Chip's CD Media Resource Center:
CD-DA (Digital Audio) 2

Reading with a Laser

CDs are read using a laser with a wavelength of 780nm (in air). But the polycarbonate has a refraction index of 1.55, so the laser wavelength in polycarbonate is 780/1.55 = 503nm. The pits/bumps are precisely calibrated to be one quarter wavelength in height: 125nm. The result (remember that the pits are really bumps) is that light reflected from a land travels 1/4 + 1/4 = 1/2 wavelength further than light reflected from a pit.

Now here's the tricky part: Since the light reflected from land travels a half wavelength further than light reflected from a pit, the two reflections are exactly a half wavelength out of phase. This means they cancel each other: the combination of the two reflections is darkness.

Diagram from Kuhn EE498 lecture notes

I don't think the diagram above captures what's happening very well, but currently it's the best I have. Here are the points that I think are important to take away, along with some questions that remain. Remember, I'm no optics expert. Just trying to make sense of what people tell me.

• The total reflected light from the laser spot is bright when it's over land.
• The total reflected light from the laser spot is dim when it's over a pit.
• Pits themselves aren't any less reflective than lands. It's the phase cancellation effect that makes them look dim.
• In order to get this kind of phase cancellation over a pit, the laser spot must be shining in equal intensity on the pit itself and on the adjacent land area between the tracks. In other words, when reading a pit, there should be equal amounts of light (a) reflected from the pit and (b) reflected from the adjacent land.
• If the laser spot is so small that it completely fits within the boundaries of a pit, then there isn't going to be any phase cancellation in the reflection. The reflection from the pit will be be just as bright as the reflection from a land. I'm sorry to belabor this, but personally I held a complete misconception on this point for a long time.

Tracking With a Three Beam Pickup

The diagram below shows the path of a three beam laser pickup as it tracks a CD. I was trying to avoid getting into pickup design at all (Kuhn has a good description), but ultimately found that some understanding of the three beam design was helpful in understanding the general optical principles behind a CD, and in seeing through some misconceptions that commonly arise.

Diagram from Kuhn EE498 lecture notes Note that this incorrectly shows the tracking beams as centered between adjacent tracks. Pit width is shown as 0.5 microns, but that might be incorrect.

There are two extra beams, one on each side of the primary beam (thus three beams). The reflections of these extra beams are used to track the land area between tracks, on each side of the track being read. The idea is that if the pickup starts to get off center (as shown below), the reflected light from one of the tracking beams will grow dimmer as part of its light reflects (out of phase) from pits. Good tracking depends on equalizing the differential between these two tracking beams.

Diagram from Kuhn EE498 lecture notes

The Size of the Laser Spot

It should be evident from these diagrams that the laser beam spot size is the critical element which determines the

1. track pitch (separation between adjacent rows of pits). There must be room for the spot of a tracking beam to fit entirely within the land area between tracks with room to spare. The tracking beams should not be centered between tracks because if they were, both beams would dim at the same rate as the tracking deviated and they crossed into adjacent tracks. Instead, as the tracking deviates one of the beams must start dimming as it begins to partially fall upon the track pits while the other beam is still at full reflectance.
   
   The track pitch is specified as 1.6 microns (1600nm).
   
   
2. minimum size of a pit or land. Conceptually, at least, there must be room for the spot of the reading beam to fit entirely between the leading and trailing edges of a pit, or in a land between pits. If the beam was crossing two edges at the same time, how would we be able to distinguish them? Therefore the pits and lands should not be smaller than the size of the laser spot.

In practice it's not quite this simple. At these small sizes, the laser does not make a simple "spot light". In the next section we will look at the actual spot size of the laser and how to compute it, and then return to the question of how it affects the sizes of these key features on the disc.