Normalize, Level, Noise Removal - what order?

[morvenwestfield]

I have a podcast episode where it's just me speaking. Because I'm not a professional and don't have the best equipment, I have microphone noise and differing levels among sections of my podcast. (For example, the Intro, Item 1, Item 2, and so on aren't necessarily the same level because I'm not good at reading my script and watching the levels at the same time.)

It's going to be straight mono.

Is this the proper order for effects?

1. Do noise removal on the whole track.
2. Here's where I'm confused -- Normalize or Leveller? I think that it's normalize for a single track, but I'd use Leveller if I had two tracks? Not sure.

3. After I do that (and save my project), I import my background music and use auto ducking to blend it in to the "spaces" between feature sections.

Is that correct? (Pardon my terminology; I know nothing about recording, having worked mostly with text processing programs in my career.)

And while we're at it, what's the File > Chain feature? It looks to me like it's a macro recorder along with a separate macro editor. I couldn't find any doc on it. Being a tech writer myself, I'm aware that docs often have to lag behind the engineering, but is there anything at all on it anywhere?

Thanks, and great job on this new release. I'm really liking it so far.

[morvenwestfield]

Never mind on the chain question. I think I found some info on this post:

http://forum.audacityteam.org/viewtopic.php?f=15&t=7549

[kozikowski]

<<<Normalize or Leveller? I think that it's normalize for a single track, but I'd use Leveller if I had two tracks? Not sure.>>>

Neither one. Leveler is to produce special effects like Air Traffic Controller. Normalize and Amplify bring the whole show up or down so the loudest part reaches a certain level.

For theatrical automatic volume correction, Chris' Compressor is the answer.

http://pdf23ds.net/software/dynamic-compressor/

I tend to not like Auto Ducking because it produces (annoying) pumping audio. Much better the Envelope tool (two white arrows and bent blue line) which lets you push the volume up and down in a highly controlled manner -- like someone at a mix desk is adjusting theatrical music to a script not word by word.

<<<I have microphone noise >>>

Hopefully hiss? Any other noises are probably part of your show now. We can't get rid of that MetroBus on the street or the Aer Lingus jet going over. Echos are now performers in your show. They're permanent.

Single performer podcasts can do very well by recording them on top of the feather duvet on the bed. That plus bedroom carpeting is how I recorded this audio test.

http://www.kozco.com/tech/LRMonoPhaset4.wav

Before we get into the thing on noise removal, what sort of noise is it?

Koz

[kozikowski]

Some of these tools don't work so well in mono. You might consider producing the whole show in stereo and then Tracks > Stereo Track To Mono just before you Export your final WAV. And I would Export As WAV. Don't just Save Project. You can convert a high quality WAV to anything else.

If your show master is a ratty, bubbly MP3, it will never get any better than that, and Projects are brittle, easily damaged and hard to back up and move.

Koz

[morvenwestfield]

Hi, Koz. Thanks for answering and thanks for the info on Chris' Compressor. In the past I've used The Levelator by The Conversations Network. According to their site, it "adjusts the audio levels within your podcast or other audio file for variations from one speaker to the next, for example. It's not a compressor, normalizer or limiter although it contains all three." (See http://www.conversationsnetwork.org/levelator/)

The only downside to the Levelator is that I have to export my recording as WAV, run the Levelator, and import the resulting WAV. I wanted to be able to do something entirely within Audacity. It sounds like Chris' Compressor will allow me to do that (run inside of Audacity) because it's a plug-in. I guess I'll have to run a test to see if I get similar results as I would from Levelator.

The microphone noise is a low hum. I'm using a Logitech USB headset. Because it's a headset, I don't tend to pick up other noises, even the cat meowing down the hall. No street noise or anything. I do have carpeting in the room I'm using. No duvet (grin), but I've got enough clutter in this "office" that there aren't very many hard surfaces left.

[kozikowski]

You posted in the 1.3 forum, so I hope you're actually using 1.3.

Hum we can do. I think I would clean everything up and noise reduce before you started cutting.

The noise reduction in 1.3 is the useful tool compared to 1.2. The better the "profile" you capture at the beginning of the tool, the better you can do with the reduction. There is a faction that likes to choose the profile segment (noise only -- no performance) and amplify it; then capture the profile. They claim much better reduction results. I believe them.

If Auto Ducking works for you, go with it. I think it sounds like a Special Effect and not a theatrical enhancement. There is a pop song I have here somewhere where they put that in intentionally. It makes my teeth hurt.

Chris made the first tool that sounds like the expensive rack of equipment at the FM station. The show levels out and you can't figure out what they did.

Koz

[morvenwestfield]

You posted in the 1.3 forum, so I hope you're actually using 1.3.

Yes, I'm using 1.3.9.

[kozikowski]

I whipped over Noise Reduction quickly. Do you have the 30,000 foot view of what it does? Pick a portion of the performance (profile) and the software tries to remove that segment from the rest of the show.

Everything else is steering tools, and the tools are, in my opinion, only useful in Audacity 1.3.

Koz