There are several ways to interpret the question. What I think you might be asking is that you have a sequence of $n$ letters in an alphabet $\Sigma$ where $\left| \Sigma \right| = 80$. You want to store this in as few as possible bits. We will assume that the letters in the alphabet are uniformly distributed.
The information-theoretic amount of space required to store this is $n \log_2 \left| \Sigma \right|$ bits. Using arithmetic coding, you can do this in linear time, using $O(\log n)$ bits of intermediate space. (Remember, that's the logarithm of the number of symbols, in bits! If the size of the sequence fits in a machine word, the intermediate storage required is a constant number of machine words at the most.)
So that's pretty good. But what about if we want random access?
It turns out that it can be done. The first technique to do it was only discovered about four years ago. We can store the sequence in $n \log_2 \left| \Sigma \right|$ bits, such that reading or writing any entry takes $O(1)$ time. If you think about it, this is a remarkable result, because it means that a computer which works with any radix is, in a sense, equivalent to a binary one.
Here's the paper: Yevgeniy Dodis, Mihai Pătraşcu, and Mikkel Thorup, An Alternative to Arithmetic Coding with Local Decodability, STOC 2010.
By the way, remember the name Mihai Pătraşcu. He was and is the closest thing we have to a modern-day Évariste Galois. He died very young, of a brain tumor at the age of 29. But in his short career as a computer scientist, his work revolutionised the field of analysis of algorithms in ways that will take decades to fully understand.