My Pi camera arrived amazingly quickly when you consider how long most of us had to wait for our original Raspberry Pi orders, in my case about 8 months!
Admittedly, after the sell out of first orders of the Pi itself, I did take the precaution of getting up at 3:00am yesterday to get my order in, after being mailed that they were taking available.
To my astonishment, the slender package popped through the letter box about and hour and a half ago, just 36 hours after the order.
The basic instructions to install the camera in the dedicated camera slot on the Pi board come on a single sheet with the camera itself.
In time honoured tradition, despite reading the instructions, I managed to install the ribbon cable the wrong way round first time, so take note:
the conductors face away from the ethernet connector
The unit itself is extremely small, and the lens also, which made me wonder what the picture quality would be like.
Having managed to get around my excitement and consequent ham-fistedness in installing the unit, I turned to software installation.
The general installation instructions are on the Element14 website
Depending on how out of date your installation of Raspbian is, the update can take a while; in my case, it took about half an hour.
Upon rebooting the first issue that I encountered is that the name of the camera software has been changed. In the instructions, it's called raspicam, in the actual distribution, the stills utility is now called raspbistill. The video utility is still called raspivid as per the instructions.
Next, upon starting raspistill, I got the rather disappointing error message:
* failed to open vchiq instance
A quick bit of googling revealed that the user must be a member of the 'video' group, thus:
usermod -a -G video [your_username]
(Thanks to Gegory Roberts - http://www.raspberrypi.org/archives/3263)
After this, and another reboot, the camera was working!
Capturing a still image does take a surprising length of time, I estimate around 5 seconds. I imagine that much of this is camera setup time and that if the a program kept the camera device open the minimum interval between shots would be much shorter.
So now you can see the current, appalling state of one of my workbenches. However, this is actually a mirror image of the actual scene, it appears that the -hf flag is needed to get the image the right way round.
Overall, the picture appears to be pretty much what one would expect from a 5 mega pixel camera phone camera, note the compression artifacts from the jpeg are very noticeable in the flat white areas.
Increasing the JPEG quality to 100% eliminated a lot of the compression artifacts.
Next, a quick focus chart.
I'd actually intended to use this as a quick check of the quality of de-bayering, however, the focus was sufficiently poor that the image was too soft to be able to form an opinion on this.
Video's also seem quite acceptable for the size and price of device. I'll post some of those shortly.
More depth later.