So I decided to rework a demo -- actually a job that I did in 2020 that has always haunted me, trying to get it to sound right, or at least better. It was a job that was given to me to do the voiceover and composite on top of a video with a sound effects bed. Easy-peasy. I have software that can do that (Reaper) . However, since I did the piece, I never liked the balance of VO to the bed. I mean, it was ok and the customer liked it (I suppose), but something had always stuck in my crawl. The VO, no matter what I did, never popped and the bed never had any depth. By that I mean it sounded flat and without any...I dunno, any life. Just plain boring.
I had reworked this piece a number of times and still could not get the sound I wanted. I tried every trick I knew at that point. EQ. Compression. Gateing. Limiting. I mean, I did more EQ that a man is allowed to do. And for what? Some thin sounding production? No. I was not going to quit. Winners never quit and quitters never win (but those who never win and never quit are...nevermind). So I wasn't going to be that guy who never wins and never quits.
To give you an idea to start, visit this set of clips off my SoundCloud page. You'll hear a short section of the original piece, a short of the newest rework, then an A/B of them both alternating so you can really get a great idea of what the differences are.
I pulled the files back up after wanting to showcase this video as part of my portfolio and just simply shook my head. I reviewed the tracks and the FX I added back then and since and noted that there are a lot of things I hadn't considered. Especially some of the techniques I've honed over the past year or so. Specifically some of the techniques I've blogged about! Mainly my blog posts about Gain Staging, Compression, how devices portray sound, and most importantly LUFS and how to use/understand it.
I also reviewed my FX chain. If you're new to sound editing (or if you have a clue), an FX chain is a post production set of effects such as EQ, compression, limiting, expansion, you name it. What a mess I had on my hands, but at that time, that's what I knew. I didn't really have an expanded grasp of what it took to make audio pop the way I wanted it to.
The first thing I did when I looked at my latest rework was to review what I didn't do in terms of gain staging. Now, I've talked about gain staging before, so go take a look at that. I go into VU meters and how that assists in getting the levels right before the FX chain.
Once I got the gain staging right, I immediately noticed a difference. The VO and the bed were much more in balance. Amazing. What a difference gain staging make.
Now go back and take a look at the blog post about LUFS. There's a link to a Youtube video from Lenny B. This video was a revolution in how I master my audio. Bear in mind, my techniques work for me and if you want to use them, do so but they will need tweaking for your voice and setup.
What did we learn from Lenny's video? Well, the most important thing is getting the sound to a listenable level. By that I mean, the quietest parts that need enhancement, need to be enhanced. A lot of folks think compression is the key, and for the most part that's true. Compression mainly works from the top down and yes, a good compressor can add back in some gain. This is the important part!
Now, lets look at compression for a second. A compressor that compresses at a ratio of 'Infinite', basically creates a wall. Past a certain point, the levels are clipped off like a lawn mower and grass.
Now, we call this a Limiter. A Limiter has the inherent ability to mow off the tops of any tracks, leaving them at a set upper limit based on a threshold (more on that in a moment). Limiting has the unique and quite useful attribute of preventing (yes, preventing) clipping in audio which is bad. As you can see below, the wave form above has been mowed off at a limit, hence being a limiter. You'll also note that the overall loudness has increased -- more so than when the gain stage was applied.
So in essence, the Limiter has done the job of a compressor (to the extreme) that has gain applied, or as we call it Makeup Gain. A lot of compressors not only compress above a threshold, but I'm seeing them compress the whole audio waveform. I don't want that to be honest.
That all said about compression and limiting, I've opted to go with the Limiter to do a few jobs: It increases the overall loudness of a track or take or waveform (or whatever you wanna call it) and prevents clipping. Along with gain staging, any audio should sound much better and can cut through and even become clearer. I think you would have heard that in my example.
Now, In addition to all this, I've learned a few things about plugins. A lot of you out there have a set FX chain as do I, and most of you rely on that FX chain. Don't change what you're doing (unless like me, it just isn't satisfying you). What I've come to discover is that when I had a long or complex FX chain, the more plugins I put to a problem, the more the problem became. Nobody likes problems and especially when its a vexing problem like I had.
My solution is to simplify (so far, and that's likely to change). Look, all you really need is to listen to your raw audio. Listen to what the mic produces. Listen to your environment and how your mic pics up what (echo, room/environmental noise, the tone of your voice, etc). Then start from scratch.
First step, and I cannot stress this enough: Gain Stage. Depending on your DAW (Digital Audio Workstation), make sure you :
1. Leave at least 12 dB of headspace when recording. Somewhere between -18 and -6 dB and median around -12.
2. Bring the levels up in your DAW to where the average tops of the audio are basically at 0.
3. In your FX chain, add in a VU meter and I usually set mine to a Nominal Level of -12dB. I got this tip from Reaper master Kenny Gioia.
4. Add in some sort of Limiter. I set mine to a threshold of -3dB and a limit to -1.5dB for the final output.
5. Add in an EQ of your choice. For me, I have a low and high cut applied and depending on what I see in some sort of frequency graph, I will enhance the mid tones mainly -- but only a little. Mainly a cut instead of a boost. Better to remove than add IMHO.
6. I add Youlean Loudness Meter 2 to review the final levels before I 'print' the final audio.
7. I will also add a limiter to the Master Track in my DAW. The Master Track is where all the other tracks are mixed and finally sent to output, whether that be a file such as MP3, MP4 or WAV, or hardware. I set that limiter to -6dB threshold and -1.5 limit. I've seen Lenny B use -8 but that is too much gain for me and my voice dynamics.
8. On the Master Track I also put the Youlean Loudness Meter and check the LUFS and True Peak dB (dBTP)
9. Then with the use the virtual mixer board to adjust the levels of each track or the master track so that the output is where I want it. The LUFS should be somewhere at -14 LUFS.
On my DAW, I have the ability to see what the 'Render' will look like and what the LUFS reading is. This example is a little hot, but you get the idea.
Now, what you add in between in the FX chain is your business. Waves DeBreath is good, but I find it unnecessary as too much breath removal sounds sterile. Some, yes. All, no.
So go out now and if any of this makes sense and is useful, make something wonderful and feel free to ask questions. email me at woodyg@woodygvoiceover.net
Comments