Here’s a step by step guide on how I edit Caustic Soda. The segment I’m editing is the “intro” portion of our show, so it’s short.
1. Import the files into Audacity. I’m using the old version because my current computer can’t handle the most recent version.
You can see already that track #2 (Joe) is about twice as loud as everyone else. That will almost definitely require some fixing.
2. Look for trouble spots. Joe’s peaking – or coming very close to it – in a couple of spots I’ve circled below. This is usually some kind of clicking (this is often done right before you speak a sentence as you open your mouth and the tongue comes off the roof of your mouth) or from the hard “k” sound in “Caustic.”
3. But I’ll come back to that in a sec. Next I mute the tracks of anyone who isn’t speaking via the judicious use of the “silence selection” button. We usually record with everyone in one room, so there’s always some bleed-over (that’s a word now) onto other tracks. If you want your podcast to sound it’s very best, you manually silence that bleed-over. TIP: If you have someone who is dominating the talking (usually me) in the podcast, put that person on the top or the bottom so that it’s easy to select everyone else with a click and drag (as seen in the big grey block below).
4. Back to Joe’s peaking (I’ve zoomed in here). I select the bit that’s louder than the rest of the track. I select Effect/Amplify.
-4 db almost always does it perfectly.
5. Next I’m going to make sure everyone’s volume (level) is the same. There are 2 ways to do this, AMPLIFY or COMPRESSOR (both under Effect). Either way, select the entire track or the chunk of track you want to adjust. I’m going to start with Joe’s track since I’m there anyway.
With compressor, I usually choose somewhere between -10 and -8 db in the Threshold field. I don’t mess with the other fields unless the track really needs some serious work.
This will make the loudest parts of the selection hit the top or bottom of the track, as shown below. It’s okay that the peaks of the waveform touch the borders, this is the maximum loudness you’re going to get with this function. If I hadn’t adjusted the peaks in step 4, ALMOST NOTHING WOULD CHANGE on the track, so that’s why it’s so important to manually soften those peaks. We’ll see more of that in a second.
Compressor works well for speakers who have poor microphone technique, which is to say speak loudly and then mumble (I’m guilty of it!), as it decreases the highs and increases the lows of the selection you modify. If you use the AMPLIFY function to increase your track, this will not happen.
Looking at the waveforms above, you can see that track 3 (Jordan) and track 4 (Kevin) are pretty uniform, so I won’t have to do any manual tinkering with them. I’ll repeat the compressor on their waveforms too (and I can do this with the ‘repeat last command’ shortcut: CTRL-R).
6. Below are all tracks compressed. I used the compressor on track 1 (my channel) too, but you can see the problem by looking at the waveform. I too had too many peaks, and so the average volume (noted in the dotted red line) doesn’t go all the way to the top and bottom border of the track. I’ll have to do some manual amplifying again. This is time consuming but it’s the method I’ve found to be the best quality. All the other shortcuts I’ve tried come out sounding crappy.
Here’s the adjusted track. While listening to it, a lot of the stuff on the track was to be cut out anyway (again, there were a couple of tongue clicks), so cutting that stuff out saved me from having to fiddle with the amplify. Here’s the waveforms after they’ve all been properly compressed:
7. Listen to the tracks. Don’t trust the waveform! Do they all sound roughly the same level? If so, you’re done. If not, you can adjust the track volume in a different way. Let’s say that Joe, just by the nature of the recording, is still louder than everyone else after all the fiddling and compression. I can use the volume slider on the left of his track to reduce his volume (below in red).
I should point out that I never use this slider to increase someone’s volume, because you risk creating “hot spots” in the peaks of the track (which you carefully crafted earlier). Really I rarely use the slider at all, preferring to use the Amplify function to reduce the entire track, usually -2 db, though this can take some processing time depending on your computer.
