After my (slightly ludicrous) experiments with projectors and video calls I became pretty into the idea of having my face and my slides in the frame of a video call.
So!
Here are some MORE pictures of what I’ve made. Check out that write-up page, the rest of this post will make way more sense if you do.
Both of these experiments are made using a virtual webcam setup - basically I’m using some software mainly used by video game streamers to intercept my webcam feed, and add extra elements to that video. The result is output as a virtual webcam that can be chosen as the camera in Zoom, Google Meet, or whatever you use.
As soon as I make something, I think of the 100 things I want to have next. That’s why prototyping is good. You don’t need to have much imagination, you just listen to what the prototype tells you.
1. What if you could count with your fingers, and big numbers would appear on the screen?
For my first experiment, I made it so that when I sketch on my iPad, the sketch is overlaid on my webcam.
The particular interaction I tried is included as a video on that write-up page: I hold up one finger and draw a figure one; I hold up two fingers and draw a figure two; etc.
It feels like this would be a neat way to provide narrative “anchors” when giving a talk. Minimum viable chapter headings. Or maybe draw a quick diagram in the air when only a diagram will do, Pulp Fiction style.
2. What if your regular slides appeared directly over your face?
This one is pretty simply but feels like it’s got some legs: I gave my regular slides a green background, which I then chromakey-removed and replaced with my face on the webcam.
Like, I’m tracking what Mmhmm is doing because I like the idea of including my slides in my video feed like I’m a talkshow host. It’s still in beta and I’m fascinated to see where they take the service.
BUT: fully blending webcam + slides, and designing for thumbnail view… that’s what I want. FOR EXAMPLE, what I found is:
It is neat to have MASSIVE TEXT over your face because it means that everyone in the audience can watch in gallery view - every face a thumbnail - yet your slides are still visible.
(There are pics of all of these on that write-up page linked at the top of this post.)
This is like speaking to an audience but keeping the house lights up, and being on the flat instead of a stage. It’s a more egalitarian feel.
You can have lists that appear over your shoulder that provide structure during long sections; you can takeover the whole image with a quote to draw focus.
And then there are some games to play: you can peep around the side of images that float in space. You can make faces at, say, a statement that undermines your point that you’ve deliberately included – as previously discussed it can be narratively useful to put yourself in the shoes of the audience by turning round and looking at a slide with them, reacting to it. And this is a way to do the same on a video call.
The system I made is pretty janky, but I’ve used it enough that I know there’s some creative potential I want to explore. Not just novelty but better storytelling.
(And although this system used pre-prepared slides I also tried live slides - typing words directly into a slide and having it appear over my face, which I was on a call. That’s intriguing, though harder to manage.)
I gave a talk on Tom Critchlow’s Discord to maybe 30 people, and it ended up being audio-only due to technical hiccups. It’s been a while since I spoke for 30 minutes straight, just my voice, no video. In that talk, I found myself wanting to be able to live scribe numbers in the air, to indicate where I was in a series of points. I wanted to play with my slides/webcam combo, and use the size of the type to communicate emphasis -= body language doesn’t work nearly as well over video as real life.
So I asked myself: when I’m doing a talk, what job am I really asking slides to do?
I think I use slides as…
- visual anchors – or rather running heads: the mini section titles at the top of every page in a magazine. People are often more comfortable when they have a sense of position and progress through a talk. Bonus: when people are watching the playback at 2x speed, this means the whole talk is easier to navigate.
- rhythm. Amazon once accidentally sent out a marketing email template with placeholder text, and it was a gold mine of copywriting tips. Top tip: Vary sentence length. It creates rhythm and engagement and without knowing it, the audience is carried along with you, and so it is for slides, just the same. It’s tough to do, using only your voice. But this is what slides are for! Quick transitions and slow transitions make a rhythm which can make a talk.
- a playful foil. There’s you, there’s the audience, and there are your slides. They can support you, they can contradict you – and you can react. You can demonstrate the emotional response you want your audience to have, rather than explaining it. Or you can be quiet and let your slide speak for you, which can be deafening effective when done right.
Sure there are graphs and diagrams and images and long quotes, and all the other things that presentation slides have in them. Content.
But a talk needs to engage or the content won’t come across anyway. Talking for a long period of time, without a conversational back and forth, is pretty unnatural, and you have to do a bunch of work to stop people tuning out. That’s what I think slides are for – at least in part, and at least for me.
What I’ve found, with this composite webcam feed, is that the slides can do a similar job for me as they do in real life - anchoring, rhythm, and play - while keeping my face full frame, not taking over the full screen, and not making it look like a pre-recorded TV show (which is distancing in its own way).
I mean, I admit this is slightly, “ooh hark at me, I’ve re-invented the freaking WEBINAR,” but I enjoy public speaking, and it seems like there’s a route to a satisfying version of the same kind of thing only from my sofa – which is how I live now, a centaur: my top half on zoom calls and my bottom half, soft furnishings.
So I’m going to keep digging in this direction.
Or rather… I’m going to try to fix my janky hacks so it functions for more than 15 minutes without accumulating up a very distracting and weird-looking 5 seconds of video lag..
The technical bit:
I’m using OBS Studio to capture the webcam and mix it with other video sources. The obs-virtual-webcam plugin (for macOS) outputs the stream as a camera source that can be used in most video software.
For slides I use Deckset as it has a built-in feature that expands type to fill entire slides, and also because I can simply type into a text document to quickly make new slides while I’m on the call.
To capture the iPad screen I’ve been using AirServer to run an AirPlay server on my Mac, and that can direct video into OBS (add a Syphon source in OBS and choose your iPad once you’ve started screen sharing).
It’s all pretty slow – I have to close applications, quit my network monitor, etc, and the lag still builds up. I’d like to hear about ways of shortening the video path if anyone has any ideas. I want to continue having this appear as a virtual webcam so it works in all kinds of video call software.
After my (slightly ludicrous) experiments with projectors and video calls I became pretty into the idea of having my face and my slides in the frame of a video call.
So!
Here are some MORE pictures of what I’ve made. Check out that write-up page, the rest of this post will make way more sense if you do.
Both of these experiments are made using a virtual webcam setup - basically I’m using some software mainly used by video game streamers to intercept my webcam feed, and add extra elements to that video. The result is output as a virtual webcam that can be chosen as the camera in Zoom, Google Meet, or whatever you use.
As soon as I make something, I think of the 100 things I want to have next. That’s why prototyping is good. You don’t need to have much imagination, you just listen to what the prototype tells you.
1. What if you could count with your fingers, and big numbers would appear on the screen?
For my first experiment, I made it so that when I sketch on my iPad, the sketch is overlaid on my webcam.
The particular interaction I tried is included as a video on that write-up page: I hold up one finger and draw a figure one; I hold up two fingers and draw a figure two; etc.
It feels like this would be a neat way to provide narrative “anchors” when giving a talk. Minimum viable chapter headings. Or maybe draw a quick diagram in the air when only a diagram will do, Pulp Fiction style.
2. What if your regular slides appeared directly over your face?
This one is pretty simply but feels like it’s got some legs: I gave my regular slides a green background, which I then chromakey-removed and replaced with my face on the webcam.
Like, I’m tracking what Mmhmm is doing because I like the idea of including my slides in my video feed like I’m a talkshow host. It’s still in beta and I’m fascinated to see where they take the service.
BUT: fully blending webcam + slides, and designing for thumbnail view… that’s what I want. FOR EXAMPLE, what I found is:
It is neat to have MASSIVE TEXT over your face because it means that everyone in the audience can watch in gallery view - every face a thumbnail - yet your slides are still visible.
(There are pics of all of these on that write-up page linked at the top of this post.)
This is like speaking to an audience but keeping the house lights up, and being on the flat instead of a stage. It’s a more egalitarian feel.
You can have lists that appear over your shoulder that provide structure during long sections; you can takeover the whole image with a quote to draw focus.
And then there are some games to play: you can peep around the side of images that float in space. You can make faces at, say, a statement that undermines your point that you’ve deliberately included – as previously discussed it can be narratively useful to put yourself in the shoes of the audience by turning round and looking at a slide with them, reacting to it. And this is a way to do the same on a video call.
The system I made is pretty janky, but I’ve used it enough that I know there’s some creative potential I want to explore. Not just novelty but better storytelling.
(And although this system used pre-prepared slides I also tried live slides - typing words directly into a slide and having it appear over my face, which I was on a call. That’s intriguing, though harder to manage.)
I gave a talk on Tom Critchlow’s Discord to maybe 30 people, and it ended up being audio-only due to technical hiccups. It’s been a while since I spoke for 30 minutes straight, just my voice, no video. In that talk, I found myself wanting to be able to live scribe numbers in the air, to indicate where I was in a series of points. I wanted to play with my slides/webcam combo, and use the size of the type to communicate emphasis -= body language doesn’t work nearly as well over video as real life.
So I asked myself: when I’m doing a talk, what job am I really asking slides to do?
I think I use slides as…
Sure there are graphs and diagrams and images and long quotes, and all the other things that presentation slides have in them. Content.
But a talk needs to engage or the content won’t come across anyway. Talking for a long period of time, without a conversational back and forth, is pretty unnatural, and you have to do a bunch of work to stop people tuning out. That’s what I think slides are for – at least in part, and at least for me.
What I’ve found, with this composite webcam feed, is that the slides can do a similar job for me as they do in real life - anchoring, rhythm, and play - while keeping my face full frame, not taking over the full screen, and not making it look like a pre-recorded TV show (which is distancing in its own way).
I mean, I admit this is slightly, “ooh hark at me, I’ve re-invented the freaking WEBINAR,” but I enjoy public speaking, and it seems like there’s a route to a satisfying version of the same kind of thing only from my sofa – which is how I live now, a centaur: my top half on zoom calls and my bottom half, soft furnishings.
So I’m going to keep digging in this direction.
Or rather… I’m going to try to fix my janky hacks so it functions for more than 15 minutes without accumulating up a very distracting and weird-looking 5 seconds of video lag..
The technical bit:
I’m using OBS Studio to capture the webcam and mix it with other video sources. The obs-virtual-webcam plugin (for macOS) outputs the stream as a camera source that can be used in most video software.
For slides I use Deckset as it has a built-in feature that expands type to fill entire slides, and also because I can simply type into a text document to quickly make new slides while I’m on the call.
To capture the iPad screen I’ve been using AirServer to run an AirPlay server on my Mac, and that can direct video into OBS (add a Syphon source in OBS and choose your iPad once you’ve started screen sharing).
It’s all pretty slow – I have to close applications, quit my network monitor, etc, and the lag still builds up. I’d like to hear about ways of shortening the video path if anyone has any ideas. I want to continue having this appear as a virtual webcam so it works in all kinds of video call software.