Thirith on 4/9/2017 at 10:39
I'm doing a podcast with a friend, and while I've found a lot of helpful information online, i have a couple of concrete questions: how do I best make the recording sound as good as possible? I know that you can't turn a recording on a cheapish mic into something that sounds like James Earl Jones is booming sweet nothings in your ear, but I'm curious what sort of tips people here have that can be done by someone with Audacity and lots of enthusiasm but little experience.
What I've already gathered is that the following sequence should work reasonably well:
1) Noise removal
Possibly 1.5) Limiter
2) Normalize
3) Compression
4) Equalization
However, I find that my waveforms are all nice, compact and relatively well balanced after steps 1-3, but when I add a minor boost to bass in step 4 (which many webpages recommend to make voices sound warmer) they're all over the place again, with 'p's and the like sounding like a minor bass drum kick and clipping happening everywhere.
So, in general, any tips for getting voice recordings to sound better? And any specifics on how to do the above (and any other recommended edits) and tweak the settings?
Aja on 4/9/2017 at 17:58
Here's what you should do:
Get a quiet, dead space to record in. Hang up some clothes or a sleeping bag in a closet or other small space to dampen the reflections off the walls. If you can hear the sound of the room, it won't sound professional no matter what.
Record with a condenser mic (doesn't have to be expensive) and use a pop filter, which should help with the P's. Try to make your average recording level around -12db; lower is fine, but too high will cause clipping, obviously.
There should be no need for noise removal. Figure out how to get it quiet at the source by having a quiet recording environment and making sure your gain is set correctly. I assume you have an audio interface, so correct gain staging means making sure you're not clipping the interfaces preamps as well as clipping the digital meters in Audacity. Normalizing is almost never necessary since you should be getting decent levels off the bat. I suspect some of your clipping is coming from that.
Limiting doesn't really make sense to me either. But compression is really important if you want a professional sounding voice. If Audacity can use VST plugins, I recommend trying the Klanghelm MJUC Jr. It's free and should work great on vocals. If Audacity can't do plugins, switch to Reaper. Learning to use compression properly is a whole other thing, but with the MJUC Jr you basically turn the left knob until you get the sound you want, and then turn the right one to bring the gain back up.
Equalization is important too. As a starting point I'd recommend putting a 3 or 4 dB dip in your EQ at around 1k or so, and pan it around the frequency spectrum until you start to hear less boxiness, nasalness, or whatever your problem is. Sometimes dipping the low mids can give the impression of boosting the bass without actually having to. Avoid boosting frequencies if possible although slight boosts can work. You'll get more of a thick sound from compression than bass boosting. And don't be afraid to compress more than once if it sounds good.
Thirith on 5/9/2017 at 07:21
Thanks, that's a lot of information to go on. I won't be able to put all of it into practice, but it's still useful.
I'm a bit surprised you say there's no need for noise removal, because pretty much everything I've read says you should do it. Perhaps you're too optimistic with respect to what sort of recording environment me and my co-podcaster can manage; for one thing, we're not physically in the same room, recording ourselves separately in different rooms and on different microphones. If the recording environment and equipment aren't perfect and most likely are only okay, would noise removal make sense? Since it's not necessarily realistic to expect professional quality, it's perhaps more of a question of how to compromise smartly.
Aja on 5/9/2017 at 14:23
Well, I'm no expert either, but I have spent a lot of time recording music in the last few years, and in my experience noise reduction usually impacts the sound quality unless used very sparingly. And I guess it depends on what kind of noise you're trying to reduce. If there's a slight hissing on the track, you might get away with it, but it'll still probably impact esses and consonant sounds. If it's traffic noise, it won't work, and if it's something like a computer or a refrigerator you're trying to cut out, you'd be better off moving the mic away from those things.
Aja on 5/9/2017 at 16:28
I think I should clarify some things I said earlier. About normalization, it shouldn't be necessary if you're recording at a good level and are using a compressor to control the dynamic range of your recorded material. The basic job of a compressor is to dynamically turn the volume down when the source material gets loud, so it makes the loud parts closer in volume to the quiet parts, reducing the peaks and allowing you to turn the overall volume up to the desired level without clipping. This consistency in volume is a big part of what makes a voice sound "broadcasty."
I don't have a lot of experience using limiters as they're more of a mastering tool than a mixing tool, and my mixing experience is, well, limited. For more info on using them correctly in a broadcasting context, you'd have to talk to someone with more experience than me. That said, I've been able to record professional-sounding voices using only a compressor, so it's possible. I imagine the limiter would be more for safety, just to ensure that nothing goes above the desired level and that any peaks that the compressor doesn't catch are controlled.
As you can probably tell, I love talking about this stuff but am still learning my way through it, too. If you have any samples of your recorded material, I'd like to hear them.
Thirith on 5/9/2017 at 16:49
The first episode should be up in a week or two; I'd definitely be interested in your feedback then, especially in view of ep 2. I'll let you know once it's up. For ep. 2 there'll be some changes; I got myself a different mic which should at least help with my side of things.
Judith on 7/9/2017 at 10:51
In terms of preparing your studio, I prefer directional microphones, like shotguns or Rode VideoMic Pro put directly over your head and slightly in the front, so they're aimed approx 45° downward, at your head. This way you have to worry mostly about two things in terms of reverb: a) the floor, so you can use carpets and blankets to dampen the sound, and b) the space directly behind you. If you have stuff like tripod or microphone stand, you can use them to hang thick towels or blankets.
As for the volume levels, you should not just use simple normalization, you need to adapt one of the Loudness Normalization standards and stick to it. Most audio tools have the Loudness Normalize/Match Loudness option these days. That will let you analyze the audio file and normalize perceived loudness (vs the mathematical average). Typically it's -23 or -24 LUFS, although for mobile phones and internet broadcasting most people use -16 LUFS. I agree that you don't need noise reduction that much, I wouldn't use it untill there are some serious problems after I compress and normalize the audio.
There isn't any universal approach to EQ though. It all comes down to voice characteristics. Typically higher mids are most important, but some people have lower or higher registers, and that may sound disctracting which will require some fine tuning.
Thirith on 8/9/2017 at 07:34
I don't think Audacity works with LUFS, which is a shame. In any case, thanks for the additional tips. My main problem when it comes to recording is that my computer is pretty loud, but I'm hoping the new mic will make that less of an issue. We'll see next week, when we'll record the next episode.
Aja on 8/9/2017 at 21:20
Having a directional mic should help, but if there's fan noise from the computer in the track, I doubt you'll be able remove it with noise reduction without noticeably impacting the sound of your voice. Might be time to get a long cable and record in a different room, or get some kind of sound barrier for between you and the computer.
Thirith on 9/9/2017 at 10:35
Yeah, I'm thinking I'll see how it goes if I block the fan sound as much as possible, but if it's bad enough to annoy me constantly, I'll see that I can put a quiet cooling solution in the PC.