The system we use creates a score for new comments based on a variety of factors. Comments with a very low score are immediately binned (though stored so we can review them), those with a high score (from known users, etc.) are immediately published. Some of those in the middle lead to a request to complete the captcha, as a way to verify the authenticity of the author.

MJ is very right that captchas have limitations, and are problematic for some users. The notion here is that by using them as one of several available techniques we can better manage what can otherwise be a deluge of spam (I’ve previously dealt with thousands of spam comments a day on Movable Type, leading me to cur off commenting ability completely and ultimately move to WordPress). However, we have now disabled the captcha, at least temporarily, and will be monitoring the situation to check the impact on the site. If we can make do without the captcha we’re more than happy to do so.

If spam is a problem, install filters that combat spam, like content-free comments with links to known-bad sites. Your image test is actually more solvable by software than humans – you’re mainly lucky that you’ve not been attacked by suitable software often yet.

If spam is a problem, install filters that combat spam, like content-free comments with links to known-bad sites.

I’m not sure of a simple way to modify a set. My electronics are rusty, to say the least, but my thinking is that changing the size of the antenna (wound around the ferrite bar) will merely change the wavelengths that are optimally picked up by the set, but not what frequency the receiver is tuned to.