This bit covers a widely used professional technique for taming unruly lead vocal dynamics: chaining two compressors in series rather than relying on a single unit to do all the work. The host walks through why a solo compressor so often forces a trade-off, either clamping down hard enough to kill the life in a performance, or being gentle enough to leave peaks and dips still fighting for balance.
The solution is to run a fast FET compressor first, letting it catch transient peaks before they cause problems, then feed the signal into a slower optical compressor to gently even out the overall energy. Each compressor handles a different part of the problem, so neither one has to work hard enough to introduce artifacts or suck the character out of the vocal.
Splitting the job also makes setting attack and release times more straightforward. With each compressor assigned a specific task, you can dial in time constants that match that one job, rather than hunting for a single set of settings that somehow handles both peak control and overall leveling at once.