Thoughts On Creating Good Tutorial Videos

There’s a new book from Robert Fischer accompanied again, of course, by the Fibotrader charting software. It’s backing up the book’s content with tons of charts and signal reports. In time with the launch of the book we also published a revamped version 3.5.8 (we grabbed the chance and jumped to this version intentionally, anyone sees the numbers?).

Now, having a book and the software is nice, but it won’t help you really much in actually using the software in case of questions. So what we did is to create a series of tutorials, open our own YouTube channel and put them all there. Check out the result, I’ll describe how we got there now.

Scope

It was quite hard for us to get a good handle on the tutorial scope in the beginning. We wanted some connection between all of them, one common scenario so that viewers find themselves in a familiar setting when looking at related videos.

Since we had some time we decided to not only do a landmark tour but to go for the whole thing, from website to installation to first use to details to maintenance to advanced topics. This also had one huge side effect. When recording and explaining every piece of your software, you will find so many parts which are hard to explain or puzzling to use. This is your chance. Note all of that down, even if it takes time and interrupts you. Take concrete notes what needs to be changed (align buttons, remove module, show better error messages…) and in the end hand it over to a developer or if that is you, fix it. It is unbelievable how many things we changed to the better because of these tutorials.

Speaker

If you can get a professional speaker, go for it. Most probably, you can’t. Then you have to decide: Let someone who knows what he’s talking about during the tour actually do it or script the whole thing in advance, every tiny detail and hand it over to someone with a good voice.

We decided for the first approach out of very pragmatic reasons: time, quality and nerves. In the end, it was me who did the complete walkthrough, even though there are much better voice actors around. My advantages as a developer compared to someone external: I know the software by hard, I can do the whole tutorial with bullet points only, I know the big picture, I can insert spontaneous cross references, I can react to unforseen situations or bugs (and trust me, in an 8h recording these will appear) and I can actually judge if what I am telling makes sense.

Time Frame

  • appr. 2 days: a rehearsal where we noted down the most important things to be covered in very short bullet points, mainly only module names or buttons. Also as mentioned above we noted down errors and things which needed to be fixed in the software.
  • appr. 5 days: we fixed the bugs and released a new version.
  • appr. 3 days: we grouped the most interesting things into sections, reorderd them to have a good usage scenario and started recording. The result were 12 hours of material.
  • appr. 3 days: post-production, cutting, exporting

Pre-Production

Using Windows 7 we set up a separate non-admin user account so that none of our settings, bookmarks etc confuse the viewer and we ensure that we really do everything from scratch as he would do it. If you have a company logo or something like that it may make sense to put it subtly in the background to get some branding.

Recording

We used Camtasia, an ordinary speaker mic for appr. 100 euro, a custom built stand and prepared lots of water to drink during the talk. When making mistakes during talking I simply started over with the sentence without stopping the recording. We stopped only every hour or so to be safe in case of power failure etc. This will make your life way easier:

  • don’t move your mouse too much. Point somewhere and hold still. Advantage: If you cut parts of the talk then your mouse will not jump around and viewers will not realize you cut the movie
  • say loudly if you restart something, this will make it easier to detect these sections in post-production
  • don’t speak too fast and remember, that in post-production pauses between words are your friend because then you can cut easily by simply looking at the waveform
  • don’t have too high of a screen resolution, most people will watch it in the tiny YouTube player so it will be much smaller there. 1200*800 seemed the maximum feasible for us
  • you can easily save disk space by reducing the frame rate. If you don’t have any animated software you can go for 10/s, we have some nice charting effects so we went for 15/s

Post-Production

This cost most of the nerves. One is actually happy to have finished the stressful recordings. But now you have to listen to the whole recording again, remove pauses, hick ups, mistakes and unnecessary parts. So it will take at least as long as the recording is, but since you need to cut and listen to some parts several times easily multiply with 1.5 or 2. We used Adobe Premiere for that and loaded the mp4 file which we exported form Camtasia.

After having a 8h video which we were satisfied with we went through one more time, much quicker now of course (appr. 1-2h), to split it into chunks of 5-15 minutes (15 is max for most YouTube users). This resulted in 49 videos. We exported each to WMV 9 format in 1200*800 original recording resolution.

Some words to the disk space used: The whole project with all intermediate recordings and exports takes 107 Gb. The raw Camtasia recording uses 76 Gb of that for 12h of movie.

Effects

Our recording software automatically created wobbly circles around the mouse when clicking, which I personally like a lot to get attention. We adjusted the voice in the end with Adobe Soundbooth to be louder, normalized and also added a tiny echo so that the dry recording room is not so audible.

We also experimented with background music. There are really good sources for free music around where this page can be a good starting point. In the end, we decided against it. First, it took too much time to find music for 8 hours. Second, it distracted too much.

YouTube

Since we split the tutorials into 15-min-max chunks already, we could now easily upload them to a new channel we created. Click upload video, select all of your files and it will start crunching. We use the same description and category for all videos, but have custom titles and tags. It makes sense to number the titles since they are consecutive and build upon each other. You can also tell customers later on to “watch tutorial number 4 on ascii data” and he will find it easily.

It might also make sense to set your channel type to director. This will result in your channel being promoted after a while so that the 15 minute limit will not apply anymore. With an empty channel you won’t get rid of the limit though, but after having the tutorials for a while you can then also upload the full tutorial in one piece if you want to.

Leave a Reply

Your email address will not be published.