I was reading the Chapter $4$ of Hubbard's Teichmüller Theory where he begins to introduce quasiconformal mappings. Here he writes a linear tranformation $T: \mathbb{C} \rightarrow \mathbb{C}$ as $T(u)=au+b\bar{u}$ and then notes that the determinant is $$\det{T}= |a|^{2}-|b|^{2}$$
He justifies this by saying that if we let $u=re^{i\theta}$, $a:=|a|e^{i\alpha}$, $b:=|b|e^{i\beta}$, then expanding $|T(u)|=|au+b\bar{u}|=1$, we get an ellipse with minor axis of semi-length $\frac{1}{|a|+|b|}$ and major axis $\frac{1}{||a|-|b||}$ and then noting that the ratio of the unit circle to the area of this ellipse is $|a|^{2}-|b|^{2}$.
Anyways the actual equation for the ellipse is as follows:
$$\bigg|(|a|+|b|)\cos\big(\theta+\frac{\alpha-\beta}{2}\big)+i(|a|-|b|)\sin\big(\theta+\frac{\alpha-\beta}{2}\big) \bigg|=\frac{1}{r}$$
I was wondering why the $r$ doesn't come up in the area or the semi-length of the major and minor axis? (Sorry, if this is a dumb question)
The $r$ comes in as the semimajor/semiminor axes' lengths when you expand $|T(re^{i\theta})|=1$, i.e., the minimum $r_\min$ is the semiminor axis length and $r_\max$ is the semimajor axis length.