Feeds:
Posts
Comments

Archive for September, 2010

550d Video: Filming Fire

One of things that many people have really enjoyed about video DSLRs are that in many circumstances you don’t need to use lights to film at night-time. The larger sensors combined with fast lenses and adjustable ISO mean that many situations which previously would have called for using lights can now be shot without them, which means without the need to carry about bulky equipment and power sources. Unfortunately this post isn’t about one of these occasions.

Filming fire spinning is basically a bit of a pain in the ass. There are two things you need to have relatively well exposed, the performer, and the fire. Although the fire props themselves kick out some light, the difference in brightness between the two tends to be far greater than a camera can cope with. Its one of those situations which really makes you appreciate how powerful and image processing tool the human eye is. Consequently without lights (even with a DSLR cranked up to 3200 ISO and a fast lens) you either get massively overexposed flames or massively underexposed performers. Either one of these means you’ve essentially failed. In the past when I’ve made fire based films with cameras like a Panasonic HVX200 I’ve needed a set of a few redheads (850w lights) to get enough light onto the performers while retaining some level of detail in the fire to make things work. For this shoot we had one redhead, and a 125w energy-saving lightbulb, and despite a near total lack of ambient light in the location, with the highlight tone priority function enabled the 550d coped pretty well, keeping a reasonable amount of detail in the flames and lighting the performers so their movements can be clearly seen.

One thing that was quite noticeable in post-production was that the image does tend to break up a bit if graded too heavily (probably because of the relatively high ISO used). The footage from a HVX stood up considerably better because of the extra colour information in the DVCPROHD codec, which has a lower resolution (720p) and higher bitrate (100mb/s) than the footage from the 550d (1080p and 44mb/s). But saying that, the DSLR’s superior low light capabilities meant that the footage didn’t need to be tampered with too heavily (although trying to make it look less like we’d hung a black sheet up did mean darkening some of the dark greys a bit).

Overall for what is a relatively cheap camera, the 550d footage held up really well in what is a really challenging set of circumstances to be shooting, there are few other situations where there’s such an enormous difference in luminance between different parts of the same image which need to be properly exposed, and once again the results are on a par if not better than what you get from ‘professional’ camcorders which cost several times as much as a 550d and a couple of lenses. This just underlined again to me what a game changer relatively cheap video DSLR’s are, they allow hobbyists/amateurs with a pretty minimal budget to create images which a couple opf years ago simply weren’t possibly without spending many thousands of pounds.

Read Full Post »

IPCC Errors

There’s a really good piece over on the Nature website about the recommendations made by an independent assessment of the way in which the IPCC currently functions. (hat tip Realclimate)

It’s worth reading because a lot of the coverage of the report has been over the top sensationalist nonsense about how the IPCC needs drastic reform as a result of serious errors. When these mainstream and sceptic pieces say errors, they all point to one example – the erroneous claim that Himalayan glaciers could disappear by 2035. One error, which was not in working group 1 (which covers the science of ACC) but appeared in WG2 (impacts adaptation and vulnerability) does not amount to a series of errors.

While right-wing media outlets and sceptic blogs spent a lot of time propagating the idea that there have been other, similar errors with the IPCC 2007 report, such as the Amazongate saga, whose main claims the Times has now retracted (the Times version is now however behind the Murdoch pay wall), the truth is that the science behind the IPCC 2007 report has stood up pretty well to the intense scrutiny which it has come under. On a related note, the IPCC chairman Rajendra Pachauri was subjected to an entirely baseless series of accusations claiming that he was making a fortune from his links with carbon trading companies. According to climate sceptics and the right-wing press he was profiteering from his position of head of the IPCC. An independent report into his financial dealings have found all of those claims to be completely and utterly false. Still, as George Monbiot demonstrates, the accusations continue to be made by right-wing media outlets.

In fact the major story about the IPCC getting something wrong is actually about sea level rise. The IPCC admitted in their assessment that they weren’t taking all factors into account, as some were too uncertain, but the reality of empirically observed sea level rise is way above the IPCC’s uppermost prediction.  This prediction wasn’t made using grey literature, it was supposedly based on solid science, and yet it turns out to be totally wrong. So why has there not been widespread outcry in the media about this erroneous prediction? Why are there not legions of sceptical bloggers using this as evidence that the IPCC reports are equal parts guesswork and science, and that their conclusions should be viewed with a large dose of scepticism?

Because the message we get from the IPCC’s most serious error is that their estimates err on the side of conservatism and caution rather than the alarmism they are routinely accused of. Which is really not any great surprise given the consensus based framework they operate within. But if the IPCC’s predictions are too conservative, then the entire sceptic argument goes up in smoke, so very little is heard about the most serious error the IPCC have made. This is somewhat telling, as it suggests that those criticising the IPCC have little interest in improving its procedures and improving its predictions. More often than not it turns out to be the same few voices attempting to sell generalised doubt about climate change in order to prevent any kind of legaslative action being take to mitigate the predictions the science makes.

Read Full Post »