From the moment we decided to create our online course, to the moment we pushed the button to publish it on Udemy, about a year went by. And it was a year full of learning and discoveries, not just on the blockchain front, but also on the “how to make an online course” front. Now that the course has been online for 6 weeks, we thought it might be interesting to share how we did it.

An online course… to learn how to make an online course

Talk about recursion.

One of my most precious resources to get started was Sarah Cordiner’s “How to create your own profitable online courses”.

Her course was really full of useful tips and theoretical knowledge specifically adapted to created an online course on Udemy. Two key takeaways:

1. You don’t teach an adult like you would teach a kid
2. You don’t teach online like you would teach in a classroom

And a lot of the things we ended up doing, I actually learned from her. So if you are interested in creating your own online course, I can only advise you to have a look at this specific course, and even at all of her other courses on her own website. It really is a worthy investment, as well as a source of inspiration.

Trello, our best planning friend

As soon as we started planning our course, Trello was our best friend. It made it easy for us to list all the topics we wanted to cover, organize them logically into chapters, attach resources to our topics, and keep track of what still had to be done versus what had already been done. We even used it to gather comments from early reviewers.

Gathering early data and leads with Typeform

Of course, having Trello was not enough. We still had to fill it with valuable content. And we had to understand what was valuable to our potential students. So we created an easy survey on Typeform that we promoted on Facebook, on this blog, on Twitter and during our Devoxx Belgium talk in November 2016.

We used this survey to learn, among other things:

• What our potential students already knew or not
• What they wanted to learn
• How they felt regarding existing learning resources
• How much money they were ready to pay for such a course
• What were their motivations to learn about the blockchain
• Generally, who they were

Overall, we received 53 responses out of 1200 unique visits (4%), and qualitatively this information was priceless. It really helped us shape our course content.

Scripting everything in Dropbox Paper

Early on, we chose to script every single lesson in details rather than just winging it on the fly. Not every teacher makes that choice, but we did it for several reasons:

• English is not our mother tongue, so we wanted to avoid hesitations and streamline video edition with a prompter
• We know that the blockchain environment is changing very rapidly and we will have to update our course content accordingly, which is way easier to do when you can index your course text instead of having to watch every video to check when a certain topic is mentioned
• We considered the possibility of reusing this text to create written material like an ebook

Dropbox Paper was really helpful in scripting everything, as it allowed Said and I to write and review each other’s content with comments, change tracking and so on. And also, since everything was automatically saved to the cloud, it was easier to link to Trello cards.

Keynote

For all the slides in the first chapters, we used Keynote and icons from The Noun Project for which I have a Pro subscription. Their Mac app was particularly useful for changing the color of the icon to fit our visual identity and integrate them into Keynote very quickly.

Prompter

I know it shows sometimes, but the prompter was absolutely essential to save a little bit of time and have fluid speaking videos without any hesitation.

On the iPad, we used the Teleprompt+ app from BombingBrain. It’s an investment but it’s very flexible and can connect to a lot of remotes, like the Satechi BT MediaRemote that we used.

For the prompter setup itself, I used the Visioprompter Tablet kit, but the website I got it from is not online anymore and I can’t find any other link to it. Fortunately, there are plenty of other options, including a couple recommended by BombingBrain. The cheapest options I could find on Amazon are this one and this one (Amazon FR). I got mine for 250€ with the case. Again, an investment, but definitely worth it.

To transfer the scripts from Dropbox Paper to Teleprompt+, I saved them as text files into Dropbox, and then imported them into Teleprompt+ via Dropbox and added a few cue points to stop scrolling every now and then.

Lighting

For the lighting, it’s the part I’m least proud of. There are a few good things in my setup but there are a few things I would still like to improve.

First for the green screen, we used a couple of fabric kits from Amazon to cover all of the walls in black, except the background wall. The green screen is not perfectly smooth, even with clips, so I wonder if this option would be better, though a little bit smaller.

For the lighting situation, I should have done a proper 3-point lighting but I was not aware of that technique when I bought my gear. So I used this kind of kit. But now I’m thinking of upgrading to this kind of LED kit from Aputure, that is way more portable and hopefully doesn’t heat up the room that much.

One thing that really saved my ass is this ring light from Neewer, that does an amazing job lighting my face from all sides. Small tip: try to stand as far away from it as possible to avoid light rings in your eyes, but remember that the closer you are to the green background, the more it will spill. More on that later.

Camera

For the camera, I used this Canon EOS 70D kit that includes a 18-135mm lens with a very silent autofocus. Small tip: switch to P mode to be able to customize the microphone settings and see them on the screen. And also, enable the timed shooting feature to be able to start/stop recording using a remote like this one. Last tip: you have to disable Wifi to be able to shoot in video. Also, I used this kind of kit to plug my camera to power instead of relying on batteries when shooting. And of course, these Amazon Basics tripods are very affordable and perfect for this kind of simple setup.

Microphones

In terms of microphones, for face shots, I used a pretty standard Lavallier: the Audio-Technica ART3350. It is very cheap and it does the job, but 2 warnings:

• Buy yourself some batteries for it as it won’t warn you when batteries run low, the background noise will just increase.
• Sometimes, I have some random pops and clicks that are almost impossible to remove in post-production and I don’t know if it comes from the microphone or from the camera, but my guess is on the microphone.

Right now, I’m thinking of upgrading to the RODE SmartLav+, with the following adapter to plug it into the Canon EOS 70D, as the SmartLav+ is originally designed for smartphones with a TRRS plug.

For screencasts, I’m using the infamous Blue Yeti microphone. Here again, another tip: it works much better and without any echo if you speak into it in the side where the volume button and logo are located 😛

Here is a short video I just did to show you the whole setup in situ:

Camtasia

For all screencast videos, I’m using TechSmith Camtasia. Here again, another tip: don’t forget to push the resolution of your screen to the maximum and above 1080p resolution to make sure video doesn’t get blurry when you scale it to 1080p.

I also used Camtasia to record all my slides with animations instead of integrating individual slides into Hitfilm Express when editing.

By the way, when recording your screencasts, don’t forget the “don’t disturb” mode on your Mac, to avoid pesky notifications that would ruin your perfect take.

Hitfilm Express 2017

Probably the discovery I’m most proud of is the application I used to edit all of the videos: Hitfilm Express. This software is completely free, and amazingly powerful, especially when it comes to Chromakey edition. It’s the only affordable software I found that could automagically remove the green spill I get on all my green screen footage.

And the tutorial above shows you how easy and powerful it is.

I did purchase an addon pack though, the Repair Pack, that contains a very effective noise reduction filter to remove white noise in the audio background.

General process

I think that pretty much covers the hardware and software part.

When it comes to the general process we used, we had 2 different processes, one for the face cam lessons, the more theoretical ones at the beginning, and another process for the screencast videos showcasing some code.

For the theoretical lessons, here is our process:

• I scripted the lesson
• Said reviewed it
• I created the slides for it in Keynote
• I shot the face cam footage, trying to go for a single shot for each lesson
• I edited the face cam footage alone to get the right timing
• I played the face cam footage while recording the slides to get the animations timed right
• I edited everything together to synchronize slides with the face cam footage, and to do the chroma key integration into composite shots
• I exported the whole thing in Youtube 1080p format to upload it to Udemy

For the more practical code-based lessons:

• Said wrote the entire applications
• Then he scripted each individual lesson with many screenshots
• I reviewed the scripts and made sure that everything worked
• I uploaded the code to Github
• I shot the screencast with Camtasia
• I edited everything in Camtasia itself
• Then I exported everything in 1080p to upload it to Udemy

And after everything was published to Udemy and the course itself went live, I discovered another excellent trick to create subtitles: Youtube has an excellent voice recognition engines based on the millions of videos it generates subtitles for. So you can upload a video in private mode, correct the subtitles to split sentences and fix everything the engine didn’t recognize, like technical terms and the likes, and then you can export it to WebVTT format, which is what Udemy uses. More on how to do that in this tutorial:

We are still working on adding subtitles to all our lessons, and it’s a painstaking process, but it would be way worse without voice recognition.

Conclusion

And in a nutshell, that’s the hardware, software and process we used to create our course. Overall, we got 10 hours of video, and I estimate it took us somewhere around 100 hours of effective work to create all that, give or take 20 hours.

For hardware and software alone, we invested about 2000€ that we already recouped after about one month of course sales. Time will tell if our time investment will be profitable as well, but we sure had a great time learning how to do this, and actually doing it. And it’s not over, as we are already updating our content to account for all the constant evolution in the Ethereum ecosystem. We are also starting the preparation for a more advanced course that we hope to publish some time in the first half of 2018.

If you have any question about our process or the gear and software we used, feel free to leave a comment below. And of course, if you are also interested in transforming yourself into a blockchain developer, feel free to check out our course.

Sébastien Arbogast

Développeur passionné et indépendant, entrepreneur dans l’âme et apprenti visionnaire, je suis toujours à l’affût de ces technologies qui vont bouleverser le monde et nos sociétés. Consultant depuis plus de 12 ans, je me suis récemment lancé à la conquête de la blockchain, je donne des conférences et des formations sur le sujet (notamment en ligne sur Udemy). A côté de ça je suis curieux de plein de choses: la conquête spatiale, la moto, la musique électronique, les hippocampes albinos ambidextres… plein de choses.