with Tim Tippets
Do you need help fine-tuning your audiobook production? Anne and VO Tech Guru Tim Tippets, discuss the tools you need to create amazing audio for audiobooks that stand out from the crowd. This episode will help you understand how to correctly normalize your audio rather than using a plug-in and getting your audio to the required specs. Listen in to learn why being a better editor is a valuable skill for every #VOBOSS.
Quick Concepts from Today’s Episode:
RMS (root mean square) normalization averages the overall sound of the audio, making the peaks quieter, and the lower parts louder.
The standard RMS levels for ACX (and most audiobook production companies) requirements are between -18 and -23
You don’t want the listener to constantly reach for their volume button, so you want to average out the overall level of the audio
There are plug-ins available to adjust your RMS, but using some of these can make your audio sound unnatural and overly pushed
After normalizing to -3db, you can use responsible compression (with a threshold set to just below your peak levels and attack set to zero) to achieve your target RMS
If you have the better sound coming out of your studio, the more chance that you’re going to get the gig
You can’t just record an audiobook from your living room. You need a pro studio to compete
Share ideas with your own network ++
Referenced in this Episode
Direct links to things we brought up ++
Learn about audio submission requirements for ACX
Learn more about audio by taking Tim’s Courses!
Hear more about tech with Anne and Tim here
Recorded on ipDTL
Full Episode Transcript
>> It’s time to take your business to the next level, the BOSS level! These are the premiere Business Owner Strategies and Successes being utilized by the industry’s top talent today. Rock your business like a boss, a VO BOSS! Now let’s welcome your host, Anne Ganguzza.
Anne: Hey everyone! Welcome to the VO BOSS podcast. I’m your host, Anne Ganguzza, along with my very favorite audio engineer, Mr. Tim Tippets! Hey Tim, how are ya?
Tim: Hey Anne. I’m good, how are you doing?
Anne: I’m doin’ good. You know, Tim, last, last podcast, we talked about normalization, and there were two different normalizations we talked about, peak and RMS normalization. And I remember we were talking about RMS was used in audiobooks. So what do you think about talking about audiobook production and how that, how people can fine-tune their audio to make their audiobooks sound amazing?
Tim: Well, I’ll try. I mean, I don’t know if people out there know, but when Anne and I meet, it’s completely unscripted.
Tim: So we just kind of wing it. So RMS normalization sounds like a good place to start. So I think I made it pretty clear when we were talking about normalization that peak normalization takes the loudest part of your signal, amps it up or brings it down, and either amps up the audio or brings down the audio by that same amount, right?
Tim: So it uses the largest or the loudest peak as the baseline. Now in RMS normalization, which again stands for “root mean squared” — which is a fancy way of saying averaging, right? So when you’re averaging the loudness of the audio, that’s what RMS normalization does, is it takes the quieter parts and makes them louder, and it makes the louder parts and makes them quieter, but then you normalize to -3 after that. After that you do actually do peak normalization, right, to make sure it’s brought up to -3 dB.
Anne: I’m glad you brought that up, I’m glad you brought that up. Now why again do we do RMS for audiobooks?
Tim: Well, the reason we do RMS for audiobooks — and by the way, we don’t always need to do RMS for audiobooks. There is a way to go about getting the levels that you need to get by using a light compression and doing it a few times until you get where you need to be. I just want to make that clear. Other people are using limiting in order to get there. I don’t like limiting, because without getting too scientific about it, it sucks out the dynamic range and it cuts off the top of the waveforms. Anyone out there who has recorded a waveform and then limited, you’ll know exactly —
Anne: Oh yeah!
Tim: — what I’m talking about, ok?
Anne: Yup! [laughs] I do.
Tim: And that is fine for even some commercial or promo stuff. Like if you’re doing longer narration, and after you’re done normalizing, you have a few, you know, peaks that are just kind of sitting out there on their own.
Tim: But they’re just a few, right?
Tim: So when you nip off the tops —
Tim: — when they do come along when you’re saying one word, no one’s gonna notice. But they certainly will notice if you limit too much, and the entire thing is just [in loud monotone] one continuous volume across the entire file, and it just sounds like this. [normal] Nobody wants that because the dynamics are gone, right? But, but here’s the thing about getting an RMS number between -18 and -23, which is ACX requirement, plus then normalized to -3 dB, is because a lot of people are listening to podcasts in their car, on their phone, and so on. And no one wants someone constantly reaching for their volume knob because they can’t hear something.
Tim: Right? So averaging out the audio is actually a smart thing to do. It’s the right thing to do when it comes to something like an audiobook.
Anne: Well, I’m gonna say, in a contrast, in terms of when I’m recording narration, if I have a high peak, I tend to want to like just take that peak and de-amplify it a little bit so that when I do my normalization, it’s not completely, you know, whacked out for that one high peak that I have. But for audiobooks, I hear what you’re saying, to have more of it normalized, or excuse me, more of it normalized so that you can hear in case there’s other extraneous noises. I think that’s great. But there’s so many different ways to do this. What do you suggest? I mean, is it just an RMS? I’ve known people that have used different tools that have helped to kind of normalize your audio that way. And I don’t know like what the effectiveness of that is. I mean, sometimes I listen. I mean, I’ve used the tools myself back in the day, and I’d listen to it afterwards, and it just didn’t sound right.
Tim: Yeah, so without calling out any names, there is a tool out there that kind of gets your loudness where it needs to be automatically. But you don’t see it happen. Right? You just throw it through this plugin —
Anne: Right, right, exactly.
Tim: Yeah, and it spits it out the other side. And we —
Anne: Right. And there’s no control, there’s no control.
Tim: Yeah, and there’s no control, exactly. When you’re taking care of your RMS levels inside of whatever you’re using, whether it’s Audacity or you know, whatever DAW you might be using — the simple fact is that when you’re using responsible compression in order to take care of those peaks that you talked about, right, and you’re bringing this down — and this has to do with some terminology that I realize is gonna confuse some people, but what we’re really looking for here is to normalize the audio to -3 dB first, and then use this compression with a threshold that’s set a little bit below that peak —
Anne: Ah, right.
Tim: — so that when it compresses, right, it’s only hitting those peaks.
Tim: And we set, and we set what’s called the attack time, which is how quickly that compressor begins to act, we set it to zero so it sort of acts like a limiter but at the same time not, because it doesn’t just cut off the top of the dynamic range. So when you do it correctly, and you’re still meeting those -18 to -23 numbers, and then you use this plugin that we were talking about earlier — I don’t know if you’d call it a plugin much less a tool — but regardless, when you run it through this leveling plugin, and then you listen to doing it the way that we talked about earlier, which is just a series of compression until you reach your numbers versus listening to it thrown through that tool, it’s not even close.
Anne: Mm yeah.
Tim: After having thrown it through the tool, it does not sound natural. It sounds as if it’s being over-pushed, the signal is being over-pushed, for lack of a better way to put it, and it’s just not preferred. So even if it’s not audiobooks that we’re having a discussion about, and let’s say it’s longer narration, which you and I have both done —
Anne: Right, right.
Tim: — elearning, etc., and you see those few straight peaks, then yeah, you know, you could limit that. But still, Anne, I’ll get with you, and we’ll put in like a little compression plugin to get you in a better spot so that you can just hit that one button —
Anne: Oh yes, nice.
Tim: — and then run with it.