Introduction

Over the last couple of months I’ve made a series of video tutorials for my company, Well Played Games.  I’ve been getting a lot of inquiries about how I created them so I want to share my process.

Always Have a Plan

Having a good plan is the key difference between a video that rambles, and a video that stays engaging.  I always write a simple script before starting even the smallest of projects.  I don’t include every sentence and action.  I just make an outline that I can work from.

A good script will have the following:

1. Arranges the information you will cover into a logical progression, where later content builds upon principles presented earlier.
2. Has just enough information to remind you what you wanted to talk about, but not so much that you can’t read it in a glance as you are recording.
3. Leaves room for improvisation and being natural.  Never script out every detail.

Make a Test Run

Using my script, I run though everything I want to cover as if I were actually recording it, talking out loud about the things I’m doing.  This process if vital because it not only provides practice, but also reveals things I didn’t think of in the planning stage.  Sometimes a step requires a lot of explanation about a particular software tool that I’m familiar enough with to take for granted.  Other times the thing I was planning on demonstrating turns out to be a lot more complex than can be worked through in one video.  Sometimes you will find bugs in the thing you are demonstrating.  Honestly, I’ve never had an original plan that didn’t need modification after making a test run.

Audio Path

The microphone is the magic key to audio quality.  Those Plantronics style headsets are always going to give you crappy audio quality, regardless of how you use them or which version you buy.  This is fine for teleconferencing or many simple YouTube videos, but I needed a more pro-level sound since my videos are being sold as a product.  I’ve been using an old Audio Technica 450d microphone.  I purchased it in 1989 and used it as my vocal mic when I was a gigging musician.  It then sat in a box for about 10 years and recently went back to work on my video tutorials.  I plugged it directly into a Presonus Firebox which is connected via Firewire into my 27inch iMac.

If you want good audio quality you have to buy a good microphone.  Not a great one, but at least a good one.  An in-depth conversation about microphones could take a whole blog series on it’s own, but as some quick advice for those of you buying a new microphone, you want to find something that is essentially “entry level” for pro audio.  Anything less than that will sound horrible, and anything more than that will be wasting your money for something as simple as a video voice over or podcast.  My 450d has an XLR plug in, but if I were buying a new microphone for this use today I would seriously investigate one that has a USB connection.  USB would allow you to plug directly into any computer, or even an iPad, be incredibly portable, and can deliver some nice audio for the money.  An XLR connection requires some form of mixer and/or pre-amp that feeds into a computer.  The microphone is by far the most important choice you will make for a project like this, so don’t take it lightly.  Look for help from an audio professional if you can, and avoid all advice from retail store employees and consumer review web sites.  Although they mean well, these people have no idea what they are talking about and will steer you to crappy sounding products.

I didn’t do any sound proofing of my home office, but I do follow best practices for recording audio.  Don’t have anything else running at the time.  No fans, no open windows, no dishwasher down the hall, etc.  I had some issues with the fan noise from my 15 inch Macbook Pro, but my 27 inch iMac quad i7 is completely silent, even when capturing 1280×720 video!  If you have a noisy computer you will need to move it to another room or otherwise silence it.  It can be a tricky thing to get a beefy enough computer to capture video in real time while also being silent, but it’s absolutely important, so put the effort in or purchase a silent machine to begin with.  Any noise that bleeds through into your recording is there forever.

Capture

With the script cleaned up and the topics fresh in my mind, I move on to the capture stage.  I use Snapz Pro X from Ambrosia Software to capture both video and audio at the same time.  I can’t rave enough about this program!  It captures video and stills, and has options for including both/either a mic input and system sounds.  That is vital when I want to get both my voice from a microphone and sounds I’m triggering in the demonstration (mac audio track) at the same time.

Setting up the video size is a little goofy, but easy enough.  I specify the capture size in the Snapz Pro X app, which gives me a preview guide for what area of the screen will be captured.  I go in and out of this preview while adjusting the size of my Unity3D window to match.  It’s important to match the Unity3D window to the preview to get a cleanly framed capture at an industry standard resolution.  I’ve used both 1280×720 and 960×520.  Regardless of the capture size, the process is the same.

If the video is set up in sections I tend to capture each one independently, but other times I do the whole thing in one run.  It’s something I just feel out.  If I need to take a break or load some new Unity3D scenes I might stop capture at that point.  I also set up some audio queues for myself to make editing easier.  I will say “Break right here”, and then let the capture keep rolling while I take a drink, clear my throat, etc.  I don’t touch the mouse during this time so there wont be any jump of on-screen content between edits.  When I’m ready to get back to the presentation I say, “Starting back in 3… 2…” and then I dive right back in.  This consistency helps me easily remember what to cut out when editing the video, often days later.

Compression and Temporary Storage

When a capture is done, I save it using H.264 compression set to “Best” quality.  There is a bug in Snapz Pro X that incorrectly displays the compression quality settings, on the main screen, but the “Settings” button reveals the real configuration.  Audio is saved as AAC at 44.1 kHz, 320 kbps.  I tried saving as both AIFF and Apple Lossless, but I was getting some bugginess when opening the files in other apps, and the file size/time involved was much larger without any perceptible improvement in quality.

When I first started making video tutorials I took one short video (about 10 minutes long) all the way from capture to publish ready to make sure I had the pipeline down.  But since working that out I complete all capture for every video in my series before moving on to editing.