What kuky said will work, though it creates more manual work (but not massively more work) having to amplify the music sections independently. Also, compression and/or volume envelopes will do the job in this case. I usually find it best, though more time consuming, to compress (or adjust with volume envelope) each applause section independently. You won't impact the quality of the music because you're not applying compression or VE to the music itself, only the applause sections. I'd compress/VE the applause sections so they're sufficiently lower than the music, and then peak normalize afterwards (this removes the need to amplify the audio manually, as in kuky's suggestion). For applause, I've found a very short attack setting for compression works best. FWIW, I think VE sounds better, but is more time consuming, than compression, and in some cases using both compression and VE works best. Even though I find it a bit more time consuming, VE can be worth it, and the learning curve is fairly quick.