Any good parametric will do the job. For example, I just happened to be reviewing the Sonitus parametric EQ for an upcoming issue of Keyboard, when I decided to take a break and check out some of the posts. After reading yours, I took a piece of music with heavy peaks at 20-30 Hz and applied the following:
Stage 1: bandpass response, 25 Hz, Q of 1.4, -18 dB gain
Stage 2: bandpass response, 20 Hz, Q, of 5.6, -18 dB gain
I then did a frequency analysis of the original and processed files, which confirmed that this totally took out 20 Hz, but left everything over 35 Hz pretty much intact. I was also able to increase the level of the processed file by about 1-2 dB without clipping by getting rid of the subsonics.
