What’s in a Decibel?

What’s in a decibel? A lot, apparently. I’m no sound engineer, you can be assured of that. I’ve done lots of stuff about physics and sound propagation and all of that, and I can describe to you how decibels work and even how to do some calculations based upon distances and densities and things. Great. Top stuff.

Doesn’t really help when it comes to the raw practicalities of getting Voice Over work done right for audiobooks, now. Well, that’s not totally true. It helps a bit. You see, when it comes to submitting an audiobook for ACX, there are some strict guidelines issued that, if breached, will result in rejection of the audio file (not audiophile, that’s something else).

As you can imagine, there’s a desire to have quality, consistent recordings sent up for the listeners to enjoy, and one way to at least ensure you won’t bust the eardrums or have the audience screaming “What? What did he just say?”, is to keep the recording between some ‘loudness’ values.

Peak? -3dB. Floor? -60dB.

Easy peasy. Just make sure you’ve cleaned out the noise from the signal, quieten out any top-end values to be under -3dB and you’re good to go.

Wait, there’s one more thing: The RMS must be between -18 and -23 dB. RMS? RMS. Remember back to your signal processing days? Root Mean Square. Take the root of the mean of the square of the signal, and you end up with an ‘average’ signal, without the need to discard or negate the negative side of the signal (since squaring will make it positive).

OK, so what’s that all about? Yeah, good question. It means the average loudness of your voice needs to fit between these two values. Not too loud, not too soft.

Well, that’s me done because I’m a naturally soft speaker. All my files fell below the -23dB cutoff. Better give it up and go play ping-pong, yeah? Nah. Because with Audacity, I can amplify the signal and boost it up to something more in line with the requirements. Thing is, amplification of a signal also amplifies the noise, so that needs to be removed as well (otherwise the clip fails the -60dB floor test).

To help out, there’s a tool for audacity called ‘ACX Check’ that can be downloaded. Get it from : https://wiki.audacityteam.org/wiki/Nyquist_Analyze_Plug-ins#ACX_Check

That helps out a lot, letting you know if you’ve managed to get into the realm of happiness, if you’ve got some bad noise somewhere, or a pip that blasts past the -3dB threshold. The only thing I’ve found is that the RMS calculation takes into account all sound blocks, including breaks and pauses, meaning that the RMS figure is artificially low for a sentence with a pause in it.

Now, if I take the value and ramp up, then check it with the ACX Check, it may pass the RMS check, but then fail the noise check. So I then take a sample of the voice along with the silence, and the RMS check fails, but the noise floor passes. Where that leaves me, I don’t know.

I’ve resubmitted the sound files after going through them all and tweaking the amplification, so I’ll let you know how it goes.

Slow the Heck Down

I’ve been in the booth, doing my audiobook recording, and I’ve learnt so much already. It can all be summarised in two words, “Slow Down”. Just take some time and slow right down.

I was keen to get started, to just hammer this thing out and get going. Yeah, well that may work for writer’s block or when I’ve got a story in my head and I know what all the bits are. That’s typing. Typing doesn’t care about speed. If I make a mistake while typing, great, I’ll go back and correct it and move on. That’s not possible when it comes to recording audio.

In the booth, milliseconds count. I can’t stop and have a scratch midway through a sentence. Once I start speaking, I must continue until it’s time to stop. If my mouth is gummy and I choke and cough and mutter profanities, all of that is recorded. To cope, I’ve developed a mechanism and it involves, you guessed it, slowing down.

But first, a horror story: I gave the story my full attention. I put aside the day to do the recording of all the chapters (surely, it won’t take all day!) and locked myself in the booth. After chapter 3, my mouth was dry. OK, lunch, water, have a rest, get back to it. Maybe I’ll have a listen to see how it’s sounding. Hmm, no good. I’m mumbling and not speaking clearly, and talking too fast. Bother. OK, that just means I start again, and this time pay attention to how I sound as I’m talking (the headphones come in real handy at this point).

I’m guilty, you see, of mumbling, of speaking (way) too fast, of dropping consonants and flattening vowels. Thing is, I’m not conscious that I do it, not until you play a video with me talking. I’ve been told many times, by many people, in many ways, “Jez, slow down.”

Start again. No probs, got three chapters done again by Joey o’clock, went to pick him up, came home and had a listen. Still no good. I’m still talking too fast and not clearly. Fine, I get it. Slow the heck down.

I worked into the night, persisting, cracking through the first couple of chapters (again). The next morning I came at it, ready to go, ready to speak clearly and slowly. I started early and got used to listening to myself as I spoke, correcting mistakes as I went, keeping a pace. By the afternoon, I had all chapters done. Perfect. Late lunch, but who cares? I clicked on the play button and died. Ba-bow. Bad news, Jez. The little microphone drop down in Audacity hadn’t been set to the studio mic, rather the internal computer mic.

I had been recording my voice, muffled and indistinct, from within the booth while the microphone on the laptop was outside the booth. There was nothing for it.

The entire day’s recording was utterly, unspeakably, useless.

If I had slowed down, taken the time to check the settings before and replay the clip after each chapter, I would have spotted it in the first hour and only lost a portion of my day. As it was, I was in a royally gloomy funk for the rest of the evening. I had lost a day’s worth of work, had a sore voice, a massive headache and nothing to show for it.

All this means I’ve gotten into a process when it comes to recording – turn on my computer outside the booth, plug in the mic, load up Audacity, make sure it’s on the right bloody microphone, sit down, lock the booth, get thirty seconds of silence and sip some water.

Then I breathe. I use my hand as a metronome. I listen as I speak and, if something doesn’t sound right, I correct myself and try again. Then I check on the chapter as soon as I’ve finished it. It’s slower, but it’s a whole lot better.

Presenting – Grosvenor Lane Ghost

Wow! I’m bushed.

I thoroughly gave a Synfig, Audacity, Anvil Studio, Gimp, Corel and good ol’ Microsoft Movie Maker a workout.

I haven’t got a lot to say except that the promotional animation for Grosvenor Lane Ghost is now up on You Tube and Daily Motion (hehehe… Daily Motion. You know, like, one’s daily constitutional?) and any other place that I can find.

Please share, enjoy and criticise. Don’t worry, I won’t be listening, I’ll be sleeping. Right now it’s a warm Milo and off to bed.Mini Jeztyr Logo

The Front Porch at Grosvenor Lane

With the backdrops pretty much ready to go, with the exception of the fireplace, which needs some serious shadows, I’ve been looking at the animation side of things.

Nothing gets stuff done like getting stuff started, eh?

Synfig

Opening up Synfig, I set the dimensions to 1280 x 720, which gives a 16 : 9 ratio, ready to go.

Synfig-properties.png

Now, the scenes are to be slow and progressive so, unlike Adaptation which was more a collection of conceptual shots, I’ll be looking for a lengthy time span of about ten seconds. At a frame rate of 30 fps, that’s, uh, wait, let me get my calculator out… carry the 1… 300 frames.

Synfig-properties-time

Hit the OK button and get ready to rumble.

I start by adding a flat colour for the background. Black is good, considering the number of shadows. Why do I need a background if I’ve got a backdrop? If any part of the backdrop image happens to be transparent, or if I use a layer over the top which modifies the alpha of the backdrop, I don’t want to make sure it doesn’t use white or something to compensate.

Anyway, with that in place, I add the backdrop of the door frame (taken from the actual cover. Yes, it’s grainy. Yes, it’s dark. That’s ok. The scene is at night, and it’s going to be a bit further away than our Professor, anyway. The door, however, is sharp. This is where the Professor will be spending some time opening up the lock.

So I insert the door and, presto, I’ve got a front porch!

FrontPorchInitial.PNG

Props

Now I want to be able to move things about. I want the Professor to walk from the right, over to the door while chatting to the protagonist, unlock the door and open it. And for extra focus and ‘night time’ness, I want there to be some evidence of a lantern.

ProfessorBody
The Professor’s Torso

 

 

ProfessorArm
The Professor’s Arm

I’ve broken up the Professor into two main parts: his arm, which will move about to give an impression that he’s not just a cardboard cutout, and his torso. Both the arm and torso will belong to a group so that, as the Professor ‘walks’, the arms and torso bob at the same rate.

FrontPorchWithProfessor.PNG

I’ve added the ‘lantern’ to the Professor group, so that it, too, moves along with the body. It’s really just a shroud, a radial gradient of zero-alpha to full, with a heavy offset, such that everything outside of the lantern’s influence is dark.

You’ll note, on the time line there, the bunch of green dots. This is the motion of the Professor, stepping and bobbing along. The green is Synfig’s TCB waypoint inference. It gives a looser waypoint than clamped or ease. If I set them all to linear or clamped, the Professor would be marching like a soldier. As it is, his gait is more natural.

All that’s left to do now it animate the door opening, add in a warm ‘lantern’ glow to the radial gradient and shade the door more as it opens to give an impression of darkness and depth.

Sound

There are three key sounds in this scene:

  1. The Professor nattering to the Protagonist about it being dry on the porch.
  2. The key turning and the door opening.
  3. The ambient rain, a crucial element of the story.

Getting the key and lock sound was fairly simple. I went out the back to the gate and practiced with the slide bolt. A few trial runs and I recorded it on my phone, picked the best sounding one and cleaned it up in Audacity. More on the cleaning-up bit later.

The voice was more difficult. Where, oh, where does one find a Victorian Professor in the middle of outer Melbourne suburbia? I tried a few online services, but I couldn’t get the voice actor I was after. The ‘British’ was either too uppety, too young, too old or, in most cases, too damned expensive. I’m working on a shoe-string, here.

Fiverr looked promising. There are a lot of voice over artists who are willing to lend their talents. Checking through the various videos and samples, though, it seems it’s mostly geared toward reading scripts for advertisements. Not what I’m after.

In the end, I put on my best ‘old-but-not-too-old’ British accent, practiced again and again and again. And again. Then recorded myself. Yeah. That’s what I did. I hope it sounds right. You know when you hear your voice on tape and you think, “Heck, is that me?”

Lastly, the rain. I haven’t got that sound clip yet. I’m expecting it to rain here in Fawkner tomorrow, and I’ve got a nice corrugated iron cantilever out the side that should sound awesome.

Anyway, back to Synfig: I tried adding these sounds as ‘sound layers’. That is, one adds a layer of type ‘sound’, points the sound file to the .wav or .mp3 and then set the offset.

Synfig-sound-layer.PNG

This seemed the perfect way to add sound to the clip apart from two things.

Firstly, it didn’t always play. Every so often, when re-running the clip, I’d have to select the layer to give it a poke, and the sound would then play. OK, no biggy. So long as it exports…

It didn’t export. No matter what format I exported it as, the sound didn’t come through in the final file. I vaguely remember having this issue with The Bullet. My solution there was to add the sound when assembling the final video. I guess I’ll have to do the same thing here.

More on the sound and Audacity in the next post.Mini Jeztyr Logo

What’s the Score?

One needs to eat the meal before one tastes the dessert. Yeah, but it’s kind of tantalising, isn’t it? That’s kind of how I’m feeling when it comes to the music. Besides, I justify, I’m going to need to do it eventually, why not look at it now? Not to mention, I add, the feel of the music and the images presented need to match, so one will affect the other. Also… ALRIGHT!

I give in. Pencils and Gimp down, I’m going to divert to score the theme for Adaptation.

Bring on Anvil Studio

Learning from the animation for The Bullet, I’m not even going to bother trying to sing or play an instrument. Midi is the way to go, and I’m a little wiser this time around. For starters, I’m not even going to bother with mucking about with the instrumentation or beat until after I’ve got the basic melody in place.

AnvilStudio
Installed the latest Anvil

I had to reinstall Anvil Studio after the upgrade to Windows 10, no biggy, just an inconvenience. It works alright, only the Virtual Midi Synth, too, is knacked. It’s just a matter of downloading and reinstalling and… I’m done.

Now, the big question is where to start. I’m not writing a song for radio, after all, nor is it muzak to go with an abstract DIY YouTube clip, nor will it have words. It’s a theme. It’s musical. It needs a tune. Last time I had the tune stuck in my head. This time round I’m going to have to start from scratch.

John Williams I ain’t.

Humming doesn’t do it for me. Any indistinct tune tends towards an established one within a matter of bars. Da-dum-dum dum-da da-da-da bop boopeti-bop… and I’ve hit the Superman theme. A few attempts later, it’s Indiana Jones. Dammit! Lawrence of Arabia! And now Doctor Who.

You know what? I’ve got my old Yamaha keyboard in the garage that I’ve recently spruced up.

Dusting off the ol' keyboard...
Dusting off the ol’ keyboard…

Let me go out there and tap on it for a bit and I’ll get back to you…

Getting back to you

Ah, that brings back memories. I wasted a bit of time remembering how all the bits worked. You know, the white keys and the black ones, chords, scales, tone-tone-semitone, tone-tone-tone-semitone. *sigh* back to business.

What did I find out? I found out that I’m not after anything hugely complicated. A simple 4/4 time, coupled with a simplistic tune is what I’m after. To reaffirm this, I went to YouTube and had a listen to some sci-fi themes to get an idea.

Blade Runner, Alien, Terminator, Doctor Who, Terra Nova, even Flash Gordon! It seems that the space related ones, like Star Wars, Star Trek, Battlestar Galactica and Stargate, are grand, epic themes with big cymbals, orchestras and drums. Alien, Terminator, Blade Runner and Flash Gordon are less about the overall tune, and more about the beat, sound and underlying feel.

Where does Adaptation sit between these two? Probably more toward the latter side.

So I figure what I’m after is a sci-fi ditty that I can prop up with a thumpa-thumpa beat or a jika-jang of a distortion guitar to bring it back down.

Ah, phooey. I might as well stop typing and start on the melody.Mini Jeztyr Logo

Animation for Adaptation

November? Is it really November? Yes, it is. That means it’s almost Christmas, and that means that the Software Development Cycle is preparing for end of year, and THAT means a bit of a scramble to get the bleeders tied off before we hibernate for the New Year’s break.

So… what does that all mean?

What it all means

Like exercise, if you only ever train your biceps, you’ll wind up with sore biceps and flabby everything else. A change of pace is a prime opportunity to have a change of creative outlet, so I am, once again, putting the writing on hold (well, a slow down. A couple of a pages a day, max) to work on some other pursuits.

Since The Bullet got some love with its own animation, I’ve been meaning to take the lessons learnt and apply to them to another animation. Atlas, Broken would be too hard, and while Grosvenor Lane would do well with dark silhouettes and spooky music (I’m counter-convincing myself now… damn), Adaptation needs to get some attention.

Why an animation? Books don’t get read unless you can attract a pair of eyes to look at them. I’ve said it once and I’ll say it again, as an indie author, you are responsible for getting your book seen. How you do that is up to you (just maintain your integrity. And your dignity while you’re at it.), and an online animation is just one way.

The Kit

So I’ve gotten myself kitted out with my tools once more:

  1. A pencil and paper for sketches and planning
  2. Gimp to handle importing and cleaning my sketches up
  3. Inkscape for converting things to Vector graphics
  4. Synfig to animate the whole show
  5. Anvil Studio to create a Midi Track
  6. Virtual MidiSynth and Soundfonts to give richer sound
  7. Audacity for any vocals, sound effects, etc.
  8. Window Movie Maker to plop the bits together and convert the final product to be presented on YouTube

With more of a physical, as opposed to a metaphysical story, to work with, the animation called for more ‘scenes’. My first thought was to make everything from the point of view of Ottavio or Ryan, but then I thought, no, the book isn’t about them, it’s about the world that they are in.

So I scrapped that idea and took a different approach: The promotional video is there not to tell the story, rather it tells the viewer about the story. It’s a front cover on steroids. Its a blurb that gets shown. It’s a chance to see the bits of the book that lets the reader know that the book is right for them.

The Plan

And so I looked at my options: I could play out a pivotal scene from the book. That sounded good, until I realised that no particular scene defines the intention of the book. Sure, it’d be easier given that I’d only have to make one set of drawings or scenes, but I’m not after easy, here, I’m after something I can look at and think, “Yeah. Happy with that.”

I then thought, “Why not a voice-over reading out the blurb”. No. No. No. I mean, that’s fine, soundwise, but a video wants some video. And it would be akin to a powerpoint presentation where the presenter reads out the dot points that the viewer can read for themselves. No.

So then I thought about movies, video games and television shows, and how they tended to present their entertainment: Snippets. Stills. Short clips of stuff. It gives a general feel of what it’s about, a couple of poignant comments or quotes, but it doesn’t hit the user over the head with information.

And that’s where I’m headed. Over the next few weeks I’ll be posting about the creative process, and I’ll share some of the drawings and music as I’m going along. The last post on animation was done retrospectively, whereas this will be a ‘work in progress’ one.Mini Jeztyr Logo

The Bullet Animation – All Together Now

Let me go back to where these updated began: As an independent author, it is up to me to organise any form of marketing or promoting of my books. To this end, I embarked on an adventure – Yes, I’ll go as far as to call it an adventure – to create an animation about The Bullet. Let’s see how this came together.

The Pieces of the Puzzle

Hindsight is a wonderful thing. Here is a rough chronological list of my tasks:

  1. I considered what I was after. I made a plan, sketched out my ideas into scenes, refined these down to what was I considered was doable, selecting five main sections.
  2. I researched software that was available for sketching, vector drawing and animations and downloaded Inkscape for creating the vector graphics, Synfig for animation and Gimp for image manipulation.
  3. I sketched out my characters faces and brought these into a digital format, converting them to vector graphics.
  4. Using Synfig, I created my scenes, one by one, according to my original design.
  5. I recorded a bunch of sounds on my phone, uploaded these to the machine and edited the soundwaves with Audacity, and hunted down a gunshot for the climax.
  6. With the aid of Anvil, I wrote the musical track.
  7. I used VirtualMIDISynth and the “Fluid GM” Midi Soundfont to get a richer sound
  8. I exported the music from Anvil and blended this as a separate track together with the sound effects in Audacity.
  9. I rendered the animation from Synfig to a movie file.
  10. Lastly, using Microsoft’s Movie Maker, I added the audio to the video and exported the whole shebam to a YouTube ready file and uploaded it.

The end result is a one minute and twenty second clip that I’m pretty chuffed with:

Sure, it’s not refined, it’s not going to win any medals. If I get to do it again, if I ever have time, there will be several things I’d concentrate on.

Post-Mortem

In the programming world, we use retrospectives or post-mortems to see what went wrong, what went right and what can be done better. Forgive me if I cannot resist giving the animation the same treatment.

The first issue that jumps at me is the lack of sophisticated motion. It was suitable for what it needed to be, and that’s fine, but as I think about how I might create other animations, I figure there will be more ‘going on’. Background motion, moving lips with synchronised speech, blinking eyes, torsos turning, limbs flailing. While too much can be distracting, too little can be boring.

The music I enjoyed. A lot. Creating it piece by piece, getting the soundfonts, discovering reverb and chorus (albeit too late to apply it) and adding tracks as layers was just fun. Pure and simple. I reckon I could lose hours just knocking out tunes and mucking about with rhythms.

Then comes the sound. That was a headache. It was the opposite of fun. It doesn’t matter how I look at it, it just didn’t sound ‘right’. I guess I just don’t have the skillset or the proper equipment for sound engineering, so I’d probably ask for help, or try and find someone to hire.

Likewise with voice-overs. I think a voice-over would have been great. Again, lousy recording equipment and an even lousier voice let me down, to the point where I omitted the voice-over altogether. For this I’d definitely hire someone with a voice appropriate for the context.

Lastly, I think the sketching and vectorising the characters worked out just fine, only I’d spend more time getting details and layers so as to add more dimension to them. And I’d really like to try the ‘bones’ feature out in Synfig and get some complex motion happening. Oh, for another lifetime!

In any case, I’ll call it a wrap. I’ve got to get back to writing, so I bid a fond farewell to the Land of Animation – for now. I’ve got my little bag of tricks for next time, and I hope to share with you my next foray when I get a bit of breathing space between titles.

The Bullet Animation – Sound

Up to this point I had been toying with graphics and animating things and sketches and learning about vector graphics – and I’d completely neglected the audio! Well, not completely. Mostly.

The issue, as I saw it, is that I had to finish the animation before I could add the sound. I could hardly hope to figure out the timing without something against which to time. Anyway, by the time I got to the first major iteration, I thought I’d better spend some time on sound.

Music, Sound, Voice

I broke up the tasks of sound into three main categories: The background music, ambient sounds and voice-over. I chose to go without a voice-over for reasons previously mentioned, but I think I’d like to give it a try in the next animation. I can imagine it would present its own challenges and I’d like to explore them one day.

For now, this post will concentrate on the ambient sound, the next will be on music.

My first task was to think about what scenes needed what sounds. Going back through my animation files, I watched the silent progress and imagined what might lend itself to the matter. I made a wee list:

  1. The hissing of the furnace
  2. The rattling of the conveyor belt
  3. The kak-klunking of the machinery
  4. Heavy breathing of the Assassin
  5. The bang of the Bullet

Armed with my dodgy microphone, I tried my hand at making noises with my mouth. I discovered a couple of things. Firstly, my microphone ain’t no good. I thought it was broken at first. No, not broken, just really crappy. The resulting sound was barely above a whisper. Upping the gain only upped the noise and clipped the sound. I couldn’t make too much noise: I’ve got a young ‘un who is usually asleep by the time I’m doing anything. On top of that, everything came through with a hum that I later tracked down to being the fan of the computer.

Secondly, while my vocal impersonations of the garbage truck on a Friday are enough to impress small children, Michael Winslow I ain’t. Even when I did manage to get a sample of something loud enough to be workable, it sounded pretty lame. The rattling conveyor sounded like an old man about to lose his lunch, the kak-klunking sounded like nutshells being rubbed together.

Take two

Microphone = Inadequate. Location = Terrible. Source = Abysmal. To address these issues, I looked at the palm of my hand. My phone! Not only can it take telephone calls, it has a recorder built into it. On the weekend, I buzzed about outside, in the garden, in the garage, trying to find sources for sounds. The roller door. A hammer. The hose. The air conditioning unit. The lawn mower. The can opener. There were clunks and rattles and hisses and sighs all over the place.

By the time I came back inside and thawed my nose (it’s Winter time), I had a phone full of sounds, ready for use. Only, they weren’t. First, I needed to download and convert them into something usable.

Audacity

A long, long time ago, Dad splashed out on a Sound Blaster Pro. Tucked into the whopping ISA expansion slot on the motherboard, it allowed, for the first time, not only playback of awesome sounds and music, but also recording of awesome sounds. As a family we huddled around the box to record funny messages for windows startup, add reverb and warp the pitch until we sounded like chipmunks.

When I tried the recorder the other day, I was sorely disappointed. Yes, I could record, but that was about it. Where had all the fun gone? Why couldn’t I fade in or out? What about the echo and hiss-reduction and all of that. We had it back in the 90’s, right?

AudacityWell, all of that is still there. A quick search on the net brought me to Audacity (http://audacityteam.org/). Simply download and enjoy. The interface was a bit daunting to look at, granted, but stick with it. Go ahead, import a sound file, boom, there’s the waveform, ready to be fiddled with. First port of call for me was to trim out the bits of the samples that I didn’t want. Highlight the section and delete it, simple as that. And if you need to insert a block of silence, sure, select menu > generate and make as much sweet silence as you need. I had to do that a fair bit: I had a two year-old shadow following me around, nattering all the while.

You can cut and and copy and paste, or select a region and make it repeat x number of times. Importantly, you can fade in or out, cross-fade between left and right channels, or apply some really cool filters to knock out high hisses or low hums. I must admit, I lost myself for quite some time as I mucked about with different filters, seeing what each one did.

One of the really cool features of Audacity is the ability to have multiple, parallel tracks. They end up working a lot like layers in the graphics programs, so you can tweak one track independently of another, speeding it up, slowing it down, adjusting the volume, whatever you like. And you can play it back, just like that, to hear how it goes.

In the end, banging a lump of wood on the roller door provided a decent ka-thunk, ka-thunk, and the air-conditioner gave up a bunch of interesting sounds, whirrings and groans and squeaks and hisses. One thing I couldn’t find in my garage wonderland of noises was the distinct sound of a rifle shot.

Essential to the animation, I simply could not recreate a convincing bang that was distinctly a gunshot. Short of rocking around to a rifle range, I poked about online to find free online noises. I listened to the report of a few different models and settled on the Springfield M1A rifle: It has that heavy crack that I was after, along with a lasting, gaseous hiss.

Pushing this all together, I must say that I’m not entirely happy with what I ended up with. If there’s anything I’d go back and do again, it’d be the sound, simply because its just not punchy and distinct enough. In fact, I’d probably seek help from a sound engineer in this department. Anyway, enough prattling about that, next time I’ll prattle on about the music.