I have a speech I’m attempting to clean up and I’m new to Audacity. Noise reduction is great. However, the audio has a number of low volume sections where someone is speaking that then spike up once people start clapping very loudly. Amplify seems to work well, but it doesn’t seem to discriminate. It increases the gain for everything selected and it would take a very long time to manually go through the full clip. Is there a function that allows for amplification of only portions of the audio that are low volume, but then either leaves the high volume sections alone or possibly the function allows you to set a limit?
That would be an “upward compressor” … Dynamic range compression - Wikipedia
Audacity is shipped with a compressor and (soft) limiter, either of which will do your job.
I have a speech I’m attempting to clean up
“Help me Clean Up.” There’s the warning sign.
Did the applause overload? View > Show Clipping. Did the screen fill up with little red bars during the applause? If so, those areas will always sound crunchy, harsh and crackly. Overload is pretty deadly.
There is a program called Level Speech.
And there is an older program called Chris’s Compressor.
Chris appears as “Effects > Compress dynamics…” when you use it. The only shortcoming I know of is needing to leave some sound after the valuable show. Chris doesn’t like to fall off the end of the show. After Chris is done, you can cut off anything extra.
Koz
Chris may win here. I use it to even out podcasts so I can hear them when I’m out hiking. It’s a look-ahead compressor and will tailor, up or down, the corrections on the fly. One show is unlistenable without this processing. Two guys, one mumbles in his beer and the other has a thermonuclear laugh.
Mumbmemmmmmbbmmbmmbmm HA! HA! HA! HA! HA!
Most annoying.
I usually change the first setting “Compress Ratio” from the default 0.5 to a stiffer 0.77. That’s it. I just let it fly.
Koz
Thanks for the help here. That guidance will help me. Actually, it doesn’t appear there was any clipping, but the audio isn’t good.
The mistake I made was using the audio from an action cam. For these family videos I typically put music in the background so sound from the action cam isn’t needed, but in this case I wanted to capture, but I only figured that out in hindsight. If I would have thought things through on the front-end I could have simply thrown down an iphone or just about anything else to do the audio recording and it would have turned out good enough.
In case you are curious here is a link to the audio file. If anyone would like to take a stab at cleaning up that would be great. My skills with Audacity leave something to be desired. 
https://drive.google.com/drive/folders/1OT_ll_skjSRR2Ghu35QDoEmrULCmTJGA?usp=sharing
A lot of hiss on that : inevitably the compressor will increase the hiss when it selectively amplifies the quiet sections, see …
In addition to compression I also had to use Audacity’s envelope tool, which manually changes the volume of sections …