Green Screen Webcam For £5

Amaze your friends and irritate your work colleagues by setting up your own custom video background to webcam calls for a total outlay of £5!

Commercial software is out there which will attempt to extract your image and replace the background – but it’s (a) expensive to buy and (b) demanding on the processor. Far cheaper to go old-school and use chroma keying – this will enable you to use open source software to replace your background, and seems to run fine on my pre-i3 box which I got off eBay for £50 a few years ago.

1. Get Green Screen Fabric

Search eBay for “green screen fabric”. You should be able to buy 1×1.6m / 3x5ft fabric for around £5 – larger amounts will cost more correspondingly but I managed to get it working with this size.

You can jury-rig a way to hang it behind you (my own solution involves a stepladder, mop pole, g-clamp and 4 large paperclips) or buy a backdrop stand for around £20 more (search eBay for “photography backdrop stand”).

2. Install Open Broadcaster Software

This is free. Go here: Open Broadcaster Software project

3. Install the OBS-VirtualCam Plugin

This is also free. Go here: OBS VirtualCam Plugin

4. Set up ChromaKey output based on your camera feed, and make that available as a virtual camera

The basics:

  • Set up a profile with 3 sources – your webcam, your microphone (for audio) and a media source (which should be below your webcam in the layering order).
  • On the webcam, right-click, pick “Filters” and add “Effect Filter” of “Chroma Key”. You may need to tweak the settings slightly to get a clear image
  • On the media source, set up whatever you want. Don’t worry if it’s wider than the camera.

Some more detailed instructions (the latter includes how to make the output available as a camera):

5. Pick your background

Both static backgrounds and videos are possible by setting them up in OBS.


We’re working our way through Corbyn and many of his supporters not being on board with representative democracy [0]. Makes it kind of tricky to work with the PLP [1]. The end-game is probably a split where you get a representative democracy party and another party based more on Momentum-style direct democracy [2]. One is likely to be a lot more effective than the other at getting things done [3].

[0] representative democracy: you elect a representative and they have autonomy once they are elected, check out Burke etc. So to say that the PLP rebels are acting against the ideals of representative democracy doesn’t really make much sense.

[1] PLP: they tend to be on board with representative democracy since that’s the water they swim in

[2] representative democracy => oligarchy. direct democracy => cults of personality and mob rule. Pays your money and takes your choice.

[3] I’ll bet my money on whichever one is aligned more effectively with the way power works in the UK political system. Which is…

The Glass Bead Game

I read The Glass Bead Game more than half my life ago. It’s about a monastery where the monks are devotees of the aforementioned game, which brings all artistic and scientific endeavour together in a single unified form.

I didn’t like it at the time, and I don’t like it now. I am, at least, clearer on why I don’t like it now, for which I am sure Joseph Knecht would give thanks.

The idea that it is somehow virtuous, somehow positive, to unify diverse fields of endeavour – I don’t buy it.

I have spent much of my working life trying to get people to work more effectively together when tackling problems with a significant analytical content, but a key part of this has always been the acceptance of multiple, irreconcilable viewpoints.

And I love mathematics that makes deep connections between superficially unrelated fields, but maths also has numerous contenders for how one should think about its very foundations (category theory, set theory etc).

“The Glass Bead Game” has the underlying assumption that some kind of global unification of human endeavour is intrinsically a Good Thing. Why would anyone think that?

Conway’s Law’s_law

“organizations which design systems … are constrained to produce designs which are copies of the communication structure of these organizations”

A more generalised statement of Conway’s Law is:

“technological systems and political systems mirror each other”


1) If you try to change the deep structure of technological systems, this will have major political implications. Don’t assume that you can put an information bus or microservices in place as a purely technological fix. The more effective any such change will be, the more of the current political landscape you will be invalidating.

2) More generally, you can’t fix primarily political problems by technological means. If you are getting too far ahead of the political change you wish to effect, you will be pulled back, no matter how good the technology is.

3) You can’t fix primarily technological problems by political means: you can’t make up for the lack of fundamental technology by building coalitions, no matter how well-intentioned everyone is.

Why Liberals Are So Goddam Smug

There’s a recent book out called The Righteous Mind that has an interesting take on how we analyse morality, and how this analysis differs between Liberals and Conservatives. The thesis is that rather than being rational, morality is largely informed by instinctive feelings – and these feelings differ in a significant way between the two political groupings. While both groupings are motivated by considerations of Fairness and Care/Harm Prevention, the Conservatives also refer to notions of Loyalty, Authority and Sanctity/Purity to which the Liberal wing are, on average, tone-deaf to.

One end result being that the Liberals get all aerated about how those dumb Conservatives spend so much energy on stuff that is Obviously Completely Irrelevant because they are So Dumb It Isn’t True (Or Maybe They Are Evil Instead).

The characterisation of Liberal concerns fits me well – the poem “Girls!” by Stevie Smith gets my take on Loyalty and Authority spot-on, and I view notions of Sanctity with the bemused incomprehension of a Martian Anthropologist.

That’s fine and all, but then the question is how a Liberal political movement can hope to engage a stable majority coalition if it’s blind to 60% of the base moral drivers of, say, 50% of the voting population. The answer is that they have to at least be conscious of these drivers and not pretend they don’t exist when framing political issues.



Girls! although I am a woman
I always try to appear human

Unlike Mrs So-and-So whose greatest pride
Is to remain always in the VI Form and not let down the side

Do not sell the pass dear, dont let down the side
That is what this woman said and a lot of balsy stuff beside
(Oh the awful balsy nonsense that this woman cried)

Girls! I will let down the side if I get a chance
And I will sell the pass for a couple of pence

- Stevie Smith, from “Mother, What Is Man?” (Cape 1942)

I Am Surrounded By Idiots

…who don’t vaccinate their children. That’s how I feel today anyway, since I am currently a few weeks into the joy of adult-onset Pertussis, more popularly known as “whooping cough” – there is a big uptick in cases in Surrey and South-West London at the moment. Not-entirely-coincidentally, uptake of the Pertussis booster vaccine is very low in at least one London borough, Wandsworth.

However once I get my irritation over my personal circumstances to one side, vaccination uptake is an interesting area.There will always be those convinced that vaccination is the work of the devil, but for the rest of us there are still some deep issues that people need to get past. Some of the issues thrown up in presentations like this:

  • ambiguity or doubts about the reliability of vaccine information, helped by the media “showing balance”
  • a preference for errors of omission over errors of commission
  • instinctive aversion to putting pathogens inside ones child
  • personal acquaintance with someone who thinks vaccines have damaged their child.
  • recognition that if many other children are vaccinated, the risk to unvaccinated children may be lowered

All of these need different strategies to counter:

  • Information could be improved by having a better web presence that was not funded by pharma companies (as part of the problem is distrust of those companies).
  • Omission/commission is an intuitive bias that is hard to counter – obviously not doing something is just as much of a choice as doing something, but this is hard to put across clearly.
  • Instinctive aversion is also hard to counter. A start would be to at least acknowledge that people feel uncomfortable about this, and to explain why they feel this way – rather than to dismiss their feelings out of hand.
  • Personal experience is very powerful. But this can be also used positively, eg by pointing up incidence of vaccination among the health procession.
  • Relying on others to provide herd immunity could be countered by explaining how some people really need others to provide herd immunity, because they are immunocompromised – so relying on herd immunity for a healthy child is re-framed as taking resources from sick people.

And in fairness, it is very easy to see associations where none exist – for instance, when one of my own children had just been vaccinated, they vomited that night. “Aha, side effects” I thought to myself. Then I started vomiting too. As did my other child. Because we all had a stomach bug.

Think I will look for more research in the vaccination uptake field. Once I can stop puking whenever I cough.

Physics Envy

Physics Envy May Be Hazardous To Your Wealth correctly diagnoses a problem in Finance – how people take models aimed at capturing aspects of material reality, and apply them to complex emergent social phenomena (such as pricing multi-tiered wobbly derivatives on top of US sub-prime mortgages).

Their solution is to have more quants - in other words to have more people with the level of mathematical knowledge required to engage with the models. I am sceptical.

If you can’t explain an idea in mathematical finance clearly, using plain English, on one side of paper, then it’s hard to see how things will end well. Mathematical finance needs more people whose primary focus is on how the maths is put to practical use – the last thing the field needs is more specialists in maths. The more people are labelling themselves as quants, the sharper the divide, the more severe the disconnect, and the bigger the eventual problems.

Let’s Kick Off With A Classic

Evidence of heuristic traps in recreational avalanche accidents

“Even though people are capable of making decisions in a thorough and methodical way, it appears that most of the time they don’t. A growing body of research suggests that people unconsciously use simple rules of thumb, or heuristics, to navigate the routine complexities of modern life. In this paper, I examine evidence that four of these heuristics – familiarity, social proof, commitment and scarcity – have influenced the decisions of avalanche victims.”