Irregular Webcomic! #1731

Dun duhn duuuuh!

The etymology of the word "titbit" is interesting. As best I can ascertain without access to a copy of the Oxford English Dictionary, the original form was "tidbit", from the Middle English tyd, meaning choice or special, and bit, meaning a small morsel. At some point the British converted this to "titbit" for some reason I haven't been able to uncover, and this spelling and pronunciation is now the most common in the UK and Commonwealth nations. The "tidbit" spelling remains as an alternative in use in the USA, although it seems to have been a relatively recent re-invention, appearing in the US only as recently as the mid-19th century. It's not that the US has preserved the original spelling, but that they have for some reason gone back to it after an intervening couple of centuries when everyone used "titbit".

There is some speculation that the (relatively) recent American change was prompted by a prudish desire to sanitise the language of "rude syllables", changing the potentially titillating (pun intended) "tit" for "tid". However, there doesn't appear to be any solid evidence for this as the reason.

2018-06-06 Rerun commentary: It's interesting that you can see the rather obvious barrel distortion of the camera lens in these shots. Notice how the vertical wooden parts of the bookcase bend in towards the top and bottom.

This is interesting because these days almost all compact cameras of the sort I used to take these photos have built in digital distortion correction, so that you don't see the distortion of the lens. This is done by calibrating the distortion of the lens during the design phase to calculate the geometric transformation from a rectilinear image to the distorted one produced by the lens. Then, after an image is captured, the camera's processor applies the inverse geometric transformation to the pixels of the sensor data, remapping pixels of the distorted image to where they would be if the image was not distorted. (This involves some interpolation to calculate the value of pixels that fall at non-integer pixel positions.)

The reason this is done is twofold. Firstly, most people prefer their images undistorted, so the image that comes out of the camera looks better for most purposes. Secondly, it's significantly cheaper and easier to design and build lenses with noticeable distortion than to build ones where any distortion is too small to notice. This means that camera manufacturers can produce smaller, cheaper cameras with better image quality than they could if they designed the lenses to be less distorting.

Some compact cameras can be modified (usually by firmware hacks such as CHDK) to save the pre-undistortion images, so you can see what they look like. Thanks to an old article by DPReview, you can see and compare shots from a Canon Powershot S90 camera: before distortion correction - only visible by hacking the firmware to save the uncorrected file, and after distortion correction - the default image saved by the camera. (The article compares three other camera models too, and explains a lot more about how and why this is done.)