Well, I did it. Episode 001 is done. I’ve received a fair amount of feedback privately, which is nice, but also makes me wonder about the amount of work I put into the polls and feedback stuff allowing for public feedback. It’s still early though, so I’m not that worried about it. It’s been said that I’m my own worst critic. So what have I learned, and what did I think?
I almost immediately started getting some basic questions as to how I did all of this. Elsewhere, I posted the following quick and dirty tutorial, so in the interests of efficiency, I will simply copy it here. These are the super basic steps, just as far as the technology goes.
Most people have the system requirements to record something like this episode, and if you have the bandwidth, you even have the technology to stream it live. I record it, because I don’t have the bandwidth to stream it live. Fun fact that made for some easy calculations without even needing to calculate file sizes. I record in 1080p, but the output is 720p. This episode was 30 minutes long, and to upload it took just over 1.5 hours. So my upload rate needs to triple to stream live. 2mbps would be more than adequate. Too bad I’d have to get Comcast if I wanted that, due to the quality of DSL lines in my area.
The technology I use is all open source, and freely available. So let’s break things down, shall we? The base production software I use is OBS Studio.
The interface you are seeing above is one of the two available views (toggled by hitting the “Studio Mode” button). This one is a bit more advanced. The screen to the left is the preview of what is next, the screen on the right is the output; what is actually being recorded or streamed. Modifications can be made in the left panel without affecting the output. I don’t actually use this mode often, because it involves two clicks to transition instead of just one. There is a click to update the preview, and then one click to transition the output. I do not do post processing, so the number of clicks is important.
For each scene, there is a list of sources, which are items that correspond with the output. Each of the elements you see are called sources. There is the main one, of the webcam (the red squares mean you can resize them):
There is an audio input source (in this case, my headset microphone) so that audio from me is recorded.
I also added an image source, which is the ‘Lower Third’ you see there. This image is just a png with transparency, and it’s natively supported:
So that’s how I did the thing with my name. Now, the upper left corner is another source. In theory, there are window captures that capture a program, or game captures that do something similar (OBS Studio was initially designed for streaming games), but since I have two monitors, I’ve stuck with a display capture, that catches the opposite monitor:
The image that you are seeing is a full screen chrome instance, pulled up with Google Sheets slideshow. I know it looks a bit ugly, but that is because of the next step. I manually changed the background of every slide to be green.
One thing that you can do in OBS is apply filters to a specific source.
I assigned a chroma key to the display and just chose green. This takes every instance of a green shade, and applies transparency. Note: If you look at the actual slide of Voltaire, you will see that there are two different shades of green. This does not matter since the threshold is so wide. Next episode, I will reduce the threshold, and only apply the chroma key to #00FF00.
Which results it full transparency on the ‘canvas’.
So when I enter presentation mode, you wind up with something like this:
And then I record.
You know what they say about the best laid plans of mice and men, so give me a moment to talk about what my initial plan looked like.
First, I’d clean up the office, and use a sheet as a somewhat uniform backdrop. After that, I would use OBS to figure out the framing of my shots. I knew I had two main things related to framing I had to deal with. The warm open, in which I would be centered, and once I had the section with content, I’d be off to the side a bit, so that I wouldn’t be in the way. After that, I’d make sure that the documents I referenced were downloaded and made available in the shared Google Drive folder. Then make sure the slideshow order matches my idea as to how the show content will progress. Grab screenshots of the articles and use Gimp to blur all but the title, so the slideshow looks a bit more professional, and there is a decent method of attribution for quotes and data. Do a run through, make sure that the position shifts are done smoothly; talk through it and create an outline; order the slides; order the scenes in OBS; make sure keyboard shortcuts work.
Once all that is complete, do one full dry run without stopping, watch it, adjust, and then record the actual version.
Now, what really happened?
I did clean up the office and put up the sheet (proof). But then new contributions started coming in, new ideas as far as the format presented themselves, and I really liked them. I still had a little bit of work to do before the first episode, plus I kept running into issues with the sun. Having a west facing office doors can create some insane glare.
I did an initial run, and realized that needing to move around was a bit of a pain. I’d have to move my chair, make sure that I didn’t bump into the sheet, and move the keyboard and mouse to make sure I didn’t fat finger anything. So instead, I decided I’d just move the camera. Alas, the camera is on top of my monitor, and while I can move the monitor, then everything gets messed up because the angle is a bit off. So I did what any enterprising Mustachian would do, and used some duct tape, 2X8 scraps from building a subfloor, a pen to note position, and a kabob skewer to ‘lock’ the camera on top of a piece of wood. Sweet! Now I know exactly where I need it, and when to shift. Unfortunately, since these are just chunks of wood, the transition was not smooth. What was a lad to do? Well, I practiced. Slightly rotate the wood from one mark, to another. I got pretty good, but I also knocked the thing over a few times. But, I also go good at picking it back up. I determined that I can manage this about 60% of the time within 9 seconds (the length of the title sequence). So even though I didn’t have a slide for the beginning, during the title sequence was the only off camera opportunity I had to move it.
Also, since this first run through was designed to come up with an actual outline (outline I used is available here), I stumbled over words a bit, the flow of the conversation made perfect sense to me, but probably not for the target audience of this particular episode. So I used the first recording to create the outline, and made the slides from that.
And here is when I noticed that I stopped focusing on the actual content, and focused on the production aspects of it. The nitty-gritty behind the scenes stuff. The clips and positioning that made it flow better. I posted about The Enemy of Good earlier, and I was able to recognize that I was falling into that trap. So I added a slide with Voltaire, mentioned that idea in the outline, and decided to just do it. I didn’t put it off until the next day, I just waited for the sun to go down. And that’s what you wound up watching.
So here is what I’ve learned, and what I am doing for the next episode (and going forward if I get good feedback, and/or I like it).
A new logo and font was created and found by my friend over at Technical Meshugana, since he found himself really enjoying getting back into animating. He was an invaluable sounding board, and created a few other things just based on some concepts I liked from other videos. As of the writing of this post, the banner you see up top is based on his work, and I used that entire concept to create the favicon for this site, along with the profile photo for the different social networks sites. You’ll also see an animated background he created, along with a nicer Lower 3rd based on his ideas (less is more, and minimalism is of value in places other than one’s own life).
The use of a background, and a slightly smaller active camera scene means that I can actually move it inside a 16:9 video, without needing to worry about shifting the physical camera. That means less opportunity for error, and being able to capitalize on OBS transitions without needing a media source as a buffer.
Also, since the first episode is up, I now have a timeline for the next episode. I still haven’t decided between every 3 weeks or every 4 weeks, but that’s okay. I’m already thinking and planning the next episode. The topics are decided, and the outline and slides are coming along. With a hard deadline (3/16 at the latest), that means I can’t slack off either.
The following was some feedback I received elsewhere, and since I don’t have enough to justify an after action episode, I will just reply here.
Intro was good. The 3d green text was weird. 80s retro. Like when Homer became 3d.
— Joe from Adventuring Along
As I mentioned in this post, the learning curve on making an animated title sequence was too steep to get something from the ground up built, so I used a template. I took out some jitter, and changed the music (this is just from the YouTube music library). Much quicker. I am embracing the simplistic design elements, but I’m comfortable taking shortcuts as I learn.
Ha – just have to point out that your flame there reminded me of my logo!
— Neo from ReTorch
The Technical Meshugana found this, as part of an actual font pack. It scales better that way, and all I have to do is load the font, and type a diphthong (æ). This allows for some awesome flexibility. It’s also how I made the favicon.
I haven’t gotten a lot of traffic to this site, but I still want to get in the habit of asking a question at the end of each post to drive conversation. To that end…
How’d you like the first retrospective? Is this interesting to read about? What was/is your favorite stuffed animal? Mine was a monkey, named Abu before Aladdin came out.