There, I have some spikes. I apply the Envelope tool. As soon as I put my first point, the entire recording seems to shrink to that first point. That is primarily what is confusing me. Is that then lowering everything?
Are you using Effect > Compressor anywhere in this process? Taming those tall spikes are pretty much what it does.
And just to throw mud in the game, aren’t these blue waves supposed to be up closer to 1. Not 0.1? Whatever that work is, it’s about 10% of where it should be. That can cause Effect > Compressor not to work right. It’s assuming a normal volume show.
And, as always, my first recording is always quieter than it ought to be. I don’t know why and haven’t yet managed to fix that but I’m not worrying about it for this book. I’m going to try a few things (most noticeably putting the microphone closer) for the next one but already making a couple of moves (the music stand and microphone) lead to the elusive noise returning.
I still manage to get the end result within ACX parameters but it likely takes longer than it ought to and certainly, without knowing how to properly use the Envelope tool will take me even longer from here on in as I’m now at the point of all the zombies nattering and yelling and I find it difficult to do their voices without being a little louder (apparently).
Yes the envelope tool can be a bit awkward to use. What you are doing is placing points (and the points show up as little dots on the graph). For what you are trying to do the first thing is to click at the beginning and end of the selection you which to modify, but just click such that a point shows up, don’t actually change the gain. Then you probably have to click at least twice within the section to get an area of reduced gain.
Suggestion for the developers: When the envelope tool is invoked and only a section of the track is currently selected, automatically add points to the beginning and end of the selection to pin the gain so that the tool only effects the selected section.
The worry about the sound level has to do directly with making ACX conformance. Noise is killer difficult to meet and still make peak (overload) and RMS (loudness). This is the process where people start digging themselves a hole with more and more aggressive patches, filters and tools to force conformance.
Aggressive patches, filters and tools are not welcome at ACX.
Interesting idea, but I’m not sure that would suit everyone. It would perhaps be worth raising that idea on the Adding Features board http://forum.audacityteam.org/viewforum.php?f=20 for further discussion. Developers rarely visit this forum, but features that are discussed on that section of the forum are eventually transferred to the Audacity wiki where they are more likely to be seen by developers.
As previously mentioned you will want to place points at the beginning and the end of the section you wish to modify that are at unity (so that the rest of the recording is not effected. Then typically I would place two more points one a fraction of a second after the beginning of the section you wish to modify and one a fraction of a second before the end. Those two points would be adjusted for the necessary level shift. (ie if the section is too loud they would be set at a lower volume.) In the attach example I’ve adjusted the volume of one bit to match with the next bit (it started out quite a bit louder).
You might want to add a lot more points, and the level might not be uniform through the section, you just need to listen to the clip and adjust as you want. But you should try to avoid abrupt changes in the middle of a word as that will almost certainly sound odd.