When downward compression is applied with a slower attack, I'm struggling to understand how the dynamic range of the signal as a whole is reduced. If the signal is a bass hit for example, where the signal crosses the threshold very quickly and the loudest part of the original signal is let through by the slower attack before compression is applied, then surely that peak will still exist in the compressed signal. Even if the compressor then clamps down and reduces the gain on the rest of the signal that has crossed the threshold, that original part of the signal isn't compressed, so surely the dynamic range of the signal as a whole isn't reduced, because neither the loudest part of the signal (the initial part of the hit that slipped through) or the quietest part of the signal have changed in volume.
So would it be true to say that compression only reduces the dynamic range of a signal if the loudest part of the signal doesn't happen before the compressor kicks in (due to a slower attack). Because in this instance it would appear that the Perceived dynamic might actually be increased because the gain of everything other than the peak will be reduced) which might well be a desirable effect in the case of a kick.