both directions. Adding and subtracting the sum and difference identities makes the mixed terms cancel, giving three product-to-sum formulas that turn a product of circular functions into a sum of first powers. Reversing them, with the two angles renamed as a half-sum and a half-difference, gives the sum-to-product formulas that turn a sum into a product — the form Section 10.7 needs, because only a product can be set to zero one factor at a time.
Subject: Trigonometry · 62 slides · symbolic lesson
Open the interactive version of this deck
Title
Trigonometry · Section 10.4c
The last identity family of the chapter, in both directions
Objectives
Everything in this section so far has been built from the cosine difference identity. This last family comes from doing one new thing with the identities you already have: adding two of them together, and then reading the result backwards.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 780-781 — Theorems 10.20 and 10.21
Section 10.4 has been one long chain. The cosine difference identity was proved once, from two chords of the unit circle, and every family since then has been squeezed out of it: the cofunction identities, the sine formulas, the double angle formulas, the power reduction formulas, the half angle formulas. This lesson closes the chain with the last two families, and they cost you no new geometry at all.
The trick is embarrassingly small. You have the cosine sum identity and the cosine difference identity sitting one above the other, differing in exactly one sign. Add them. The terms carrying that sign cancel, and what is left is a product on one side and a sum on the other. That is a product-to-sum formula, and there are three of them.
Then read those three backwards and you get the sum-to-product formulas, which is where the halves come from. By the end you should be able to derive any of the six in under a minute rather than recalling them, and, more importantly, to say which of the two directions a given problem is asking for.
Warm-up
Discussion prompt
Write down the cosine sum identity and the cosine difference identity, one directly above the other. Now add the two right-hand sides. Which terms survive, and which cancel?
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 773 — Theorem 10.16, from lesson 10.4a
Do this on paper before you look. Writing the two identities one under the other, with the like terms in the same columns, is most of the insight, and it is an insight you will lose if you read the answer first.
What you should see is that the two right-hand sides are built from the same two ingredients: a cosine times a cosine, and a sine times a sine. The only difference between the identities is whether the second ingredient is added or subtracted.
Once two expressions differ in exactly one sign, adding and subtracting them are both productive moves. Adding kills whatever carried the sign; subtracting kills whatever did not. This is the same manoeuvre you use to solve a pair of linear equations by elimination, and it is doing the same job here.
Section
Part 1
Concept
Take the two cosine identities at one fixed pair of angles and look at the four numbers involved. Both right-hand sides are built from the same two blocks.
Figure (svg): Four bars on a common baseline for the angles fifty and twenty degrees. The first bar, cos(a minus b) equal to 0.866, is split into a lower block of 0.604 and an upper block of 0.262. The second bar, cos(a plus b) equal to 0.342, is the same 0.604 block with 0.262 taken away, shown as a dashed red gap. The third bar is their difference, 0.524, exactly twice 0.262. The fourth is their sum, 1.208, exactly twice 0.604.
\[ \cos(\alpha-\beta) + \cos(\alpha+\beta) = 2\cos\alpha\cos\beta \]
\[ \cos(\alpha-\beta) - \cos(\alpha+\beta) = 2\sin\alpha\sin\beta \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 780 — before Theorem 10.20
The bars fix one pair of angles, fifty and twenty degrees, so you can watch the cancellation happen to actual numbers rather than to letters. The cosine of the difference, thirty degrees, is 0.866. The cosine of the sum, seventy degrees, is 0.342. Neither of those numbers looks related to the other until you see how each is built.
The first bar splits into a block of 0.604, which is the cosine of fifty times the cosine of twenty, and a block of 0.262, which is the sine of fifty times the sine of twenty. The second bar is the same 0.604 block with the same 0.262 taken away, shown as the dashed gap. Same two pieces, opposite signs.
So adding the two bars gives 1.208, which is exactly twice 0.604, and the sine blocks have vanished. Subtracting gives 0.524, exactly twice 0.262, and now the cosine blocks have vanished. The two equations under the picture say that in general, and dividing each by two gives you the first two product-to-sum formulas.
Worked example
Add the two cosine identities and solve for the product.
Write the difference identity
Why: This is the one that was proved from equal chords in lesson 10.4a.
\[ \cos(\alpha - \beta) = \cos\alpha\cos\beta + \sin\alpha\sin\beta \]
Write the sum identity beneath it
Why: The same expression with the sign of the sine term reversed.
\[ \cos(\alpha + \beta) = \cos\alpha\cos\beta - \sin\alpha\sin\beta \]
Add the two equations
Why: The sine-times-sine terms are opposites and cancel; the cosine-times-cosine terms are equal and double.
\[ \cos(\alpha-\beta) + \cos(\alpha+\beta) = 2\cos\alpha\cos\beta \]
Divide both sides by two
Why: This isolates the product on its own.
\[ \cos\alpha\cos\beta = \tfrac{1}{2}\big[\cos(\alpha-\beta) + \cos(\alpha+\beta)\big] \]
Verify at fifty and twenty degrees
Why: The left side is 0.643 times 0.940, which is 0.604. The right side is half of 0.866 plus 0.342, which is half of 1.208, also 0.604.
\[ \cos 50^\circ\cos 20^\circ = 0.604 = \tfrac12(0.866 + 0.342) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 780 — Theorem 10.20, first formula; Exercise 101
This derivation is four lines long and you should be able to write it from memory by the end of the lesson. That matters more than it sounds: if you can rebuild the formula you never have to gamble on whether the difference angle or the sum angle comes first.
Notice which identity is written on top. Putting the difference identity first is a deliberate habit, because it makes the difference angle appear first on the right of the finished formula, which is exactly the order the book states. Write the sum identity on top instead and you will end up with the terms swapped, which for this formula is harmless but for the sine-times-sine formula is fatal.
The verification at fifty and twenty degrees is the same arithmetic as the bar picture, and it is worth doing at least once with real numbers. Every one of these formulas can be checked in about twenty seconds with a calculator, and given how similar the six of them look, twenty seconds is cheap insurance.
Worked example
The same two identities, subtracted instead of added.
Subtract the sum identity from the difference identity
Why: Now the cosine-times-cosine terms cancel and the sine-times-sine terms double.
\[ \cos(\alpha-\beta) - \cos(\alpha+\beta) = 2\sin\alpha\sin\beta \]
Divide both sides by two
Why: The product is isolated.
\[ \sin\alpha\sin\beta = \tfrac{1}{2}\big[\cos(\alpha-\beta) - \cos(\alpha+\beta)\big] \]
Read the order carefully
Why: The difference angle still comes first, and now it is the one carrying the plus sign. Swapping the two terms would change the sign of the whole answer.
\[ \tfrac12\big[\cos(\alpha-\beta) - \cos(\alpha+\beta)\big] \ne \tfrac12\big[\cos(\alpha+\beta) - \cos(\alpha-\beta)\big] \]
Verify at fifty and twenty degrees
Why: The left side is 0.766 times 0.342, which is 0.262. The right side is half of 0.866 minus 0.342, which is half of 0.524, also 0.262.
\[ \sin 50^\circ\sin 20^\circ = 0.262 = \tfrac12(0.866 - 0.342) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 780 — Theorem 10.20, second formula; Exercise 101
Nothing new happens here except a sign. You subtract the second identity from the first instead of adding, which makes the cosine products cancel and the sine products double.
The third step is the one to slow down on. For this formula the order of the two terms on the right genuinely matters, because they are subtracted rather than added. The difference angle keeps its position at the front and it keeps the plus sign, so the sum angle is the one being taken away. Swap them and the whole answer changes sign, which is a real error rather than a cosmetic one.
A memory aid that survives pressure: the difference angle always goes first in all three product-to-sum formulas, without exception. Once you accept that, the only thing left to recall is whether the middle sign is a plus or a minus, and the minus belongs to sine times sine alone.
Worked example
The mixed product needs the two sine identities rather than the two cosine ones.
Write the sine difference identity
Why: Note that this one has mixed products on the right, not matched ones.
\[ \sin(\alpha - \beta) = \sin\alpha\cos\beta - \cos\alpha\sin\beta \]
Write the sine sum identity beneath it
Why: Only the sign of the second term changes.
\[ \sin(\alpha + \beta) = \sin\alpha\cos\beta + \cos\alpha\sin\beta \]
Add the two equations
Why: The cosine-times-sine terms cancel; the sine-times-cosine terms double.
\[ \sin(\alpha-\beta) + \sin(\alpha+\beta) = 2\sin\alpha\cos\beta \]
Divide by two
Why: The third and last product-to-sum formula.
\[ \sin\alpha\cos\beta = \tfrac{1}{2}\big[\sin(\alpha-\beta) + \sin(\alpha+\beta)\big] \]
Verify at fifty and twenty degrees
Why: The left side is 0.766 times 0.940, which is 0.720. The right side is half of the sine of thirty plus the sine of seventy, that is half of 0.5 plus 0.940, half of 1.440, also 0.720.
\[ \sin 50^\circ\cos 20^\circ = 0.720 = \tfrac12(0.500 + 0.940) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 780 — Theorem 10.20, third formula; Exercise 101
The mixed product cannot come from the cosine identities, because their right-hand sides contain only matched products, a cosine times a cosine and a sine times a sine. You need the two sine identities, whose right-hand sides contain only mixed products.
Adding them cancels the cosine-times-sine terms and doubles the sine-times-cosine terms, which is why the formula you get is for sine times cosine in that order, not for cosine times sine. In practice this never blocks you, because multiplication commutes and you may reorder the factors first, but it does mean you must decide which angle is playing the role of alpha before substituting.
One consequence worth noticing now: only the mixed product produces sines on the right. Both matched products produce cosines. That single sentence lets you sanity-check an answer before you have checked any angles.
Concept
Product to sum formulas — For all angles alpha and beta, each product of two circular functions equals one half of a sum or difference of two single cosines or sines, evaluated at the difference and the sum of the two angles.
\[ \cos\alpha\cos\beta = \tfrac12\big[\cos(\alpha-\beta) + \cos(\alpha+\beta)\big] \]
\[ \sin\alpha\sin\beta = \tfrac12\big[\cos(\alpha-\beta) - \cos(\alpha+\beta)\big] \]
\[ \sin\alpha\cos\beta = \tfrac12\big[\sin(\alpha-\beta) + \sin(\alpha+\beta)\big] \]
| product | comes out as | sign in the middle |
|---|---|---|
| cos times cos (matched) | cosines | plus |
| sin times sin (matched) | cosines | minus |
| sin times cos (mixed) | sines | plus |
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 780 — Theorem 10.20
Here are the three formulas together, exactly as the book states them. The table underneath compresses the differences between them into two columns, and those two columns are all you actually have to hold: what kind of function comes out, and what the sign in the middle is.
Read the pattern rather than the symbols. Matched product, meaning two cosines or two sines, produces cosines. Mixed product produces sines. The middle sign is plus in two cases and minus only for sine times sine. The difference angle is first every time, and the leading constant is one half every time.
The book leaves the proofs of these as Exercise 101, which is what the three previous slides just did. That is worth knowing if you are working through the exercise set: the derivation is the exercise, and you have already done it.
Notation
Annotate
On: \( \cos\alpha\cos\beta = \tfrac12\big[\cos(\alpha-\beta) + \cos(\alpha+\beta)\big] \)
Click through the four annotations rather than reading the formula as one lump. Each piece of it is doing a job, and once you can name the job you stop having to memorise the arrangement.
The piece students lose most often is the leading one half, and it is worth understanding why it is there rather than just remembering it. Adding the two identities did not produce the product, it produced twice the product, because the same product appeared in both lines. Halving undoes that doubling. Every product-to-sum formula carries the one half for exactly the same reason.
The last annotation is the one that saves you in an exam. If you can recall that a matched product gives cosines and the mixed product gives sines, you can discard half the possible answers before doing any arithmetic. Then a single substitution, usually zero or forty-five degrees, settles the rest.
Picture it
Figure (svg): Two stem plots against a horizontal axis marked in multiples of theta. On the left, two full-height stems at 2 and 6, the angles being multiplied. On the right, two half-height stems at 4 and 8, the difference and the sum of those angles.
Multiplying a wave at two theta by a wave at six theta does not give you anything at two theta or six theta. It gives you exactly two things, at the difference and at the sum, each at half the height.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 781 — Example 10.4.6, part 1
This is the most useful mental picture in the lesson. A product of two waves is not a wave at either of the frequencies you multiplied. It is two waves, and they live at the difference and at the sum.
Look at where the stems moved. On the left they stand at two and six, the angles you fed in. On the right they stand at four and eight, which are six minus two and six plus two, and they are only half as tall because of the one half in the formula. Nothing at all remains at two or at six.
Everything applied in the second half of this lesson is a consequence of this picture. Radio transmission, beats between two tuning forks, and the orthogonality that makes Fourier analysis work are all this same observation, read in one direction or the other.
Picture it
Figure (svg): The curve of cos 2θ times cos 6θ drawn from zero to pi, with the two half-cosines one half cos 4θ and one half cos 8θ drawn thin beneath it, and their sum drawn as a thick dashed line lying exactly on the product curve.
An identity says two recipes build the same curve. Here the thick solid curve is the product and the thick dashed curve is the sum of the two thin ones, and they lie on top of each other at every value of theta.
An identity is a claim that two different recipes produce the same object. When the object is a function, the claim is that the two recipes draw the same curve, and this is what that looks like when it is true.
The thick solid curve is the product. The two thin curves are the two halves of the answer, one oscillating four times over the interval and the other eight. The thick dashed curve is their sum, and it lies on the solid one everywhere, not just at a few points.
Use this as a check when you are unsure of a conversion. Plotting both sides is quicker than re-deriving, and a wrong sign or a wrong angle shows up instantly as two curves that part company. The one warning: two curves agreeing at a handful of points proves nothing, so look along the whole interval.
Matching
Match the pairs
Why: The three products are Theorem 10.20 in order. The fourth item is a sum, not a product, so it belongs to the other family entirely: it is Theorem 10.21, which the second half of this lesson derives. Matched products give cosines, the mixed product gives sines, and only the sine-times-sine formula has the minus sign in the middle.
Three of these four are the product-to-sum formulas and the fourth is a deliberate intruder from the other family, which you have not derived yet. Spotting that it does not belong is the real question here.
For the three genuine ones, do not work out any angles. Decide only two things: cosines or sines, and plus or minus in the middle. Matched products give cosines, the mixed one gives sines, and the minus lives with sine times sine.
The intruder is a sum, not a product, so no product-to-sum formula touches it. It needs the reverse direction, which turns it into twice a product of two cosines, and that is what the next part of the lesson builds. If you matched it correctly you have already guessed the shape of Theorem 10.21.
Prediction
\[ \sin\theta\cos\theta \]
Predict first
Apply the mixed product-to-sum formula with both angles equal to theta. What do you get?
Correct: One half of sin 2θ
Why: With both angles equal to theta the difference is zero and the sum is two theta, so the formula gives half of the sine of zero plus the sine of two theta, and the sine of zero is nothing. That leaves half the sine of two theta, which is exactly the double angle formula for sine read backwards. The product-to-sum family contains the double angle formulas as the special case where the two angles coincide.
\[ \sin\theta\cos\theta = \tfrac12\big[\sin 0 + \sin 2\theta\big] = \tfrac12\sin 2\theta \]
Pushing a new formula to a degenerate case is one of the fastest ways to build trust in it, and here the degenerate case turns out to be something you already know by heart.
With both angles equal, the difference angle collapses to zero. The sine of zero is nothing, so one of the two terms simply disappears, and what is left is half the sine of the double angle. Rearranged, that is the double angle formula for sine from lesson 10.4b.
The same thing happens to the other two formulas. Cosine times cosine with equal angles gives half of one plus the cosine of the double angle, which is the power reduction formula for cosine squared, and sine times sine gives half of one minus it. So the product-to-sum family quietly contains the whole of the previous lesson as its special case, which is a good reason to trust it and a good reason to prefer deriving over memorising.
Real world
Figure (svg): Two unit circles. On the left, rays at 36.87 and 53.13 degrees drop to the horizontal axis at 0.8 and 0.6. On the right, rays at 16.26 degrees, the difference of those angles, and 90 degrees, their sum, drop to the axis at 0.96 and 0.
Discussion prompt
The footnote in the book calls these the prosthaphaeresis formulas and says they have a rich history. Using only a cosine table, an addition, a subtraction and a halving, how would you multiply 0.8 by 0.6? Work it through with the picture.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 780 — footnote 3
The book adds a footnote recommending that you research the history of these formulas, so here is the short version, drawn rather than told.
Before logarithms, multiplying two long decimals by hand was slow and error-prone, and sixteenth-century astronomers did an enormous amount of it. Somebody noticed that if you scale your two numbers into the range covered by cosine, you can look up the two angles in a table, subtract them, add them, look up two more cosines, add those and halve. Six table lookups and three easy operations, and not one multiplication.
Follow the two circles. On the left, the two numbers sit on the horizontal axis as the cosines of their angles. On the right, the cosines of the difference and of the sum of those angles. Half of their total is the product you wanted. The name prosthaphaeresis is just Greek for addition and subtraction, which is all the method uses. Logarithms replaced it within a generation, but the identities stayed.
Section
Part 2
Worked example
Write the product below as a sum.
\[ \cos(2\theta)\cos(6\theta) \]
Name the two angles
Why: The book takes them in the order written.
\[ \alpha = 2\theta, \qquad \beta = 6\theta \]
Apply the cosine-times-cosine formula
Why: Half the sum of the cosines at the difference angle and the sum angle.
\[ \tfrac12\big[\cos(2\theta - 6\theta) + \cos(2\theta + 6\theta)\big] \]
Simplify the two angles
Why: Two theta minus six theta is minus four theta; two theta plus six theta is eight theta.
\[ = \tfrac12\cos(-4\theta) + \tfrac12\cos(8\theta) \]
Clear the negative angle
Why: Cosine is even, so the minus sign inside simply disappears.
\[ \cos(-4\theta) = \cos(4\theta) \]
State the answer
Why: Two cosines, each at first power, each with coefficient one half.
\[ \cos(2\theta)\cos(6\theta) = \tfrac12\cos(4\theta) + \tfrac12\cos(8\theta) \]
Check at theta equal to pi over eight
Why: The left side is the cosine of pi over four times the cosine of three pi over four, which is 0.707 times minus 0.707, namely minus one half. The right side is half the cosine of pi over two plus half the cosine of pi, which is zero plus minus one half.
\[ \cos\tfrac\pi4\cos\tfrac{3\pi}4 = -\tfrac12 = \tfrac12(0) + \tfrac12(-1) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 781 — Example 10.4.6, part 1 (also Exercise 77)
This is the book's own example, and the only genuinely new thing in it is the last piece of bookkeeping. Everything before that is substituting into a formula you derived four slides ago.
Notice the book takes the angles in the order written, so alpha is the smaller one and the difference comes out negative. That is fine, and it happens often. Cosine is an even function, so the minus sign inside it simply evaporates, leaving the cosine of four theta. If the negative angle had landed inside a sine you would have had more work to do.
The check is worth copying as a habit. Pick a value of theta that makes the angles land on the special values, here pi over eight, so that both sides come out as exact numbers you can compare at a glance. Any value would do, but one that gives clean numbers means you can do the check in your head instead of on a calculator.
Worked example
A matched product of two sines. The middle sign is the minus one.
\[ \sin(2\theta)\sin(7\theta) \]
Name the two angles and apply the formula
Why: Half the cosine at the difference angle minus the cosine at the sum angle.
\[ \tfrac12\big[\cos(2\theta - 7\theta) - \cos(2\theta + 7\theta)\big] \]
Simplify the two angles
Why: The difference is minus five theta and the sum is nine theta.
\[ = \tfrac12\cos(-5\theta) - \tfrac12\cos(9\theta) \]
Clear the negative angle
Why: Cosine is even again, so only the label changes.
\[ = \tfrac12\cos(5\theta) - \tfrac12\cos(9\theta) \]
Write it as the book does
Why: One fraction over the whole difference.
\[ \sin(2\theta)\sin(7\theta) = \frac{\cos(5\theta) - \cos(9\theta)}{2} \]
Check at theta equal to ten degrees
Why: The left side is the sine of twenty times the sine of seventy, that is 0.342 times 0.940, namely 0.321. The right side is half of the cosine of fifty minus the cosine of ninety, half of 0.643 minus 0, namely 0.321.
\[ 0.342 \times 0.940 = 0.321 = \tfrac12(0.643 - 0) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 785 — Exercise 75; answer p. 789
The one thing that separates this from the previous example is the middle sign, and that sign is the single most likely place to lose a mark in this whole lesson. A matched product of two sines is the only one of the three formulas with a minus in the middle.
Everything else runs the same way. Difference angle first, which is negative here, and cosine is even so the negative vanishes. Then the sum angle, carrying the minus.
The book writes its answer as a single fraction with the two cosines over a two, rather than as two separate halves. The two forms are the same thing, and either is acceptable, but matching the book's layout makes it much easier to compare your work against the answer key on page 789.
Worked example
A mixed product written the other way round. The formula is stated for sine times cosine, so the first job is to get it into that shape.
\[ \cos(\theta)\sin(3\theta) \]
Reorder the factors
Why: Multiplication commutes, so this is a sine times a cosine with alpha equal to three theta.
\[ \cos(\theta)\sin(3\theta) = \sin(3\theta)\cos(\theta) \]
Apply the mixed formula
Why: Half the sine at the difference angle plus the sine at the sum angle, with alpha three theta and beta theta.
\[ \tfrac12\big[\sin(3\theta - \theta) + \sin(3\theta + \theta)\big] \]
Simplify the two angles
Why: Both come out positive, so no cleanup is needed.
\[ \cos(\theta)\sin(3\theta) = \frac{\sin(2\theta) + \sin(4\theta)}{2} \]
See what reordering saved
Why: Had you taken alpha as theta and beta as three theta you would have written half of the sine of minus two theta plus the sine of four theta, and then needed the odd identity to fix the first term.
\[ \tfrac12\big[\sin(-2\theta) + \sin(4\theta)\big] = \tfrac12\big[-\sin(2\theta) + \sin(4\theta)\big] \]
Check at theta equal to fifteen degrees
Why: The left side is the cosine of fifteen times the sine of forty-five, that is 0.966 times 0.707, namely 0.683. The right side is half of the sine of thirty plus the sine of sixty, half of 0.5 plus 0.866, namely 0.683.
\[ 0.966 \times 0.707 = 0.683 = \tfrac12(0.500 + 0.866) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 785 — Exercise 79; answer p. 789
Read the question carefully before reaching for anything. The formula in Theorem 10.20 is stated for a sine times a cosine in that order, and this product is written the other way round.
Because multiplication commutes you may simply swap the factors, and doing so is not just tidiness. If you take alpha to be theta and beta to be three theta, the difference angle is negative and it lands inside a sine, so you pick up an odd-identity cleanup at the end. Swap first and both output angles come out positive and no cleanup is needed at all.
The fourth step shows you the road not taken, so that you can see both answers really are the same. The extra minus sign in front of the sine of two theta is exactly what the odd identity produces, and it cancels against the order of the terms. Neither route is wrong; one is shorter.
Fill the middle
Write the product of the sine of five theta and the sine of three theta as a difference.
Fill in the blanks
\sin(5\theta)\sin(3\theta) = \tfrac12\big[\cos(2\theta) - \cos(8\theta)\big]
Why: The difference of the angles is five theta minus three theta, which is two theta, and it goes first, carrying the plus sign. The sum is eight theta and it goes second, carrying the minus sign. Checking at theta equal to pi over sixteen: the left side is 0.831 times 0.556, namely 0.462, and the right side is half of 0.924 minus 0, also 0.462.
Two blanks, and they test the two things that go wrong most often: which angle comes first, and what it is.
The first blank is the difference of the two angles, five theta minus three theta, and it belongs in the first slot because the difference always does. The second is the sum. If you wrote them the other way round, notice that the answer would come out as the negative of the truth, because these two terms are subtracted.
The explanation includes a numerical check at pi over sixteen. Get into the habit of doing that check yourself before you look, because it costs a few seconds and it catches everything: a wrong angle, a wrong order and a missing one half all show up as two numbers that fail to match.
Trap
A common slip on the sine-times-sine formula:
\[ \sin\alpha\sin\beta \]
\[ = \tfrac12\big[\cos(\alpha+\beta) - \cos(\alpha-\beta)\big] \]
Wrong. That is the negative of the right answer.
In all three formulas the difference angle comes first. For the two cosine formulas it is the term that keeps the plus sign, so writing the sum first flips the sign of everything.
\[ \tfrac12\big[\cos(\alpha-\beta) - \cos(\alpha+\beta)\big] \]
This is not a silly mistake. It is what happens when you remember that the sine-times-sine formula involves a difference of two cosines but not which of them comes first, and then guess. Half the time you will guess right, which is worse than always being wrong, because it stops you from noticing.
The cure is the derivation. If you can add and subtract the two cosine identities in your head, the difference angle is obviously the one that arrives with the plus sign, because it was the identity you started from. Two lines of work replaces one coin toss.
The check on the right of this slide is the fastest available. Both angles at forty-five degrees makes the left side the square of the sine of forty-five, which is one half. The correct formula returns one half; the swapped one returns minus one half. A single substitution, and you never have to trust your memory of the order again.
Error analysis
Annotate
On: \( \cos(4x)\cos(2x) = \tfrac12\big[\sin(2x) + \sin(6x)\big] \)
\[ \cos(4x)\cos(2x) = \tfrac12\big[\cos(2x) + \cos(6x)\big] \]
Work out what is wrong here before opening the annotations. The angles are right, the leading one half is right, and the line still fails, so it is not the kind of error that careful arithmetic prevents.
What has happened is that the mixed formula has been applied to a matched product. That is a formula-choice error, and formula-choice errors are invisible to the usual checks: everything downstream is done correctly on the wrong starting point.
The fourth annotation is the defence. Substituting x equal to zero costs nothing, because every cosine becomes one and every sine becomes zero, and it separates cosines from sines immediately. Make it a reflex: after any conversion in this lesson, put the variable equal to zero and compare the two sides.
Sorting
Sort into buckets
Sort each product by the kind of thing it becomes.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 785 — Exercises 74 to 79
These six products are Exercises 74 through 79 from the section's exercise set, and sorting them is the whole first stage of solving them. You are not being asked for any angles here, only for the shape of the answer.
Two questions decide every one of them. Are the two functions the same kind, or different kinds? And if they are the same kind, are they cosines or sines? Matched means cosines come out, mixed means sines come out, and only two sines multiplied together produce the minus sign in the middle.
Notice the fifth item, a cosine written before a sine. It still belongs with the mixed products, because you may reorder the factors. Do that reordering as the first line of your written solution rather than in your head, because that is where the decision about which angle is alpha gets made.
Worked example
This is why calculus courses teach Theorem 10.20. Find the area under the product below between zero and pi over four.
\[ \int_0^{\pi/4} \sin(4x)\cos(2x)\,dx \]
Notice that no rule integrates a product
Why: There is no product rule for integrals, and neither factor is the derivative of the other, so substitution fails too.
\[ \int \sin(4x)\cos(2x)\,dx: \quad \text{no direct rule} \]
Convert the product to a sum
Why: Mixed product, so sines come out: the difference angle is 2x and the sum angle is 6x.
\[ \sin(4x)\cos(2x) = \tfrac12\sin(2x) + \tfrac12\sin(6x) \]
Integrate each first power separately
Why: Each is a single sine, which integrates to a cosine divided by its own coefficient.
\[ \int\left(\tfrac12\sin 2x + \tfrac12\sin 6x\right)dx = -\tfrac14\cos 2x - \tfrac1{12}\cos 6x \]
Evaluate between zero and pi over four
Why: At the top, the cosine of pi over two is zero and the cosine of three pi over two is zero. At the bottom both cosines are one.
\[ \left[-\tfrac14\cos 2x - \tfrac1{12}\cos 6x\right]_0^{\pi/4} = 0 - \left(-\tfrac14 - \tfrac1{12}\right) = \tfrac13 \]
Figure (svg): The curve of sin 4x times cos 2x from zero to pi over four, shaded under the curve, with the two curves one half sin 2x and one half sin 6x drawn thin. The shaded area is labelled one third.
Check against the picture
Why: The shaded region is roughly a bump of height 1 over a base of about 0.5, so an area of about a third is exactly the right size; a numerical estimate of the integral gives 0.3333.
\[ \int_0^{\pi/4}\sin(4x)\cos(2x)\,dx = \tfrac13 \approx 0.3333 \]
This slide is a preview rather than a calculus lesson, but it is the honest answer to why a trigonometry course teaches these formulas at all. When you meet integrals, products of trigonometric functions are a wall, and this family is the door through it.
The reason is worth stating plainly. Integration has no product rule. If an expression is a product of two things, neither of which is the derivative of the other, there is nothing to reach for. But a sum of first powers integrates term by term, and each term is a single sine whose antiderivative is a cosine divided by the coefficient inside it.
The picture shows the size of the answer. The shaded bump has height around one and width around eight tenths of a unit, but it is far from rectangular, so a third is a believable area. That is the sort of estimate worth making before you trust any integral you have computed, and here a numerical estimate agrees to four decimal places.
Section
Part 3
Concept
Reading the product-to-sum formulas from right to left turns a sum into a product. But the angles change: what comes out of a product-to-sum formula is the difference and the sum of the inputs, so the inputs must be recovered from those.
\[ \text{given } \alpha \text{ and } \beta, \text{ find } A, B \text{ with } A - B = \alpha \text{ and } A + B = \beta \]
Figure (svg): A number line marked in multiples of x from zero to eight, with beta at 3x and alpha at 7x marked by dots, their midpoint at 5x marked by a larger dot, and two arrows below showing that the midpoint is the half-sum and that the distance from the midpoint to either end is the half-difference, 2x.
\[ A = \frac{\alpha + \beta}{2}, \qquad B = \frac{\beta - \alpha}{2} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 781 — before Theorem 10.21
Here is the hinge of the lesson. You already have three formulas that turn a product into a sum. Reading a true equation from right to left is always allowed, so you already have three formulas that turn a sum into a product. The only problem is that the sums they produce are at the difference and the sum of the input angles, not at the angles you happen to have.
So pose it as a question. Given the two angles you actually have, what would you need to feed in so that the difference and the sum come out as those two? That is two equations in two unknowns, and it has exactly one answer.
The number line shows that answer without any algebra. The angle you feed into the first slot is the point halfway between your two angles, and the angle you feed into the second is the distance from that midpoint to either end. Halfway is the half-sum; the distance out is the half-difference. Once you have seen that picture, the fractions in Theorem 10.21 stop looking arbitrary.
Worked example
Start from the finished product-to-sum formula and rename the angles so that the two outputs are the two angles you actually have.
Write the cosine-times-cosine formula with fresh letters
Why: Use A and B for the inputs, so that alpha and beta can be the outputs.
\[ \cos A\cos B = \tfrac12\big[\cos(A-B) + \cos(A+B)\big] \]
Multiply through by two and swap the sides
Why: Now the sum of two cosines sits on the left, where it is wanted.
\[ \cos(A-B) + \cos(A+B) = 2\cos A\cos B \]
Demand that the two outputs be beta and alpha
Why: Set the difference angle equal to beta and the sum angle equal to alpha.
\[ A - B = \beta, \qquad A + B = \alpha \]
Solve the little system
Why: Adding the two equations gives twice A; subtracting them gives twice B.
\[ A = \frac{\alpha + \beta}{2}, \qquad B = \frac{\alpha - \beta}{2} \]
Substitute back
Why: The half-sum and the half-difference, exactly as the number line showed.
\[ \cos\alpha + \cos\beta = 2\cos\!\left(\frac{\alpha+\beta}{2}\right)\cos\!\left(\frac{\alpha-\beta}{2}\right) \]
Verify at alpha equal to 60 degrees and beta equal to 0
Why: The left side is 0.5 plus 1, namely 1.5. The right side is twice the cosine of thirty times the cosine of thirty, that is twice 0.866 squared, namely 1.5.
\[ \cos 60^\circ + \cos 0^\circ = 1.5 = 2\cos^2 30^\circ \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 781 — Theorem 10.21, first formula; Exercise 102
Follow the letters carefully on this one. Capital A and capital B are the angles you feed into the old formula, and alpha and beta are the angles you want to come out. Using different letters for the two roles is the whole reason this derivation stays readable.
The third and fourth steps are the little two-equation system from the previous slide, solved. Adding the two equations eliminates B and gives twice A; subtracting them eliminates A and gives twice B. That is where both halves come from, and it is why every sum-to-product formula has halved angles while no product-to-sum formula does.
The verification uses sixty degrees and zero, which is a good pair to keep for this purpose because it makes both output angles thirty degrees and turns the right-hand side into twice a square. If you can do a check entirely with special angles, you can do it without a calculator, and a check you can do without a calculator is a check you will actually do.
Concept
Sum to product formulas — For all angles alpha and beta, a sum or difference of two cosines or two sines equals twice a product of two circular functions evaluated at the half-sum and the half-difference of the two angles.
\[ \cos\alpha + \cos\beta = 2\cos\!\left(\frac{\alpha+\beta}{2}\right)\cos\!\left(\frac{\alpha-\beta}{2}\right) \]
\[ \cos\alpha - \cos\beta = -2\sin\!\left(\frac{\alpha+\beta}{2}\right)\sin\!\left(\frac{\alpha-\beta}{2}\right) \]
\[ \sin\alpha \pm \sin\beta = 2\sin\!\left(\frac{\alpha\pm\beta}{2}\right)\cos\!\left(\frac{\alpha\mp\beta}{2}\right) \]
Three lines, four formulas: the last one is read twice, once with the top signs throughout and once with the bottom signs throughout.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 781 — Theorem 10.21
Three lines on the slide, four formulas in the theorem, because the last line is read twice. Take the top signs all the way through and you have the formula for a sum of two sines; take the bottom signs all the way through and you have the formula for their difference. What you must never do is mix a top sign in one place with a bottom sign in another.
Compare this block with Theorem 10.20 on the earlier slide. Every feature is inverted. The leading constant is two instead of one half. The angles are halved instead of doubled up into a sum and a difference. A sum goes in and a product comes out. That is what you expect of two operations that undo each other, and noticing it halves what you have to hold.
One asymmetry is not explained by the inversion and simply has to be noticed: the second formula, cosine minus cosine, carries a minus sign out in front. None of the other three does. It is the single most commonly dropped sign in the family, and there is a trap slide about it a few screens from here.
Notation
Annotate
On: \( \sin\alpha \pm \sin\beta = 2\sin\!\left(\tfrac{\alpha\pm\beta}{2}\right)\cos\!\left(\tfrac{\alpha\mp\beta}{2}\right) \)
\[ \sin\alpha + \sin\beta = 2\sin\!\left(\tfrac{\alpha+\beta}{2}\right)\cos\!\left(\tfrac{\alpha-\beta}{2}\right) \]
\[ \sin\alpha - \sin\beta = 2\sin\!\left(\tfrac{\alpha-\beta}{2}\right)\cos\!\left(\tfrac{\alpha+\beta}{2}\right) \]
This line packs two formulas into one, and the packing is where people come unstuck. Work through the annotations in order and keep one finger on the line.
The rule is that the sign inside the sine matches the sign on the left, and the sign inside the cosine is the opposite. So for a sum of two sines, the sine gets the half-sum and the cosine gets the half-difference. For a difference of two sines, they trade places: the sine gets the half-difference and the cosine gets the half-sum.
The two formulas are written out separately underneath so that you can see the swap rather than decode it. If you find the packed version slippery under pressure, learn the two unpacked lines instead. There is no virtue in the compressed form beyond saving a line of print, and a formula you can read correctly is worth more than one you can write compactly.
Picture it
Figure (svg): The curve of cos 3θ plus cos 5θ over zero to two pi, oscillating quickly, with the dashed curves plus and minus two cos θ hugging it from above and below as an envelope that pinches to zero near pi over two and three pi over two.
Two cosines at nearby angles add up to a single fast wave at the half-sum angle, whose height is being squeezed and released by a slow wave at the half-difference angle. The sum-to-product formula is that sentence in symbols.
This picture is the reason the sum-to-product formulas matter outside the classroom, and it is worth staring at until the structure separates.
The wiggling curve is the sum of two cosines at nearby multiples of theta, three and five. The formula says it equals twice the cosine of four theta, the average of the two, multiplied by the cosine of theta, half the gap between them. Those two factors play completely different roles: the fast one is the wave you see oscillating, and the slow one is the dashed pair of curves squeezing it.
Trace the dashed envelope with your eye. Where it is wide the sum reaches nearly two; where it pinches to zero near pi over two the sum is momentarily flat against the axis. Everything in the rest of this lesson about beats, about radio and about amplitude comes out of reading that one picture in one direction or the other.
Worked example
Write the difference below as a product.
\[ \sin(\theta) - \sin(3\theta) \]
Name the two angles
Why: The book takes them in the order written.
\[ \alpha = \theta, \qquad \beta = 3\theta \]
Apply the sine formula with the bottom signs
Why: Twice the sine of the half-difference times the cosine of the half-sum.
\[ 2\sin\!\left(\frac{\theta - 3\theta}{2}\right)\cos\!\left(\frac{\theta + 3\theta}{2}\right) \]
Simplify the two angles
Why: Minus two theta over two is minus theta; four theta over two is two theta.
\[ = 2\sin(-\theta)\cos(2\theta) \]
Clear the negative angle
Why: Sine is odd, so the minus sign comes out to the front of the whole expression.
\[ \sin(\theta) - \sin(3\theta) = -2\sin(\theta)\cos(2\theta) \]
Check at theta equal to thirty degrees
Why: The left side is 0.5 minus the sine of ninety, which is 0.5 minus 1, namely minus 0.5. The right side is minus two times 0.5 times the cosine of sixty, which is minus one times 0.5, also minus 0.5.
\[ 0.5 - 1 = -0.5 = -2(0.5)(0.5) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 781 — Example 10.4.6, part 2
The book's second example, and the first one where a negative angle causes real trouble rather than evaporating.
Because this is a difference of two sines, the bottom signs apply: the sine takes the half-difference and the cosine takes the half-sum. The half-difference is theta minus three theta, all over two, which is minus theta, and it lands inside a sine. Sine is odd, so it pulls a minus sign out with it, and that minus multiplies the two at the front.
That is the whole difference between this example and the easy ones. If a negative angle lands inside a cosine it disappears silently; if it lands inside a sine it changes the sign of your answer. Before you write a final line, look at every angle in it and ask whether any of them are negative.
Worked example
The friendliest member of the family: no minus sign out front and no negative angle to clear.
\[ \cos(3\theta) + \cos(5\theta) \]
Compute the half-sum
Why: Three theta plus five theta is eight theta, and half of that is four theta.
\[ \frac{3\theta + 5\theta}{2} = 4\theta \]
Compute the half-difference
Why: Three theta minus five theta is minus two theta, and half of that is minus theta.
\[ \frac{3\theta - 5\theta}{2} = -\theta \]
Apply the cosine plus cosine formula
Why: Twice the product of the cosines at those two angles.
\[ 2\cos(4\theta)\cos(-\theta) \]
Clear the negative angle
Why: Cosine is even, so the sign inside vanishes with nothing left behind.
\[ \cos(3\theta) + \cos(5\theta) = 2\cos(4\theta)\cos(\theta) \]
Check at theta equal to fifteen degrees
Why: The left side is the cosine of forty-five plus the cosine of seventy-five, that is 0.707 plus 0.259, namely 0.966. The right side is twice the cosine of sixty times the cosine of fifteen, that is twice 0.5 times 0.966, also 0.966.
\[ 0.707 + 0.259 = 0.966 = 2(0.5)(0.966) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 785 — Exercise 80; answer p. 789
Take a moment with the easy one, because it shows you what the formula looks like when nothing goes wrong, and that is the baseline you are comparing against when something does.
A sum of two cosines uses the only formula in the family with no leading minus sign, and the half-difference lands inside a cosine, where a negative angle costs nothing. So the whole exercise is two divisions and a substitution.
The verification at fifteen degrees uses a small trick worth stealing. Choosing theta so that one of the two output angles becomes a special value, here sixty degrees, makes one factor exactly one half and leaves you comparing two numbers instead of four. Deliberately choosing a convenient test value is not cheating; it is what makes the check quick enough to be worth doing every time.
Worked example
Consecutive multiples, so the halves do not come out whole. That is normal and nothing to fix.
\[ \cos(5\theta) - \cos(6\theta) \]
Compute the half-sum and half-difference
Why: Eleven theta over two, and minus theta over two.
\[ \frac{5\theta+6\theta}{2} = \frac{11}{2}\theta, \qquad \frac{5\theta-6\theta}{2} = -\frac{1}{2}\theta \]
Apply the cosine minus cosine formula
Why: This is the only one of the four with a minus sign built into the front.
\[ -2\sin\!\left(\tfrac{11}{2}\theta\right)\sin\!\left(-\tfrac12\theta\right) \]
Clear the negative angle
Why: Sine is odd, so a second minus sign comes out and the two cancel.
\[ \sin\!\left(-\tfrac12\theta\right) = -\sin\!\left(\tfrac12\theta\right) \]
Collect the signs
Why: Minus times minus is plus, so the finished answer is positive.
\[ \cos(5\theta) - \cos(6\theta) = 2\sin\!\left(\tfrac{11}{2}\theta\right)\sin\!\left(\tfrac12\theta\right) \]
Check at theta equal to pi over eleven
Why: The left side is the cosine of five pi over eleven minus the cosine of six pi over eleven, namely 0.2846. The right side is twice the sine of pi over two times the sine of pi over twenty-two, that is twice 1 times 0.1423, also 0.2846.
\[ 0.2846 = 2(1)(0.1423) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 785 — Exercise 82; answer p. 789
Two things are going on at once here, which is why it is worth its own slide.
First, the angles are consecutive multiples, so their sum is odd and their difference is a single theta. Halving gives eleven halves and one half, and there is nothing to be done about that. Fractions in the answer are normal and are not a sign that you have gone wrong. The book's answer on page 789 has exactly these two fractions in it.
Second, two minus signs meet. The formula for a difference of cosines already carries a minus out in front, and the negative half-difference inside the sine produces another. Two minus signs multiply to a plus, so the finished answer is positive. Track them separately rather than trying to see the outcome at once, because the commonest way to get this wrong is to notice one of the two and not the other.
Fill the middle
Write the sum of the sine of five x and the sine of x as a product.
Fill in the blanks
\sin 5x + \sin x = 2\sin\!\left(\frac3x2x\right)\cos\!\left(\frac______\right) = 2\sin(___)\cos(___)
Why: Six x over two is 3x, and it goes inside the sine because this is sine PLUS sine, where the sine takes the half-sum. Four x over two is 2x, and it goes inside the cosine. Checking at x equal to pi over twelve: the left side is 0.966 plus 0.259, namely 1.2247, and the right side is twice 0.707 times 0.866, also 1.2247.
One blank for each factor, and the thing being tested is which of the two computed angles goes where.
This is a sum of two sines, so the top signs apply throughout and the sine gets the half-sum. Six x over two is three x, so three x goes inside the sine. Four x over two is two x, so two x goes inside the cosine. Had the question been a difference, those two would have swapped places.
That swap really is the whole content of the plus-or-minus line, and it is the thing most worth drilling from this part of the lesson. Say it out loud once: sine plus sine puts the half-sum in the sine; sine minus sine puts the half-difference in the sine.
Trap
Converting a difference of cosines:
\[ \cos 5x - \cos x \]
\[ = 2\sin(3x)\sin(2x) \]
Wrong. The sign in front is missing.
Both angles are right: the half-sum is 3x and the half-difference is 2x, and a difference of cosines does give a product of sines. But this one formula carries a leading minus sign, and it is the only one of the four that does.
\[ \cos 5x - \cos x = -2\sin(3x)\sin(2x) \]
Everything about the wrong line is right except one character, which is exactly what makes it dangerous. The angles were computed correctly. The recognition that a difference of cosines produces a product of two sines was correct. The answer is still wrong by a factor of minus one.
There is no deep reason for that leading minus sign, and no trick will make it feel inevitable. It has to be noticed and remembered. The one consolation is that it is unique: no other formula in Theorem 10.21 has anything in front except the two, so you only have one exception to carry.
Test at a value where the two sides are clearly different, not at one where both happen to vanish. At x equal to pi over six the correct answer is negative and the wrong one is positive, so the sign error is obvious. A test value that returns zero on both sides tells you nothing at all.
Prediction
\[ \cos(4x) - \cos(4x) \]
Predict first
Push the cosine minus cosine formula to its limit: what does it give when the two angles are identical?
Correct: Zero
Why: The half-sum is 4x and the half-difference is zero, so the formula returns minus two times the sine of 4x times the sine of zero. The sine of zero is nothing, so the whole product is zero — which is obviously right, since the original expression was a quantity minus itself. A formula that survives its own degenerate case is a formula you can trust.
\[ -2\sin(4x)\sin(0) = 0 \]
Formulas deserve to be pushed until they break, and the useful discovery is usually that they do not break.
Set the two angles equal and the half-difference collapses to zero, so the second factor becomes the sine of zero, which is nothing. The whole product is zero. And that has to be right, because the expression you started with was a quantity minus itself.
This is a small thing but it is a real habit. Whenever a new formula arrives, feed it a case whose answer you already know, and see whether it agrees. Here the degenerate case confirms the leading minus sign is harmless, confirms the angle placement, and costs one line. The same trick on the product-to-sum family recovered the double angle formulas a few slides ago.
Section
Part 4
Concept
The half-difference goes negative whenever the second angle is the larger one, and half the time it lands inside a sine. The even and odd identities finish the job, and an answer is not finished until they have.
Figure (svg): A unit circle with a ray at forty degrees reaching a point above the axis and a ray at minus forty degrees reaching the point directly below it. A dashed vertical segment joins them, a dashed horizontal segment marks the shared first coordinate, and the two second coordinates are marked as opposites.
\[ \cos(-t) = \cos t, \qquad \sin(-t) = -\sin t \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 781 — the even and odd identities used in Example 10.4.6
This part of the lesson is bookkeeping rather than mathematics, but it is where marks are actually lost, so it gets its own section.
The source is always the same. The half-difference is negative whenever the second of the two angles is the larger, and roughly half the time it lands inside a sine rather than a cosine. The picture is the reason the two cases behave so differently: flipping the sign of an angle reflects its point on the unit circle across the horizontal axis, which leaves the first coordinate alone and negates the second. Cosine reads the first coordinate and does not notice; sine reads the second and changes sign.
The third bullet offers a way to avoid the whole issue, by taking the larger angle first so the half-difference is never negative. It works, but be honest about what it costs: if the expression was written as a difference in a fixed order, reordering it means carrying a minus sign out front instead. The bookkeeping moves rather than vanishing. Either route is fine as long as you finish it.
Worked example
A sine difference where the second angle is the larger, so the half-difference is negative.
\[ \sin(2\theta) - \sin(7\theta) \]
Apply the sine formula with the bottom signs
Why: The sine takes the half-difference, the cosine takes the half-sum.
\[ 2\sin\!\left(\frac{2\theta - 7\theta}{2}\right)\cos\!\left(\frac{2\theta + 7\theta}{2}\right) \]
Simplify the two angles
Why: Minus five theta over two, and nine theta over two.
\[ = 2\sin\!\left(-\tfrac52\theta\right)\cos\!\left(\tfrac92\theta\right) \]
Clear the negative angle
Why: Sine is odd, so the minus comes out in front of the 2.
\[ \sin(2\theta) - \sin(7\theta) = -2\cos\!\left(\tfrac92\theta\right)\sin\!\left(\tfrac52\theta\right) \]
Check at theta equal to fifteen degrees
Why: The left side is the sine of thirty minus the sine of one hundred and five, that is 0.5 minus 0.966, namely minus 0.466. The right side is minus two times the cosine of 67.5 times the sine of 37.5, that is minus two times 0.383 times 0.609, also minus 0.466.
\[ 0.5 - 0.966 = -0.466 = -2(0.383)(0.609) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 785 — Exercise 81; answer p. 789
This is the situation the previous slide described, in a real exercise from page 785. The second angle is the larger, so the half-difference is negative, and because this is a difference of sines the half-difference is exactly what lands inside the sine.
Follow the minus sign through. The sine of a negative angle equals the negative of the sine of the positive angle, so the minus comes out and multiplies the two at the front, giving minus two. Both surviving angles are then positive and the answer is finished.
The book writes the answer with the cosine first and the sine second, which is the same product written in the other order. Do not be thrown by that when you compare against the key on page 789: multiplication commutes, and the only things that have to match are the two angles, the two function names and the sign.
Worked example
Here the negative angle is in the question rather than in the answer. Clear it before you reach for any formula.
\[ \sin(9\theta) - \sin(-\theta) \]
Clear the negative angle first
Why: Sine is odd, so subtracting the sine of minus theta is adding the sine of theta.
\[ \sin(9\theta) - \sin(-\theta) = \sin(9\theta) + \sin(\theta) \]
Now it is a sine plus sine
Why: Top signs: the sine takes the half-sum and the cosine takes the half-difference.
\[ 2\sin\!\left(\frac{9\theta + \theta}{2}\right)\cos\!\left(\frac{9\theta - \theta}{2}\right) \]
Simplify the two angles
Why: Ten theta over two is five theta; eight theta over two is four theta.
\[ \sin(9\theta) - \sin(-\theta) = 2\cos(4\theta)\sin(5\theta) \]
Notice what the first move bought
Why: Going straight in with alpha nine theta and beta minus theta gives the half-sum four theta and the half-difference five theta, which is the same answer with the roles of the two factors swapped. Clearing first makes the bookkeeping shorter.
\[ \frac{9\theta + (-\theta)}{2} = 4\theta, \qquad \frac{9\theta - (-\theta)}{2} = 5\theta \]
Check at theta equal to eighteen degrees
Why: The left side is the sine of one hundred and sixty-two minus the sine of minus eighteen, that is 0.309 plus 0.309, namely 0.618. The right side is twice the cosine of seventy-two times the sine of ninety, that is twice 0.309 times 1, also 0.618.
\[ 0.309 + 0.309 = 0.618 = 2(0.309)(1) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 785 — Exercise 83; answer p. 789
This one hides its difficulty in the question rather than the answer, which is a favourite trick of exercise sets.
Before any sum-to-product formula, look at what you actually have. The second term is the sine of a negative angle, so the odd identity applies immediately and subtracting it becomes adding the sine of the positive angle. Now you have a clean sum of two sines and the top signs apply.
The fourth step shows what happens if you skip that first move and feed the negative angle straight into the formula. You get the same answer with the two factors swapped, because the half-sum and the half-difference trade values when one angle changes sign. It is not wrong, just longer. The general lesson is to clear every negative angle you can see before choosing a formula, not after.
Sorting
Sort into buckets
A conversion has left each of these behind. Sort them by whether clearing the negative angle changes the value or only the appearance.
Six leftovers from six conversions. For each one, ask a single question: is the negative angle inside a cosine or inside a sine?
Inside a cosine it is cosmetic. The expression already has the right value and rewriting it changes only how it looks. You should still rewrite it, because an unfinished answer invites the reader to make the correction for you and possibly to make it wrong, but nothing is at stake numerically.
Inside a sine it is genuine. A minus sign is waiting to come out, and if you stop there anyone reading your work, including you tomorrow, is liable to drop it. Look closely at the last item, where a minus sign is already sitting in front. The new one meets it and the two cancel, which is precisely what happened in Exercise 82, so a genuine change does not always mean the final sign changes.
Error analysis
Annotate
On: \( \sin(2x) - \sin(6x) = 2\sin(-2x)\cos(4x) \)
\[ \sin(2x) - \sin(6x) = -2\sin(2x)\cos(4x) \]
This is a different species of error from the earlier one, and it is worth being clear about the difference.
Nothing on that line is false. Substitute any value of x and both sides agree, because the sine of minus two x really does equal what it equals. What is wrong is that the conversion is unfinished, and an unfinished conversion is not an answer.
The last annotation says why this matters more for sines than for cosines. A reader who simplifies a leftover cosine in their head gets it right automatically. A reader who simplifies a leftover sine in their head has to remember to move a minus sign, and that is exactly the step that gets skipped. Finish the line yourself and nobody has to.
Section
Part 5
Concept
Add two waves whose frequencies are almost equal. Sum-to-product says the result is a single wave at the average frequency, multiplied by a slow wave at half the gap.
\[ \sin(2\pi f_1 t) + \sin(2\pi f_2 t) = 2\sin\!\left(2\pi\tfrac{f_1+f_2}{2}t\right)\cos\!\left(2\pi\tfrac{f_1-f_2}{2}t\right) \]
Figure (svg): The sum of two sine waves at twenty and twenty-four hertz plotted over half a second. The rapid oscillation is enclosed by a dashed envelope of plus and minus twice the cosine of two pi times two t, which pinches to zero twice in that half second.
The ear hears the fast factor as the pitch and the slow factor as a throbbing in the loudness. Because loudness does not care about sign, the throb happens twice per cycle of the slow factor.
This is the first of several applications, and it is the one you can hear. Two notes that are nearly but not quite in tune do not sound like two notes. They sound like one note that pulses.
The formula explains both halves of that. The fast factor sits at the average of the two frequencies, so the pitch you hear is a single note midway between them, not a chord. The slow factor sits at half the gap, and because the gap is tiny compared with the frequencies themselves, that factor changes over seconds rather than thousandths of a second. It is the loudness, not the pitch.
The detail that catches everyone is the factor of two in the beat rate. The envelope is the size of the slow cosine, and size does not care about sign, so the envelope pinches to zero twice per cycle of that cosine, not once. Hence the two red dots in the picture within one cycle, and hence four beats a second for a gap of four hertz.
Worked example
A piano tuner strikes a fork at 256 hertz and a string sounding 260 hertz. Predict the pitch heard and the rate of the throbbing.
\[ y = \sin(2\pi \cdot 256 t) + \sin(2\pi \cdot 260 t) \]
Compute the half-sum frequency
Why: Add the two frequencies and halve.
\[ \frac{256 + 260}{2} = 258 \text{ hertz} \]
Compute the half-difference frequency
Why: Subtract and halve.
\[ \frac{256 - 260}{2} = -2 \text{ hertz} \]
Apply sine plus sine and clear the negative angle
Why: Cosine is even, so the minus inside the cosine disappears.
\[ y = 2\sin(2\pi\cdot 258t)\cos(2\pi\cdot 2t) \]
Read off the pitch
Why: The fast factor sets the note: 258 hertz, midway between the two, and the listener hears one note, not two.
\[ \text{pitch} = 258 \text{ hertz} \]
Read off the beat rate
Why: The envelope is the size of the slow cosine, which reaches zero twice per cycle, so the loudness dips twice per cycle of a 2 hertz wave.
\[ \text{beats} = 2 \times 2 = 4 \text{ per second} \]
Check against the difference of the two frequencies
Why: Four is exactly 260 minus 256, so the rule a tuner actually uses, that the beat rate equals the gap in hertz, comes straight out of the formula. As the string is tightened the beats slow, and silence means the gap has closed.
\[ f_2 - f_1 = 260 - 256 = 4 \text{ per second} \]
Every step here is a step you have done twice already; only the letters have changed into frequencies. That is worth noticing, because it is what makes trigonometry worth learning: the same conversion serves a homework exercise and a piano tuner.
The negative half-difference appears again, and again it is harmless, because it lands inside a cosine. If you had written the two frequencies in the other order you would have got a positive two hertz and the same answer.
The final check is the point of the slide. The beat rate comes out as four per second, which is exactly the difference between the two frequencies. That is the rule tuners actually use, and it is a very sharp instrument: nobody can hear the difference between 440 and 440.5 hertz as a pitch, but counting one throb every two seconds is easy. The mathematics has turned an inaudible difference into a countable one.
Tweak it
Parameter explorer
The curve is the sum of a cosine at frequency one and a cosine at frequency k. Slide k from 1.6 down towards 1 and watch the envelope, not the individual wiggles. What happens to the length of one throb as the two frequencies close up?
\[ \cos x + \cos({k}x) \]
Drag the slider slowly and watch the outline of the curve rather than the individual wiggles. The two frequencies here are one and k, so the fast factor sits at their average and the envelope sits at half their difference.
At the right-hand end of the slider the gap is largest, so the envelope is at its fastest and you can count several throbs across the window. As you bring k down towards one, the half-difference shrinks and each throb stretches out, until only one or two fit in the whole picture.
Keep going and the throbs all but disappear, because the half-difference frequency is heading for zero and its period is heading for infinity. At exactly equal frequencies there is no throb at all: the sum is one wave of twice the height, which is what the formula gives when the half-difference angle is zero and its cosine is one. That limiting case is the one a piano tuner is listening for.
Estimation
Predict first
A tuner hears one throb every two seconds against a 440 hertz fork. How far off is the string?
Correct: Half a hertz
Why: The beat rate in throbs per second equals the difference in frequency. One throb every two seconds is half a throb per second, so the string is half a hertz away from 440: either 439.5 or 440.5, and which one is settled by tightening slightly and hearing whether the beats speed up or slow down. Notice how sensitive the method is — the ear cannot hear a half-hertz pitch difference at all, but it can easily count a throb every two seconds.
\[ \text{beats per second} = |f_2 - f_1| = 0.5 \]
Answer before you read the options carefully, then check yourself against the reasoning.
The only fact you need is the one the worked example just established: the number of throbs per second equals the difference between the two frequencies in hertz. One throb every two seconds is half a throb per second, so the gap is half a hertz.
Two things are worth taking away. First, the method does not tell you which side of 440 the string is on, so a tuner nudges the tension slightly and listens for whether the beats speed up or slow down. Second, look at the sensitivity. A half-hertz pitch difference is far below what anyone can hear directly, but as a beat it is a slow, obvious pulse. Converting a sum into a product has moved information from a place your ear cannot reach into a place it can.
Real world
Figure (svg): Two stem plots. On the left, a single stem at five kilohertz on an axis running to ten, the audio tone. On the right, on an axis from 990 to 1010 kilohertz, a dashed ghost stem at 1000 for the bare carrier and two solid half-height stems at 995 and 1005, the frequencies actually transmitted.
Discussion prompt
An AM transmitter multiplies a carrier at 1000 kilohertz by an audio tone at 5 kilohertz. Convert the product to a sum, say which frequencies are actually radiated, and explain why a station is allocated a band rather than a single frequency.
The same mathematics, run in the other direction. Beats came from a sum; broadcasting starts with a product.
A transmitter multiplies a high-frequency carrier by the audio it wants to send. Theorem 10.20 says the result contains nothing at the carrier frequency and nothing at the audio frequency. What it contains is a pair of tones, one either side of the carrier, separated from it by exactly the audio frequency, and each at half the strength.
Now think about real audio, which is not one tone but a whole range of them at once. Each one produces its own pair, so the station occupies a band of frequencies reaching out on both sides of its carrier by the highest audio frequency it transmits. That width is why the dial has stations at intervals rather than packed together, and why limiting audio quality lets more stations fit. A regulation about channel spacing is a product-to-sum formula wearing a hat.
Worked example
This is the fact the whole of Fourier analysis rests on. Take two sines at different whole-number frequencies and average their product over a full turn.
\[ \text{average of } \sin(3x)\sin(5x) \text{ over } 0 \le x \le 2\pi \]
Convert the product to a sum
Why: Matched product, so cosines, with the minus sign in the middle.
\[ \sin(3x)\sin(5x) = \tfrac12\cos(2x) - \tfrac12\cos(8x) \]
Average each cosine separately
Why: A cosine at a whole-number frequency completes whole cycles over the turn, and a whole cycle has as much above the axis as below.
\[ \text{average of }\cos(2x) = 0, \qquad \text{average of }\cos(8x) = 0 \]
Conclude for different frequencies
Why: Half of nothing minus half of nothing.
\[ \text{average of } \sin(mx)\sin(nx) = 0 \quad (m \ne n) \]
Figure (svg): The curve of sin 3x times sin 5x over one full turn, with the regions above the axis shaded green and the regions below shaded red. The green and red areas are visibly equal, so the average over the turn is zero.
Now let the two frequencies agree
Why: With m equal to n the difference angle is zero, and the cosine of zero is one, which does not average away.
\[ \sin^2(mx) = \tfrac12\cos 0 - \tfrac12\cos(2mx) = \tfrac12 - \tfrac12\cos(2mx) \]
Check both cases numerically
Why: Averaging the product over ten thousand points of a full turn gives 0.0000 for frequencies 3 and 5, and 0.5000 for frequency 3 against itself, exactly as the two lines above predict.
\[ \overline{\sin 3x\sin 5x} = 0, \qquad \overline{\sin^2 3x} = \tfrac12 \]
This slide is the deepest thing in the lesson, and it is worth more than one reading. Everything in signal processing that separates a mixture into its ingredients rests on it.
The claim is that if you multiply two sine waves at different whole-number frequencies and average the result over a full turn, you get exactly zero, but if the two frequencies are the same you get one half. The proof is one product-to-sum conversion and one obvious fact: a cosine at any whole-number frequency has as much area above the axis as below over a whole turn, so it averages to nothing.
When the frequencies differ, both terms of the converted sum are such cosines and both average away. When they agree, the difference angle is zero, and the cosine of zero is the constant one, which does not average away. That surviving constant is the one half.
So multiplying a signal by a test wave and averaging asks the signal a question: how much of you is at this frequency? Everything else answers zero. That is what a Fourier coefficient is, and the green and red areas cancelling in the picture are what it looks like when the answer is none.
Worked example
A sum of a sine and a cosine does not look like it fits any of the four formulas. A cofunction identity makes it fit.
\[ \sin(\theta) + \cos(\theta) \]
Rewrite the cosine as a sine
Why: The cofunction identity from lesson 10.4a: a cosine is the sine of the complementary angle.
\[ \cos(\theta) = \sin\!\left(\tfrac{\pi}{2} - \theta\right) \]
Now it is a sine plus a sine
Why: Two sines, so the formula with the top signs applies.
\[ \sin(\theta) + \sin\!\left(\tfrac{\pi}{2}-\theta\right) \]
Compute the half-sum
Why: The two thetas cancel, leaving a constant.
\[ \frac{\theta + \left(\tfrac{\pi}{2}-\theta\right)}{2} = \frac{\pi}{4} \]
Compute the half-difference
Why: Two thetas minus pi over two, all over two.
\[ \frac{\theta - \left(\tfrac{\pi}{2}-\theta\right)}{2} = \theta - \frac{\pi}{4} \]
Assemble and simplify the constant
Why: Twice the sine of pi over four is twice the square root of two over two, which is the square root of two.
\[ \sin\theta + \cos\theta = 2\sin\!\left(\tfrac{\pi}{4}\right)\cos\!\left(\theta - \tfrac{\pi}{4}\right) = \sqrt{2}\,\cos\!\left(\theta - \tfrac{\pi}{4}\right) \]
Figure (svg): The curve of sin t plus cos t over minus pi to two pi, a single clean sinusoid, with the curve of the square root of two times cos of t minus pi over four drawn on top of it as a thick dashed line, and horizontal dashed lines at plus and minus 1.414 touching the peaks.
Check at theta equal to fifteen degrees
Why: The left side is 0.259 plus 0.966, namely 1.2247. The right side is 1.4142 times the cosine of minus thirty degrees, that is 1.4142 times 0.8660, also 1.2247. The picture agrees: the peak sits at pi over four and reaches 1.414.
\[ 0.259 + 0.966 = 1.2247 = \sqrt2\cos(-30^\circ) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 785 — Exercise 84; answer p. 789
At first glance this is not in the family at all, because the sum-to-product formulas want two cosines or two sines, and this is one of each. The first step fixes that with a cofunction identity from lesson 10.4a.
Watch what happens to the angles. Because the two are complementary, the thetas cancel in the half-sum and it becomes the constant pi over four. So the first factor is not a function of theta at all; it is a number, twice the sine of pi over four, which is the square root of two. All the theta dependence is in the second factor.
That makes the whole thing a single shifted cosine, and the picture shows it: one clean sinusoid of amplitude about 1.414, peaking at pi over four rather than at zero. This is the shape Section 11.1 needs when it asks for the amplitude and phase shift of a sinusoid, and sum-to-product is how you find them. Exercise 85 is the same question for a cosine minus a sine, and it comes out as a shifted sine.
Worked example
This is the move that makes a whole class of equations solvable, and it is the reason Theorem 10.21 appears in a trigonometry course at all.
\[ \cos(6x) + \cos(2x) = 0 \]
Notice that a sum cannot be split
Why: Setting one term to zero is not allowed: a sum vanishes when the two terms are opposites, which need not make either of them zero.
\[ \cos 6x = 0 \text{ or } \cos 2x = 0: \quad \text{not justified here} \]
Convert the left side to a product
Why: The half-sum is 4x and the half-difference is 2x, both positive.
\[ \cos(6x) + \cos(2x) = 2\cos(4x)\cos(2x) \]
Set the product to zero
Why: Now the split IS justified: a product of real numbers is zero exactly when one of its factors is.
\[ 2\cos(4x)\cos(2x) = 0 \;\Longrightarrow\; \cos(4x) = 0 \text{ or } \cos(2x) = 0 \]
Solve the two basic equations
Why: Each is the kind solved in lesson 10.2b.
\[ 4x = \tfrac{\pi}{2} + k\pi \quad\text{or}\quad 2x = \tfrac{\pi}{2} + k\pi \]
Write the solutions
Why: Divide each by its coefficient.
\[ x = \tfrac{\pi}{8} + \tfrac{k\pi}{4} \quad\text{or}\quad x = \tfrac{\pi}{4} + \tfrac{k\pi}{2} \]
Figure (svg): The curve of cos 6x plus cos 2x over zero to pi, crossing the axis six times, with the two factor curves two cos 4x and cos 2x drawn dashed. Each crossing of the sum sits exactly on a crossing of one of the factors, marked with a dot.
Check one solution of each factor
Why: At x equal to pi over eight, the cosine of three pi over four plus the cosine of pi over four is minus 0.707 plus 0.707, namely zero. At x equal to pi over four, the cosine of three pi over two plus the cosine of pi over two is zero plus zero. Both crossings appear in the picture.
\[ \cos\tfrac{3\pi}{4}+\cos\tfrac{\pi}{4} = 0, \qquad \cos\tfrac{3\pi}{2}+\cos\tfrac{\pi}{2} = 0 \]
This is the reason Theorem 10.21 is taught in trigonometry rather than being left to calculus, and the first step is the one to understand.
There is one thing you can do with a product that you cannot do with a sum: set it to zero one factor at a time. A product of real numbers is zero exactly when one of its factors is, which is the zero product property you used on quadratics. A sum has no such property. Two cosines can add to zero with neither of them anywhere near zero, and the picture shows that happening at every crossing marked with a purple dot.
So the conversion is not a simplification in any absolute sense. It is a change of form that unlocks a technique. Once the left side is a product, the equation splits into two basic equations of the kind you solved in lesson 10.2b, and the picture confirms the arithmetic: every crossing of the solid curve sits on a crossing of one dashed factor or the other, and there are no others.
Worked example
Real verification problems combine the two directions. Convert the top and the bottom, then look for what cancels.
\[ \frac{\cos(3t) - \cos(t)}{\sin(3t) + \sin(t)} = -\tan(t) \]
Convert the numerator
Why: Cosine minus cosine: half-sum 2t, half-difference t, and the leading minus sign.
\[ \cos(3t) - \cos(t) = -2\sin(2t)\sin(t) \]
Convert the denominator
Why: Sine plus sine: the sine takes the half-sum 2t and the cosine takes the half-difference t.
\[ \sin(3t) + \sin(t) = 2\sin(2t)\cos(t) \]
Form the quotient
Why: Both conversions produced the same two angles, which is what makes the cancelling possible.
\[ \frac{-2\sin(2t)\sin(t)}{2\sin(2t)\cos(t)} \]
Cancel the common factors
Why: The 2 and the sine of 2t appear top and bottom.
\[ = -\frac{\sin(t)}{\cos(t)} \]
Recognise the quotient
Why: Sine over cosine is the tangent, by the quotient identity.
\[ = -\tan(t) \]
Check at t equal to twenty degrees
Why: The numerator is the cosine of sixty minus the cosine of twenty, that is 0.5 minus 0.940, namely minus 0.440. The denominator is the sine of sixty plus the sine of twenty, that is 0.866 plus 0.342, namely 1.208. Their quotient is minus 0.3640, and minus the tangent of twenty degrees is also minus 0.3640.
\[ \frac{-0.440}{1.208} = -0.3640 = -\tan 20^\circ \]
Verification problems in the exercise set rarely tell you which family to reach for, so here is what the decision looks like in practice.
You have a quotient, and quotients are simplified by cancelling. Nothing cancels between a sum on top and a sum on the bottom, so both need to become products. Convert each with the appropriate sum-to-product formula and look at what appears.
The lucky thing, which is not luck at all, is that both conversions produce the same two angles, because the numerator and the denominator started from the same pair. So the factor of two and the sine of two t appear in both and cancel, and what survives is a sine over a cosine, which is a tangent by the quotient identity. The leading minus from the cosine-minus-cosine formula is what makes the answer negative, and if you had lost it you would have proved the wrong identity.
Section
Part 6
Pattern
Figure (svg): A decision diagram. A starting box asks what is blocking you. One branch leads through a product that will not integrate or average to the product-to-sum family and out to a sum of first powers. The other branch leads through a sum that will not factor to the sum-to-product family and out to a product of two factors. A final box reminds you to clear negative angles.
This is the slide to photograph. The diagram is a decision procedure, and the list under it is the arithmetic once the decision is made.
Everything hangs on the first box. Neither direction is a simplification in any absolute sense, and asking which form is simpler is the wrong question. Ask what you are stuck on. A product you cannot integrate or average wants to become a sum. A sum you cannot factor, or from which you cannot read an amplitude, wants to become a product.
The last line is the one that saves the most marks in practice. These six formulas differ from each other only in signs and in which angle goes where, so your memory of them is exactly as reliable as a coin toss under pressure. One substitution of a convenient value into both sides catches every slip of that kind, and it takes under a minute.
Comparison
Comparison matrix
| product to sum | sum to product | |
|---|---|---|
| what goes in | a product of two functions | a sum or difference of two functions |
| what comes out | a sum of two first powers | twice a product of two functions |
| leading constant | one half | two |
| the two new angles | the difference and the sum | the half-sum and the half-difference |
| the formula with a minus in it | sin α sin β | cos α − cos β |
| what it unblocks | integrating, averaging | factoring, solving, reading an amplitude |
Fill the blanks from memory before revealing them. Every row is a feature of one family with its mirror image in the other, so if you can recall one side you can reconstruct the other.
The inversion is complete and it is not a coincidence: these two families are the same three equations read left to right and right to left. One half becomes two. A sum and a difference of the inputs become a half-sum and a half-difference. A product goes in where a product comes out.
The only row you have to learn independently is the fifth one, because there is no symmetry connecting the two exceptions. In one direction the odd formula out is the product of two sines, which carries a minus in the middle. In the other it is the difference of two cosines, which carries a minus in front. Those are different positions and different reasons, and you simply have to know both.
Ranking
Put in order
Put the steps of a complete conversion in order.
Why: The first step catches things like a sine plus a cosine, which need a cofunction identity before any formula applies. The two angles are computed once and then placed, and the sign bookkeeping comes before the placing, not after. Clearing the negative angle is part of the answer, not an optional tidy-up, and the substitution check is worth its thirty seconds every single time.
The order matters more than it looks, and two of the six steps are the ones people skip.
The first is the check that you actually have the right kind of expression. A sine plus a cosine, as in Exercise 84, needs a cofunction identity before any of the four formulas can touch it, and noticing that after you have already written down two angles wastes the work.
The other commonly skipped step is the second to last. Clearing a negative angle is part of producing the answer, not an optional tidy-up afterwards, and with a sine it changes the value of what you wrote. Put it in the routine and it stops being something you might forget.
The final substitution check is worth its half minute on every single problem in this lesson. Six formulas that differ only in signs and angle placement is precisely the situation where a numerical check earns its keep.
Check
Check your understanding
Which expression equals the product of the cosine of 5x and the cosine of 3x?
Answer: A
Why: A matched product of two cosines gives cosines with a plus in the middle, at the difference angle 2x first and the sum angle 8x second. At x equal to zero the product is 1 and half of cos 0 plus cos 0 is also 1.
Two cosines multiplied, so the answer must be cosines, and matched products take the plus sign in the middle. That eliminates two options before any angle is computed.
Between the two that survive, the difference angle comes first. Five x minus three x is two x, so the cosine of two x leads, and the sum angle eight x follows. The option with the angles reversed is the negative of the truth, which you can see by putting x equal to zero: the product is one, and the reversed version gives nothing sensible.
The fourth option is there to catch a different confusion. It has a leading two and half-sum and half-difference angles, which are the markings of the other family entirely. If you chose it, reread the question: it asked you to convert a product, and that answer converts a sum.
Check
Check your understanding
Which expression equals the sine of 6x minus the sine of 2x?
Answer: B
Why: For a sine difference the sine takes the half-difference, which is 2x, and the cosine takes the half-sum, which is 4x. At x equal to pi over eight the left side is 0.707 minus 0.707, namely zero, and the right side is twice 0.707 times zero, also zero; at x equal to pi over twelve both sides are 0.5.
A difference of two sines, so the bottom signs apply throughout and the roles swap: the sine takes the half-difference and the cosine takes the half-sum.
Six x minus two x over two is two x, which goes in the sine. Six x plus two x over two is four x, which goes in the cosine. The wrong option with those the other way round is the answer to a sum of two sines, and the two expressions disagree almost everywhere.
If you were unsure between the first two, a substitution settles it instantly. At x equal to pi over twelve the original is one half, and only one of the two candidates gives one half. Notice that a badly chosen test value, such as pi over eight, makes both come out as zero and tells you nothing, which is a general warning about checks: pick a value where the two candidates must differ.
Check
Check your understanding
You must solve the equation cos 5x + cos x = 0. What is the correct first move?
Answer: C
Why: A product can be set to zero one factor at a time, so converting the sum into 2 cos 3x cos 2x turns one unsolvable equation into two basic ones. The half-sum is 3x and the half-difference is 2x.
This is the question the whole lesson has been building towards, and it is about strategy rather than arithmetic.
The equation is a sum set equal to zero, and a sum set equal to zero gives you nothing to work with, because two terms can cancel without either being zero. Converting the left side into a product changes that completely: the zero product property then splits one hard equation into two easy ones.
Consider why the wrong answers are tempting. Setting each cosine to zero is exactly the move the zero product property would license if you had a product, and importing it to a sum is the single most common error in Section 10.7. There is no product to convert, so product-to-sum is out. And squaring is the reflex from radical equations, where it helps; here it introduces false solutions and leaves a product behind anyway.
Explain it to yourself
Discussion prompt
In two or three sentences, explain to someone who has just met these formulas why the sum-to-product angles are half-sums and half-differences, while the product-to-sum angles are plain sums and differences. Use the fact that the two families are inverse to each other.
Write your answer out before revealing this one, because explaining it is what turns the fractions from something remembered into something understood.
The core of a good answer is that the two families are inverse to each other, and that a product-to-sum formula reports the difference and the sum of what you feed it. If you want the report to come out as the two angles you already have, you have to feed in the pair whose difference and sum are those two angles, and solving that small system gives the half-sum and the half-difference.
A good answer should also mention the constants, because the same reasoning explains them. Going one way you divide by two; going back you multiply by two. Nothing in Theorem 10.21 is an extra fact to memorise. It is Theorem 10.20 with the inputs relabelled, and that is worth being able to say out loud.
Exit ticket
\[ \text{(a) } \sin(7x)\sin(4x) \qquad \text{(b) } \cos(7x) - \cos(3x) = 0 \]
Discussion prompt
Convert (a) to a sum and (b) to a product, clear any negative angle, and say in one line what each conversion was for.
Two conversions, one in each direction, and one sentence of purpose for each. Do them on paper before revealing anything.
Part a is a matched product of two sines, so cosines come out with a minus in the middle, the difference angle three x leading. Both angles are positive here, so there is no cleanup. Part b is a difference of cosines, which is the one formula with a leading minus, and the half-sum and half-difference are five x and two x.
The sentence of purpose is the part worth grading yourself hardest on. If you can convert fluently but cannot say which obstacle each conversion removes, you will convert in the wrong direction under exam pressure. Products block integration and averaging; sums block factoring and block reading off an amplitude.
Recap
| if the question says | your first move |
|---|---|
| write this product as a sum | difference angle first, leading constant one half |
| write this sum as a product | half-sum and half-difference, leading constant two |
| integrate or average a product | product to sum, then integrate the first powers |
| solve a sum of two terms equal to zero | sum to product, then set each factor to zero |
| a negative angle is left over | cosine: drop it; sine: pull the minus out front |
\[ \cos\alpha\cos\beta = \tfrac12\big[\cos(\alpha-\beta) + \cos(\alpha+\beta)\big] \]
\[ \cos\alpha + \cos\beta = 2\cos\!\left(\frac{\alpha+\beta}{2}\right)\cos\!\left(\frac{\alpha-\beta}{2}\right) \]
That is the end of Section 10.4 and of the identities. Section 10.5 leaves them behind and starts graphing the six functions, and Section 10.7 comes back for the sum-to-product formulas when it solves equations by factoring.
OpenStax Algebra and Trigonometry 2e, §9.4 Sum-to-Product and Product-to-Sum Formulas §9.4 — the same two families in OpenStax
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 780-781 — Theorems 10.20 and 10.21, and Example 10.4.6
The table is the lesson in six rows. Notice that the left column is written in the language of questions you will actually be asked, not in the language of formulas, because choosing the direction is the part that is hard and substituting is the part that is easy.
Two formulas are printed underneath, one from each family. If you remember only those two you can rebuild the other four: the sine-times-sine formula by subtracting the cosine identities instead of adding, the mixed one by using the sine identities, and the remaining sum-to-product formulas by running each of those backwards through the half-sum and half-difference substitution.
That closes Section 10.4. Section 10.5 changes subject entirely and starts graphing the six circular functions, and you will find that the identities you have built here make those graphs far easier to read. Section 10.7 comes back for the sum-to-product formulas specifically, because they are the only way to solve a whole class of equations.
Want this taught 1-on-1? Alexander tutors Trigonometry — $55/session, free consultation.