~ Or: This is the part where I talk nerdy to you
Whence we left off in the previous installment, we were discussing the ins and outs and odds and ends and bits and bobs of stuff that go into the creation of an eLearning voiceover demo. From my perspective. In a behind-the-scenes look at the creation of a voiceover demo for Roshelle Simpson. Quite a bit of it related to the creative side of things: getting to know the person in question, ironing out the details of the script topics, and the actual recording session. So much of the creative side of things! Also slightly technical but mostly creative.
Once the recording session is complete, it’s onto the more technical side of diving into manipulating audio in a way that is appropriate to best represent a talent in question.
Side note: the one section on the music writing that was specifically done for this voiceover demo will lean more creative (obvi) and I’ll include it for the purposes of proper selection of music for these productions. Under normal circumstances, existing library music would be selected. I needed a new music reel so this was perfect timing. More details on that ahead.
Anyway! Kick back, relax, grab a plate of wæffles and dive into the more technical side of things with me!
Which is still artistic in a way. And also technical. But also artistic. But al–
[divide by zero. goto loop detected. brain œffline. please reboot]
Audio engineering voodoo witchcraftery in a voiceover demo
I’m dividing the following process of processing the processable voiceover processional into processed audio into three steps. First is an evaluation of the recorded voice. Next is some degree of sleecey slicey deecey dicey. And then finally the final finalization in finality.
(words are fun. welcome to my insanity.)
More details!
Step 1 – Evaluate and “fix” audio
One of the first things that I will do when receiving the audio from an actor – if I didn’t record it directly to my machine via Source Connect – is give it a good listen for anything that stands out as being out of order. Subtle room nodes, harmonic resonances, subfrequency noise, the fridge down the hall, comb filters, brush filters, hair clip filters, and so forth.
If there are obvious quality concerns during the recording session, it’s possible that those might need to be addressed in advance. Things like that would be inordinate amounts of untreated room echo. Electrical noise in the signal. Microphone noise due to moisture or debris on the capsule. A chopper right over the house with a banner announcing to the county that there’s a voice actor below attempting to record things and to make certain that this evilment cannot take place. You know, that kind of thing.
One of the steps that I take prior to the recording session is to request a raw audio recording of the room. This is the individual speaking in a normal voice, a slightly elevated volume, and then 20 seconds of silence in the room. Those major concerns mentioned above would be found during that evaluation and could then be addressed at that point.
More subtle things? I can fix those after the fact and balance it out. No big deal. That’s step one of the voodoo magic.
Step 2 – Mix ‘n match all the things
Step two is alluded to above: where every take of every line of every script is evaluated, played side to side and back to back, and decisions made about which takes are the ones to hang onto. In the case of Roshelle, we had bits of both. A couple of the reads were absolutely perfect right out of the gate. A couple of them were “cut-and-paste stitchery” because of moments of brilliance in each. And sometimes? The takes are fine on their own but I hear something that I didn’t notice in the recording session that fits better when mixed and matched so I’ll do that instead.
Step 3 – Final processing and tweaking
The final step of this is taking those segments, putting them all in order, and then doing a little bit of extra processing to even things out for consistency in audio signal. Sometimes that can take quite a bit of nerding around with equalizers and compressors and sometimes it’s super quick and we continue on to the next step.
That’s my three-step program to putting together all of the recorded voiceover stuff and things. It probably took longer to write out the paragraphs above than it would take to go through that process, but documenting is fun and here we are!
Tunesmithing
Apparently I’m all about the smithingness of things in this post today…
So! This is where there’s going to be a bit of a deviation from the norm. Under normal circumstances, I would select music that matches the mood of a segment from a library of music that I have access to that is specifically licensed for the purpose of using in productions like this. Usually searchable by mood or instruments or tempo, it’s decently easy to find a matching track for any segment.
This time was a little bit different.
I have been needing to update my library of music beds for quite a while and it seemed like as good a time as any to do so. So I asked Roshelle if she was interested in having music written for her segments rather than just having it selected from an existing library and she agreed.
The jury is still out on whether or not that was a good idea.
On to the merging of art and science!
Each of the songs is uploaded to SoundCloud presently and I will include links to them in each title, along with the titles that randomly popped into my head when it came time to name them.
In n Out – “Bob, the Ironic Dancing Moose”
So in this case, I wrote music that I felt fit the mood of each of the segments. Sat down and asked myself “What does an ironic employee training video music sound like?” At the time, the answer was with pizzicato strings, a smattering of orchestra percussion, and a clarinet. Some mallets. A nice little dance between minor and major keys. The rest of it kinda wrote itself.
Before you can ask “But George, why do pizzicato strings, a smattering of orchestra percussion, and a clarinet with some mallets sound ironic?” I’ll say that I am not well studied enough to be able to qualitatively answer that question. Music theory is not something that I have any major strength in; I just hear things in my head and then try to recreate them. In the case of the song in question, this is what came out of the keto-compatible fatty computer made of meat in my brain pan.
If there was an ironic moose with the capacity to actually get his dance on, his name would be Bob. Thus, the title.
DEI – “PianoCello with Pings and Things”
The moment I started to listen to Roshelle read the DEI script again, the entire mood was piano and cello. And a bit of ambient pingy synthesizer. It was interesting working with the Cello samples that I have because the higher the notes got, the more they wanted to hog the spotlight. Which is contrary to most cello players that I know who are content without the need for the limelight… Since that was not the hill I wanted to die on, I let the cellos have their way.
Cuisinart – “Why Argentinians Weep for their Art”
When it came to the ironic equipment training for Cuisinart, the ONLY thing I had in my mind was Tango. My original thought was to write Tango music that was more authentic street-music from Argentina, except that as that was coming along I realized that it was more of a distraction than an enhancement. So I went back to a more straight tempo and less excitement in the melody and chords. Why Argentinians weeping for their art? Have you ever seen Argentine tango done authentically? And then have you seen American or International ballroom tango?
I rest my case.
Funny side story: it’s the first time I’ve ever recorded the accordion that I received from my Great Aunt a good 20something years ago. And it was already old by then. The leather straps actually snapped while I was recording and I had to figure out how to solve that to keep the music flowing!
The solution ended up being attempting to sew up the broken parts of the straps and they held just long enough to get through the song. I’ve since replaced those poor straps. They were ancient but they served us well.
Ovarian Cancer – “Guitars of All Shapes and Sizes”
The ovarian cancer research background absolutely screamed acoustic guitar and guitar harmonics. Except not actually screamed because it was nice and quiet. It was a quiet scream. Go with me on this.
Calm and soothing were the criteria in my head for a script like this one. Front porch rock. –ing chair background music. Except to take that to the next level by adding electric guitar harmonics on top of the acoustic guitar accompaniment. A nice merging of the old and new world into a very soothing piece of music with hints of optimism based on the research in question.
ADHD – “Electronica Hydroponica with Springy Bassica”
ADHD had to be a light electronica track. Why? I’m not sure. Full disclosure and I think I’ve touched on this before: I’m diagnosed with ADHD myself. It’s possible that I was thinking back to all of the years of music written and the early days in high school. Back when I could hyperfocus on the writing process. Most of the music then was electronica in some nature. We didn’t have realistic sample libraries then outside of tracker software that used samples. But that’s getting a bit too nerdy for this post.
In any event, the baseline track came out pretty fast – like 20 minutes – but then I A G O N I Z E D over the inclusion of a background beat. For at least an hour and a half. A war was waged inside my head. There was weeping and gnashing of teeth.
Ultimately the background beat won the battle and there we have it.
Link to the music beds for this specific voiceover demo
If you’d like to give those pieces a listen as a playlist instead of individually and without the accompanying voice (I don’t, but that’s me), you can find them at a SoundCloud Link here.
Final thought on that: big thanks to Roshelle for taking a gamble on that. The backup plan was in the event that the music didn’t turn out good we would go with stock library stuff. Again, this was outside the normal scope of what I would do in the creation of a voiceover demo but I had a need and it seemed like a good opportunity to get that taken care of. Ultimately I think she liked it and we ran with it!
Cold reailty: the music doesn’t matter that much. This is a voiceover demo, not a multimedia project. Music needs to be mood/style appropriate enhancement but not be even remotely the star of the show. This was just a teensy bit more sugar in the icing but nothing more than that.
Balancing the voiceover demo sonic scales
This is where some of the art starts to get merged in with some of the science.
Once upon a time, did you ever encounter that thing where you would listen to one song on a playlist and then the song after it was SO LOUD THAT YOU DIDN’T THINK YOU WOULD EVER HEAR AGAIN?
And then the track after that is super quiet.
Different songs from different albums are mixed differently. In the context of an album, it sounds fine because all the songs in the album are going to be uniform in volume and dynamic. From album to album, artist to artist, this is not always the case. If track one is a Queen song from the 1970s and then it switches to a track from Metallica’s “Death Magnetic,” you are going to need new ear drums.
In the case of a voiceover demo, the genre is going to determine some aspect of the level of loudness and dynamics involved. eLearning demos are, thankfully, generally not in the category of “VOLUME WHAT COULD PEEL THE PAINT OFF YOUR WALLS” and you have a moderately pleasant listening experience and a good method of evaluating the talent in question.
Where does balance come from?
No, not the stork. Or under cabbage leaves. Definitely not Tai Chi in this case.
The short answer is that once the voiceover has been appropriately processed and music added in another track, it’s time to balance them out. The featured voice needs to be a percentage louder than the music backing it up, which makes sense. The voice actor is the one who is looking to demonstrate their capabilities to a prospective business partner. That prospective business partner is interested in what the voice in question might sound like out in the field amidst an œfficial production and the voice must be the dominant entity.
Sometimes balancing those things out involves a small cadre of tools and measuring devices. When mixing for long periods of time the ears will lie to you and having reference tracks and taking breaks helps quite a bit. For those times when breaks are not available, visual guides are helpful to figure out where you’re at. I’ll spare you some of the more mathematical terminology because m*th is a four letter word in every sense. Some of those terms are measurements of sound in decibels and include Average and peak RMS and Loudness Units Full Scale (LUFS).
(gesundheit)
Thank you.
(Будь здоров)
Uhh, Спасибо?
(Alhamdulillah)
Ok this is getting ridiculous. Moving on!
Sorry, the short answer got really not-short. That happens to me sometimes.
Anyway. Unnecessarily long story short: the voice must be prominent in the mix. The end.
Smashing it all together into one big happy track of tracks
There are two different directions to go with this one. Many times it will be one long mp3/wav file with all of the demo segments blended together. Creative use of crossfade, and so forth. Many will then take this single file and make it available for listening and downloading on their websites and in social media profiles.
In the case of folks who use players like VoiceZam that take the individual clips and merge them together into a playlist, there is a different solution. The individual clips are provided for the actor to upload into the player. The player will then make a full version plus the individual segments available for listen and download. Kinda catchy, eh?
However it is done in that sense, the order has already been determined and the file provided. Up next is presentation to the client to find out if they approve of the mess we’ve made or would like modifications.
The signoff to end all signoffs
That NERVE-WRACKING MOMENT when the first draft of the full voiceover demo is presented to the client. Is it going to be everything they hoped it would be? Is there something about the order they’re not digging at the moment? Are they not fans of that poor moose Bob? Are they pleased as punch and can’t wait to put it on their web site??
One of a couple things takes place here. Either it’s exactly what they’re looking for and they are super hype, or there are tweaks they would like to have made. Perhaps the order isn’t flowing the way they would like. It’s possible that one of the selected pieces of music is too similar to a traumatizing childhood movie viewing soundtrack and it needs to be swapped out for something. They might be in the middle of eating soup and their opinion of the order will vary greatly depending on how much soup is left to eat.
From that point, it’s either back to the drawing board and making some tweaks or it’s done and signed off on and they have their audio demo in hand!
BUT WAIT! We’re not done yet!
(I was going to say “There’s more!” But that would be tacky and cheesy. So I didn’t.)
Visual representation of audio stuff and things
Depending on which marketing guru of the month you ask, video content can be considered unnecessary or the benevolent ruler of all that which is to be presented.
Better to be safe than sorry.
Once the final voiceover demo order has been signed off on, the audio then gets brought into a video editing software application and the video version of it is concocted!
There’s usually a short conversation about the type of theme/aesthetic involved. Items like logos or headshots that could be included, any other visuals requested. Once that conversation takes place, editing all the bits and bobs happens. The video is then sent to the client for, once again, final review.
Once the video is settled upon and transferred over via WeTransfer (video files are large), that’s the end of the video process!
Hypothetically.
There could be edits to the order in the future or possibly even new segments to include down the road. Since I never delete anything ever, the resource files are all available to modify as necessary.
That’s the end! Voiceover demo journey COMPLETE!
So, uhh, yeah! Here we are!
The entire process from the very beginning to the very end and every agonizing detail in between.
In the event that you didn’t know anything about how a voiceover demo creation process works – or at least mine – you are now more greatly informed!
In the event that you already knew about the entire process, this was probably really boring for you!
Wherever you land, I hope this retelling was at least somewhat entertaining. Or informative. Possibly a bit of both.
Thanks for hanging out with me for TWO PARTS of this whole thing. Let me know if you have any questions about any of the specifics!
Until next week.
-= george =-
