What makes Blu-Ray (TM) Blue?

In the last blog, I talked about the difference between continuous (analog) data and discrete (digital) data. One of the most popular, "hands-on", types of data that people use each day are that for audio and video.

In the case of audio, analog data are often considered to be the most "faithful" to the sound. Digital adherents say that the sound stays crisp and clear. They are actually both correct. When analog recordings are made they are able to reproduce all of the "between" sounds that are dropped during analog recordings. While one can debate as to whether it can be heard by most people, it does exist and, therefore, there may be a substantial difference even if only noticed by the subconscious.

Analog recordings primarily fall into two categories -- an engraved reproduction of the sound waves or a magnetic version. Each has the capability of continuous data recording. However, the use of such recordings requires destructive mechanical mechanisms to be "read" after being recorded. For the "engraved" (vinyl, records, wax cylinders -- yes all have been used) version this means a sharp object following the path of the engraving which will eventually start cause eroding the engraving and a deterioration of the sound. For magnetic versions, the media (tape usually) wears while being pulled and the magnetic material on the tape also gets worn by friction with the reading "head".

Thus, as time goes on and the recording gets used, the recording will get worse -- while, in general, a digital recording will stay the same. So the audiophiles and the digital adherents are both "right".

CHALLENGE: It should be possible to create a commercially viable analog recording medium that can be read non-destructively. With all the bright people and companies employing bright people this should be possible. Make it so!

As discussed in the previous blog, digital media (for audio and video especially) requires decisions as to how much data will be omitted. This is precision and sampling rate. For human speech, it is considered acceptable to take a sample 8000 times per second and the data can be recorded with the use of 8 binary units ("bits"). This means that a digital recording of human speed will require 480,000 bytes (8 bits), or 480 KB per minute of recording. In the case of "high fidelity" digital recordings of music, the sampling rate can be increased and the precision may also be increased. This ends up with a greater amount of data.

Currently, a popular way to record this data is on optical disks. The digital data are marked on the optical media with very, very small pits. A pit can be considered to be a "1" and a land (non-pit) can be considered to be a "0". Note that this is actually very similar to analog engravings except for the nature of the data. Please also note that the exact encoding is actually more complicated than I am saying -- check other sources for more precise descriptions.

A larger difference form analog data, however, is how the data are read once recorded. An optical disk makes use of a laser which can tell whether there is a pit or a land by the timing of the reflection from the medium. This reading is non-destructive and, as long as the optical medium is not otherwise damaged, should retain data unchanged for a long time.

We now enter into the third area of data recordings which is storage space. An audio CD makes use of a near-infrared laser (wavelength of 780 nm). This wavelength determines how dense the data can be placed on the optical medium. For an audio CD, using 780 nm wavelength lasers, about 737 MB (megabytes) of data can be stored in a single layer (a disk CAN have multiple layers with the laser reading separately from the different layers of the disk). Since this amount of data is considered to be around 80 minutes of music, we can see that, for an audio CD, each minute takes about 10 MB of data so the precision and sampling rate are much higher than considered acceptable for human speech -- greater "fidelity".

The wavelength of the laser determines density -- how "packed" the data can be. This limits the total amount of data in a predetermined physical size.  One method of increasing the density is by decreasing the wavelength of the laser.

A Blu-Ray disc uses a "blue" (blue is officially considered to be 475 nm) laser with a wavelength of 405 nm. A single layer Blu-Ray disk can store about 25 gigabyte (GB) of data. For audio, this would be about 100 hours using the same encoding as CDs. DVDs use a wavelength of about 650 nm ("true" red)..

Analog and digital data

I thought that I would talk about digital media -- CDs, DVDs, Blu-Ray, and so forth. But then I realized that I really needed to first talk about what digital media are -- and that, in turn, means that it is important to talk about analog.

Analog data are a reflection of events that occur on a continual basis. Such things include time, temperature, sound, moving images, water flowing, and so forth. An analog watch is known by its "face" -- where the "hands" are located to allow a person to interpret the data (information).

It would be completely possible for a watch to have a single hand. All the information is present in the hour hand. However, it is difficult to "read" (interpret) the value with a single hand and, therefore, analog watches and clocks normally have a minute hand and may even have a second hand.

In my old university, they had an analog computer. Set up correctly, it would be able to be used to calculate an exact value for Pi. But this brings up further the problems with analog data -- being able to actually make use of the data in a precise manner.An analog thermometer can give a precise value but can a person really read it that clearly? Is it saying 98.6 or saying 98.53?

Digital data can only create approximations of continual information. There are a lot of non-continual data in the world -- particularly in the area of finances. However, when it comes to continual data, you are involved with sampling rates and precision. The sampling rate is how often you "mark down" the information. You take a sample of sound at 1 hour, 20 minutes, 15 seconds, and 180 milliseconds. You then take a sample of the sound every 20 milliseconds -- but, whatever interval you choose, you are also choosing to ignore the data that exists when you are not taking a sample. You will never really know what happened within that 20 millisecond gap. You can guess what it might be -- that is called interpolation -- but you cannot know.

The second part of digital data is precision. For money (or other non-continual data), the precision is self-defined by what exists (although other units may exist for formulas -- like taxes). For continual data, the precision is a choice. Do you record 98.5, 98.54, 98.536, 98.5359, or what? Once again, you lose data/information and your choice CAN make a difference if the data are used in a repetitive fashion (such as calculating trajectories for a space ship).

So, analog data are accurate but very difficult to interpret precisely. Digital data are an approximation but have ease of interpretation as a built-in aspect of the choices that are made.

And this leads into the next blog "What makes Blu-Ray (TM) Blue?"