31 days to go…
I was quite happy about how well I was doing last night, then towards the end of my work day I spotted A Big Problem with one of my chapters. See, early on in the chapter I show an experiment where I shine two lasers onto my sample – one is a strong pump laser and one is a weaker measurement laser. By measuring the spectrum of the latter, I show that you get an unwanted disturbance when you put the pump laser on full power (but not when you attenuate it a little). So, surely the next experiment in the series takes account of that.
… no. We did the experiment at full power. That, it seems, explains perfectly the signal that we saw. Now, the next experiment after THAT solves this problem, but it’s sort-of frustrating to have to defend an experiment which I know in retrospect wasn’t very good. On the other hand, hindsight is 20-20. An interesting fact of the environment in which I have done all of the experiments for my PhD is the severe time constraints under which we’re forced to perform. Our experiments need light from a machine called a free electron laser (FEL), which is so expensive and difficult to maintain that it takes an international facility and a team of tens of people to keep it running smoothly. As you might imagine, time with the laser don’t come cheap. That constrains the time you have to do experiments in – 8 hours at a time, usually – as well as how easy it is to iterate/improve upon your experiments. Essentially: you can’t unless you have a really good reason.
Oh, and there are no alternatives, no ways to test your experimental plan before it goes live. And it’s always at night, so you’re sleep deprived. Frankly, I find it remarkable that anyone gets science done under those conditions.
I’d like to think that I’m not one to give up easily, so I make the best of this catastrophe. It turns out to make me think a bit harder about the science behind the measurement, and in the end I think that my chapter is better for it. I’m pleased, and I spend the rest of the afternoon finishing the chapter off by polishing the introduction to the chapter. It does tie together quite well. As a thing I’ve been working on since November, I’m pleased to say I finally don’t hate it.
As I’m wrapping up, I decide to compile into the full document and update my supervisor with all my progress (two chapters since Tuesday!) – and the University web domain goes down, along with my connection to its servers, and any chance of me finishing my work for the night. Great. I don’t think I lost anything except the feeling of satisfaction.