# screencasting: make narrated screencasts for demoing stuff
A picture is worth a thousand words. A narrated screencast is worth thousand pictures.
As an experiment I've recorded some screencasts:
gorun:
efftesting:
effdump:
goerrors:
# Format
The video size is 720x1280 so that they are convenient to consume from phones. A lot of people consume media on phones so I thought the videos should be convenient to consume for them too.
I really like how expressive this media is. When I'm explaining something, I often need to set context, e.g. write some code that demonstrates something. In the video format I can present the code and highlight the important bits from it and quickly jump over the unimportant bits. If the viewer is interested in the details then they can pause the video and study the frame. In the written format it's much harder to mark some section as non-important. I tend to err on including more context but then that makes the posts too long and too rambling.
I made the videos relatively short. I had an upper limit of 3 minutes in mind because that's what youtube's upper limit for shorts are. I sort of agree: I don't have a long attention span either. If I can't express the idea I want in 3 minutes, I probably need to split up the video into separate parts anyway. This allows the listeners to take a break and think through what I just said and decide if they want to continue or not. I often speed up the desktop action 10x so there's a lot of stuff happening quickly, I don't think it would be easy to sustain attention for longer periods for the above video style.
# Setup
Here's the process I used:
# Microphone
I got a fancy Rode PodMic USB. I can't record with it in my own room though. My room is empty so there's lot of echo. I have to go out to the living room and record there.
But for some reason the recording's volume is too low on the machine I have in the living room. I've ensured that the relevant volume settings in ALSA are high (I use bare ALSA there) and the mic's volume is 100% in rode's companion android app. Probably the usb port I'm using doesn't give it enough power or something. Doesn't matter much because I found a workaround: I put my mouth next to microphone at a 90 degree angle and shout in front of the mic. Thanks to the 90 degree angle the wind from the B, P, T sounds doesn't get picked up even from this closeness.
The gorun video had my ordinary quiet voice. I had to add 24 dB volume gain in shotcut onto the audio track. And I hate that voice. Too boring, no liveliness in it, it sounds like I'm talking from my deathbed.
I think the shouting voice I use in the other videos is much nicer so I'm sticking with that. I only add 12 dB additional volume gain in those videos. I'm actually glad that the mic was too quiet so that it forced me to find this voice.
# Hosting
The video size is about ~10 MB for each. They are small enough to keep them in git. But to keep my blog repo small I've created a separate git repo for them at https://github.com/ypsu/blogdata. Then I pointed data.iio.ie at that repo and host them from there. E.g. the goerrors video is at https://data.iio.ie/goerrors.mp4. It's pretty simple.
I thought about uploading them to youtube too. But I decided against that. I might start obsessing about views, subscribers, statistics, comments, etc. That shit is not healthy for my psyche. I think I'm better off keeping these videos just for myself on this secret blog without any statistics. I write most of the text posts for myself so the video posts shouldn't be different either.
# Takeaway
I'm quite happy with the result. I might create a few more screencasts later on. It makes me think harder on what I actually want to say and so my own thoughts become clearer to me.
I highly recommend creating such screencasts!
published on 2025-01-13
new comment
see @/comments for the mechanics and ratelimits of commenting.