I used to travel the world and train engineers on this very topic a few years ago, and it’s somewhat disheartening to see that even the basics of optical recording have become Wisdom of the Ancients. It took decades for thermionic valve (vacuum tube) skills to almost become extinct; it’s taken less than 10 years for CDs to go from maturity to dimly-remembered folk tale. The howstuffworks article is peppered with glaring errors (no disrespect to powerjacob, it’s a well-presented article and only an expert in this very narrow field would pick up on it).
In answer to the main question, there are indeed two lasers, 780nm (infra-red) for CD, and 650 nm (red) for DVD, and for the most part they share the same optical path within the optical pickup head (OPU). The amount of data that can be written to a disc, all other things being equal, is inversely proportional to the wavelength of the laser, which is why a 405 nm (violet verging on ultra-violet) laser is used for the higher capacity Blu-Ray.
One might wonder why the more accurate 650 nm DVD laser isn’t used to write to a CD as well, as surely this would be a bit of a holiday for it? There’s a good reason why not, but sadly I don’t recall the details. Sorry. There’s some awkward parameter in the way, like the CD disc refracting too much at this wavelength, or somesuch. This is why this knowledge is being lost, it’s the old fools getting forgetful…
As a general rule, the lower the laser wavelength, the more complex the system. This is because dealing with the smaller geometries of the data written by a lower wavelength laser becomes increasingly fiddly, and the servo systems to control the OPU sled, laser spot positioning and focus need to be increasingly accurate; for example, DVD needed to introduce a tilt mechanism into the optics to compensate for disc bowing, and Blu-Ray needed to introduce even more axes of movement. In fact, even the relatively simple CD mechanism is so complex that no one person can be an expert in all aspects, just in a small area. My particular field was operation of the laser, and write strategies. The latter are the complex series of pulses sent to the laser to form the written pits in the disc, and these will vary for different disc types, different speeds, and different manufacturers.
The same laser is used for reading and writing, the difference is in the optical power levels and the signal sent to the laser. Reading involves a steady average optical power level, modulated at a few hundred MHz or so to help reduce fringe/etalon noise (an unwelcome optical effect), while writing involves a series of much higher power pulses. Additionally, CD-RW, DVD+/-RW and Blu-Ray can be erased, and this involves a power level somewhat lower than the write level, but still many orders of magnitude greater than the read level.
This is a vast over-simplification, of course, it’s a horribly complex topic, and the people who spearheaded the development of these systems were geniuses. Now I’m not too shabby at my job, but I know when to be humbled in the presence of greatness, and I was privileged indeed to work with the people I did back then.
If you like I can dig out some old presentations on the topic; strictly speaking they’re the IP of Philips Optical Storage, but since the Philips parent company did the dirty and closed down this division I don’t feel like I owe the CEO any favours. In the meantime, here’s a historical perspective from acknowledged authority on the subject Dr Sorin Stan, genius and very nice chap indeed.