Lasers! Rainbows! The eighties! Drum Machines! Digital! Big hair! This is some righteous stuff! No kidding, the compact disc was a radical
departure from how, well really how a lot of stuff worked. You could argue that the CD, with its vast
data capacity, relatively robust nature, and with the further developments it spurred along,
changed how the world did virtually all media. That is at least until physical media became
the seemingly undesirable thing it is today, a time when streaming services and libraries
on hard drives are all the rage. Unless of course it’s vinyl, then by all
means please build your collections. Wow the snark’s coming early today. Well, this is the third video in a series
on digital sound. You can find a playlist to the previous two
videos, in which we covered how digital sound works (as well as the Nyquist-Shannon sampling
theorem), but for now, sit back and relax as you feast your eyes on the silver platter
that is the compact disc. No discussion of the Compact Disc is a good
discussion unless it pays homage to its predecessor, the LaserDisc. Now, I’ve done a series on Laserdisc if
you’d like to learn more, but in brief, this was the very first commercial optical
storage format. First released in test markets in 1978, Laserdiscs
were usually 12 inch or 30 centimeter discs, and they were an analog video format, holding
up to one hour of video per side. Competing in a world where the videocassette
recorder which could record from live TV already existed, the Laserdisc failed to capture the
hearts and minds of many individuals even though it had steller video quality. Throughout its life, it stayed a videophile-only
format in most markets. But, the development of the Laserdisc, which
was done in large part by Philips, presented an obvious solution to the problem of digital
sound storage. See, although Laserdisc is an analog format,
the signals encoded on it are as a series of pits and lands. OK, optical disc fundamentals time! You might already be aware of this, but optical
discs are read by shining a laser up at a reflective disc that is covered with little
pits. These pits are roughly one quarter as deep
as the wavelength of laser light that will hit them. I ran into some inconsistencies regarding
the exact depth of the pits, because the only source I found which specifies it has the
wrong wavelength of light listed for the CD, but all you need to know is that when the
focused laser hits a pit, the increased depth causes the reflected light to destructively
interfere with the projected light, which reduces the overall intensity of the light
reflected back. That’s actually a pretty neat part that often
gets overlooked. Endless articles talk about the fact that
the pits change how the light is reflected, but very few mention the destructive interference
aspect of it. I’ve shown this diagram before, and it perfectly
demonstrates what happens. When the laser hits a not-pit, the light gets
reflected right back down to the laser, and a prism reflects some of this into the photocell. But when it hits a pit, the destructive interference
greatly reduces the intensity of the reflected light, so very little light hits the photocell. This is how a laser pickup system can tell
the difference between a pit and a land. In the LaserDisc system, these pits and lands
were used to encode analog video and audio signals via a weird hybrid of pulse-width
modulation and frequency modulation-- Don't ask, it’s complicated --and the result is
a usable analog video signal from a shiny plastic disc. So, having already invented a thing that used
lasers to read information on a disc, when it came time for a digital music format to
hit the scene, Philips was poised to knock it out of the park with a new disc, this one
no less lasery, but quite a bit more compact. They named the format along the lines of their
previous compact invention, the compact cassette, and in a non-coincidence, the diameter of
the CD is roughly the same as the diagonal length of the Compact Cassette. Ah, but let’s not forget Sony’s role. Sony, the people who seem to only make either
runaway successes or disastrous failures, had been working on digital audio for some
time. They were the ones that developed a PCM adapter
for use with U-Matic videocassette recorders as discussed in the previous video, and in
fact they were working on an digital optical audio disc before Philips released the Laserdisc. One of their early prototypes from 1977 was
the same size as one of these honkers and held only an hour of digital audio, though
at the same exact quality of the soon-to-be Compact Disc-Digital Audio standard. Philips and Sony were sort of working on the
same thing at the same time, though Philips had the notable advantage of having already
developed and manufactured the Laserdisc by the time things really heated up. Still, Sony contributed a lot. I don’t want to get too into the weeds of
who did what, so let’s just jump to 1979, the year that Sony and Philips first decided
to for realsies collaborate on the project. By this time, optical disc fundamentals had
been established. An optical disc seemed the perfect format
for digital data, because you could just easily call a pit a one and a land a zero. But that’s not how it works! No, the pits and lands are important, but
it is the change between a pit and a land that encodes a 1, and no change encodes a
zero. This is time-based, so a four-bit span of
time which is comprised of either all pits or all lands will produce 4 zero bits. Start the sequence with a change and then
continue as is, and you get 1000. No change, change, no change, no change, and
you get 0100. And you get the idea. This is called Non-Return-To-Zero inverted
encoding. Now, this has limitations because a long span
of zeros requires an outside clock to keep track of how many “dead” spaces have passed. The longer you have between ones, or changes,
the more ambiguous the number of zeroes becomes. I know what you’re thinking. Was that six zeros or only five? Well, to tell you the truth in all this excitement
I kinda lost track myself. Which is why you need an outside timing source
keeping track of how many shots, I mean bits, have passed. This isn’t somewhere you should just feel
lucky. But that’s not a problem. What is a problem is the nature of everything. How likely do you think it is that a pressed
CD is absolutely perfect, has no scratches at all, and that the CD player reading the
disc will perfectly, without any errors, reproduce the correct sequence of ones and zeros? If you answered “not likely at all” you’re
a winner! To help make the data less susceptible to
stupid little things, two robust error-fighting mechanisms are built into the data stream. The Red-Book standard, the first of the Rainbow
Books that defines the physical, digital, and other various standards surrounding each
type of Compact Disc, not only specifies the sample rate and bit depth of the Compact Disc
Digital Audio format, which as previously discussed is 44.1 kilohertz, 16 bit, but it
also defines how the data is encoded on the disc. Now this is rather technical but I think interesting,
so bear with me. So, each audio sample is a signed, which means
it can be positive or negative, 16 bit two’s complement, which is a mathematical operation on binary numbers that we’re just gonna not worry about right now because it hurts my brain, integer. I mean, we knew we’re dealing with 16 bit
audio so each sample is gonna be 16 bits. Now, on the disc, 12 samples, 6 each for the
left and right channels, are stored together in a frame of 192 bits (that’s 24 bytes,
for those playing along at home). These frames are then encoded using an error
correction scheme called CIRC, which stands for Cross-interleaved Reed-Solomon coding. In addition to adding one data parity byte
to every three raw data bytes, the effect of CIRC is basically to spread the data out
over a longer distance. That’s where the interleaving part of the
name comes from. By jumbling up multiple frames and adding
parity bytes, CIRC can correct up to 3,500 bits of error-filled or even missing data,
and can compensate for up to 12,000 bits by masking errors via interpolation. This translates to up to 2.4mm gaps in data
being completely corrected for, and up to an 8.5mm scrambling of data, either through
a scratch or some other damage, being reasonably interpolated and masked. The end result is that the error is either
completely corrected, or it’s fudged well enough that you won’t hear it. Now we’re gonna go on a minor tangent here. I apologize. I’ve linked in the description the source
material from the Wikipedia article on CIRC. This source material is from a book written
by Kees Schouhamer Immink, one of the central engineers involved in the development of the
Compact Disc. He even won an Emmy for his work on coding
technology for optical recording formats. I’m bringing this up because for every person
that tells you Wikipedia is a worthless starting point for research, I want you to show them
the references section at the bottom of articles. Yes, you would be very unwise to cite “Wikipedia”
in any research, but let me tell you, you may be amazed at the quality of the source
material, and you’d be a fool to not at least look at these references when doing
any preliminary research of your own. Rant over. In addition to CIRC, the use of eight-to-fourteen
modulation limits the total number of possible combinations of ones and zeros. EFM translates each 8 bit word into a 14 bit
word. These 14 bit words are translated back into
their original 8 bit words using a lookup table. The point of doing this is to reduce the possibility
of errors. The use of EFM makes it such that binary ones
are always separated by at least two zeros, and a maximum of 10 zeros. This means that every pit and land is at least
3 clock cycles long. It also means that if the CD player reads
one, zero, one, it knows it must have made an error because that’s not a possible sequence. Plus, with a maximum of 10 zeroes, it reduces
the accuracy required in the clock of the CD player for worst-case clock recovery, as
there must be a one after every eleventh bit, and this will in effect synchronize and restart
the zero-counting clock. Keep in mind that with simple 8 bit encoding,
in theory you could have 14 zeros in a row. When you realize all of the processing that
has to be done by the CD player before it can even extract the data it needs to send
to the DAC and play the dang music, you gain a newfound respect for its achievements of
the time. Granted, using a look-up table and performing
some basic arithmetic is easy for a computer, but the fact that this was being done at a
consumer hardware level in 1982, with a data throughput faster than what any contemporary
microcomputer would reasonably be expected to process, impresses me. To help provide logical access to the contents
of the disc, the disc’s data stream is divided into three parts. The lead-in, the program area, and lead-out. The lead-in contains the disc’s Table of
Contents, which is basically an announcement to the player of how long it is, how many
tracks it contains, and what the timecode is for each of the tracks. The CD player can, through reading the table
of contents, determine where each track is for accessing it nearly instantly. A CD can have up to 99 tracks, which themselves
can be divided further into 100 indexes, though this feature was rarely ever used and few
CD players could access the index information. One thing that computers have made a little
confusing about the Compact Disc-Digital Audio standard is that audio CDs do not contain
files. If you pop one into a PC, it’s gonna show
you each track as its own object. But the computer is interpreting that for
you. In raw form, the data on a CD is just one
continuous stream, and the table-of-contents in the lead-in simply defines where along
the stream each track is. The concept of using a CD for computer data
storage just wasn’t really in the cards yet. I mean, a single audio track would easily
fill up entire hard drives of the time, so the idea of creating files to contain the
audio data was just absurd. The later CD-ROM, following the Yellow-book
standard, would allow for file structures on discs like a computer is used to accessing. But we’ll get to that later. OK, so let’s look a little bit closer at
the disc itself. First, I want you to see if you can spot a
big difference between a Laserdisc and a CD. Looking at the edge of a Laserdisc, can you
see that seam in the middle? Laserdiscs are obviously double sided, so
that seam is the join between the two halves of the disc, and the data is sandwiched between
two sides. But if you look at the CD, there is no seam. One of the weirder things about the CD is
that the data layer is actually at the top of the disc, just beneath the label. I’ve linked to a How It’s Made episode
on the Compact Disc that does a pretty good job of demonstrating how mass-produced discs
are made. Their manufacture is surprisingly similar
to that of conventional vinyl records, as they are stamped (or rather molded) from a
master disc. The video does a great job of showing this
process, but it skips over the actual etching of the data onto the glass master. In short, the master is covered in a solution
that will either evaporate or harden when exposed to laser light, depending on the mastering
process used. By using the laser to etch pits into the coating,
which is then hardened with a development process, the glass master is now covered in
bumps that will represent the lands in the molded discs. The master disc is metallized to harden these
bumps, and now polycarbonate discs are molded from it. And that’s something that I find really
neat about the Compact Disc. The data in stamped discs is actually in the
plastic itself. When the disc comes out of the mold, it’s
completely transparent. But it has all of the data on its surface. To make the disc readable, it’s then covered
in a thin film of aluminum via a vapor metallization process, and now the disc could theoretically
be read by a CD player. But remember, those pits and lands are on
the top of the disc, exposed. If they could be touched, the data would be
destroyed. So before the disc can be handled, it needs
a thin coat of varnish to be spread along the top, and now the data is protected from
damage. Putting the data layer on the top of the disc
meant that the disc was even more tolerant of scratches, as these scratches are out of
focus to the laser reading the disc from below. This was a pretty smart move. The laser would read through almost the entire
1.2mm thickness of the CD. As a consequence, double-sided CDs were never
a legitimate thing. But there were some weird shenanigans tried
with making double sided hybrid discs. One of these I have somewhere in my collection
but I couldn’t find it because I’m disorganized, (sorry) is the DualDisc. “Weird Al” Yankovic’s twelfth studio
album, Straight Outta Lynwood, was released in 2006 as a DualDisc, with one side being
a fully compliant DVD containing music videos and other goodies, and the other side being
a… almost CD. DualDiscs were .3mm thicker than a standard
CD or DVD, comprising a total thickness of 1.5mm rather than the standard 1.2. The CD layer was placed .9 mm into the disc,
with the DVD layer .6 into the disc from the other side, which is correct for a DVD. But it’s off by at least .2mm according
to the Red Book standard. Because of this, DualDiscs did not contain
the Compact Disc Digital Audio logo on them because they technically weren’t CDs. They featured language saying that they were
intended to work in standard CD players, and almost certainly they would given the tolerances
a standard CD player is designed to deal with, but they could not actually be officially
referred to as a Compact Disc. So there’s some fun trivia for you. AAAND, that’s where we’ll hit pause. There’s a lot of neat stuff to uncover about
the CD, even though it’s teetering on the edge of obsolete and insignificant. OK, it is pretty much obsolete. This dime-sized SD card holds more data than
this entire column of CDs. That’s kinda sad, but also amazing. In my next video on this subject, we’ll
explore more of the innards of the compact disc, such as the optical pickups mechanisms
and the rather major difference between how Philips designed its laser pickup compared
to most others. If you take a look at the disc tray on this
Magnavox unit, you might get a hint. Of course we’ll also touch on the other
various uses of the CD, such as CD-ROM, CD-R and CD-RW, and other stuff. And then, we’ll relish in the fact that
everything is on the Internet now. Thanks for watching, I hope you enjoyed the
video! If this is your first time coming across the
channel and you liked what you saw, please consider subscribing! As always, thank you to everyone who supports
this channel on Patreon, especially the fine folks that have been scrolling up your screen. If you’re interested in making a contribution
to the channel to help it grow, please check out my Patreon page. Thanks for your consideration! Don’t forget there’s a Technology Connections
subreddit now, so if you’re a redditer you can go over there and watch me not know how
to use it, and I’m getting better a Twitter so if that’s something you do, you can follow
me @TechConnectify. Anyway, that’s it for now. I’ll see you next time! Cue obnoxious music! ♫ uncomfortably smooth jazz ♫ This line read better before… With its vast data capacity, relatively lobrust… this line. This line is going to kill me! When it came time for digital music formats
to hit the scene, Philips was poised to knock it out of the par--poised? (stares into camera doubting himself) No, that’s what I wrote. That’s what I wrote! I’m regretting it now… The Red Book Standard, the first of the rainbow
books that define the physical, dizhidal, and other… dizhidal dizidal diblelr…. (clears throat) ...reduces the accuracy required in the clock
of the CD player for weush… [exasperated sigh] The master disc is metAL…. Metalized.
