Thanks! In a sense, the method I’m proposing here is a sort of “dynamic EQ”, in that the response of the filter changes with the dynamics of the input signal. The advantage here is that because we assume the input sound is a drum, we can exploit certain physical properties that we know are true for drum sounds. This is what allows us to set a “decay time” rather than an attack and release. The modal analysis tells us what the original decay time is at the ringing frequency, so we can just pick a new decay time that matches up with the rest of the drum.

The other advantage is that - in the future - much of this process could be automated. So rather than having to find the problem frequencies and fix them yourself, the modal analysis could find those frequencies for us, and automatically generate the filters needed to “fix” the drum sound.

Hope this answers your question!

-Jatin