I recorded all day on the 4th of July so I just did this. There are probably other tools you could play with, but in my experience you have two basic choices; volume limiter and spike-by-spike.
So, on my recent recordings, the levels of the pops of the fireworks were significantly higher than the prevailing music so it was easy for the software to hone in on the offending sections. I applied a hard limiter with the peak level being a db or so above the level of the prevailing music and applied a zero or negative amplification level. It worked like a champ.
If the pops don't spike above the level of the prevailing music, it's alot more problemmatic. In my experiene, you'll probably need to use the hunt and peck method and minimize each spike is you really want to fix it. Otherwise, you might try an envelope filter to isolate the frequency of the pops and see if it can find the pops. Unfortuntately, I haven't had real good luck with this tool on fireworks. Similarly, I've never had much luck with the click or pop filters on fireworks, though they can work great on diginoise.
Hint: Whichever method you use, assuming you recorded with a bit of headroom, do it before normalization so that you can still see the spikes in case you need to zoom in and do any manual manipulation. If you normalize first, it might be harder to find the spikes when you zoom in.
Hint No. 2: I've found that indoors with pyro or any loud bang, these spikes are alot tougher. The reverb from the loud noise decays through the room and the pop just isn't short enough. Outdoors, is much easier.
Topic: Dealing with pyrotechic spikes from fireworks in a wave file. (Read 2285 times)