Frequently Asked Questions

What's The Deal With Audio Versus Data Disc Formats And All That Jazz?

Recently I received an email message asking me about this. I'm going to reprint it and my response in its entirety, as it will likely answer many of the most commonly-asked questions.

On Fri, 25 Jan 2002 20:05:20 -0600, James wrote:

>     I just noticed that my cd-r says 700 MB but 80 minutes. I have a cd 
>with almost 30 half hour shows on it. Is there something I need to know 
>about recording from my zip drive where I have download otr files to my 
>cd-r? How do I get several hours of shows on an 80 minute cd-r?

James, you definitely wanna do some more research on this. :-) But I'll try to give you the capsule version.

When CD's were first created, they were intended to be used in one way, and one way only: for music. As such, they had a maximum capacity of just slightly over 80 minutes, though such long-playing discs were rare (though I do have a few from the mid-80's that run that long, usually two-album sets combined into one CD).

When this standard was created, it made the following assumptions about the sound sampling:

44.1 KHz (44100 samples per second)
16 bits per sample (two bytes)
Then multiply all of this by two for stereo

This comes out to 172.27 kilobytes per second. So when you stick an audio disc into your CD player, this is the rate of data delivery that your CD player reads the data from the disc. It then transforms this data into audio, and delivers it to your speakers (or receiver, or however you've got it wired).

Now, at 172.27K per second, that adds up to an impressive total of data after 80 minutes. In fact, it adds up to 807.495 megabytes, which is a LOT.

Then they came up with the brainstorm of: "what if we made a data version of CD's? We could put a CD unit into a personal computer, and then make CD's filled with lots and lots of data, which could be accessed just like any other file on a computer, except that you could only read, and not write." And that's how the CD-ROM was born.

In doing this, though, they realized that CD's are easily damaged (scratches, fingerprints, etc.) and that the actual pits and grooves on the disc were SO small that even a speck of dust could cause a read-error that would stop you from correctly reading a file. Unlike an audio disc, where that same read-error is a very very small part of the overall music "picture" and thus the D/A converter could simply "smooth over" that part of the music, invisibly to you, ANY non-recoverable error on a CD would spell disaster for that particular file. So, they came up with a scheme that creates a fair amount of redundancy in the data, which means that if a particular sector of data becomes unreadable, but the surrounding sectors ARE readable, the system can "reconstruct" the missing data accurately. This actually works great, at least to a point - once you reach a certain amount of damage, though, even this scheme fails - but over the long haul, this scheme has worked spectacularly well.

It comes at a cost, though. The "redundancy" means that space on the disc that could otherwise be used for other data needs to be used for "redundancy" data. Thus, the figure of 807.495 megabytes that we arrived at earlier is not possible for data discs. The actual figure is about 702+ megabytes, which is still one heck of a lot of data.

Anyways, point is, we now have two completely different ways of using CD discs, both of which are NOT compatible with each other because they use the available storage space in entirely different ways. Each has its own way of finding files or "tracks" on the disc, and of storing the data itself. Most CD-ROM units in personal computers are capable of using either type of disc (data or audio) but you do need to understand in your mind that they are DIFFERENT standards - it just happens that your CD-ROM player can switch modes and understand either standard. Your home CD player, on the other hand, cannot: if you stick your Windows 98 disc in it, it won't play any music. :-)

If you understand this point - that there are two different FORMATS of disc (audio and data), we can move on. If you're still confused here, stop and ask me or do some online research.

We're going to put the discussion of disc formats aside for now, and talk about something else: file formats on your PC.

Audio files can exist in a very, very broad variety of file formats. Only one of these file formats is similar to anything we've discussed, and that is the ".WAV" format. Even then, .WAV allows for a wide variety of bitrates, sample rates, frequencies, and so on, and only one combination of these values, 44.1KHz, 16-bit, stereo, is compatible with the "audio disc format" mentioned earlier.

Typically, if you have a .WAV file on your hard drive (and if you're running Windows you have several, starting with c:\windows\media), it is UNCOMPRESSED. That is, the sound was converted from analog to digital at a particular frequency and sample rate and then stored as-is. This means that the higher the quality of the sample, the higher the amount of disc space required to store it.

Now, remember the audio disc format? For "CD quality" sound, that means that for every one second of sound you record, you need 172.27 kilobytes of disc storage. That adds up in a large hurry: a 5-minute song will consume 50.47 MEGAbytes of space on your disc. If you then, say, wanted to send that file to a friend via email.... well, most ISP's won't let you send files that large. It would take a hell of a long time, too. Bear in mind also that we've only recently entered the age where a 40-gigabyte hard drive is standard equipment on new PC's - my first PC hard drive had a grand total of 30 megabytes of storage, which is less than the size of this single hypothetical 5-minute .WAV file that we've just created!

So along came some folks in Europe who dreamed up something that has come to be known as "MP3". They did a lot of research and found that they could "compress" audio dramatically. I'm not sure precisely how it works, but basically you run this .WAV file through an MP3 converter, and it will essentially convert the sound into a series of mathematical equations. Much of the actual audio data will be THROWN AWAY, because they determined that the human ear actually couldn't tell the difference if that data were present or not. The "important" data was kept.

Then, later, you can "play" this MP3 file using an MP3 player program, such as WinAMP or Windows Media Player. This player actually has a lot of work to do. When playing a .WAV file, all a player has to do is convert a digital data stream back into analog, which is comparatively trivial. But playing back an MP3 file means making thousands of complex computations per second, in order to reconvert the mathematical representations of audio contained in the file back into something that sounds like the original sound. Actually this happens in two stages: first the mathematical representations get converted into a digital stream, and then that digital stream is sent through the same digital-to-analog conversion process that a regular .WAV file would go through. Therefore, playing back a CD-quality MP3 file requires a computer with a reasonable amount of horsepower: at least a Pentium 200 or thereabouts.

MP3 is what's known as "lossy" compression. You may also be familiar with .ZIP files, which are "lossless" compression. Anything you put into a .ZIP file will come back out of the .ZIP file exactly the same as it went in. Anything you put into an MP3 file will NOT be the same, it will have lost something. The genius here, though, is that with MP3 and a high enough bitrate, the human ear generally cannot tell the difference (but you could see if it you compared the original versus the MP3 on an oscilliscope).

A "CD quality" MP3 file will be around 11 to 12 times SMALLER than the original .WAV file. So in our example above, that 50.47 megabyte file is now (dividing by 11) only 4.59 megabytes, and that's why Napster got into so much trouble. At this size, sending full-length CD-quality songs to your friends suddenly became not only possible, but a very attractive idea.

MP3, like .WAV, also allows for a variety of options. If you're willing to sacrifice some quality, you can make the resulting files even smaller. And when you're dealing with Old Time Radio files, the original files were ALREADY of low quality to begin with.

A CD-quality MP3 file will usually have a bitrate of 128kbps. A typical OTR file will have a bitrate of 32kbps, meaning it will be four times smaller. You'll have to sacrifice something to make this work, and usually this tradeoff goes something like this:

  1. First, you drop stereo. Few OTR files were in stereo to begin with (and none at all from the "Golden Age Of Radio", only some of the newer things were done in stereo, like Hitchhiker's Guide and stuff like that). So, now, you no longer need 128kbps: 64kbps works just fine.
  2. Next, you drop your frequency response. At 128kbps, you're getting the same 44.1KHz frequency response that you get with a CD. But if you decide to drop down to 22.05KHz, which is what most FM radio stations use, now you can drop your bitrate to 32kbps and get NEARLY the same level of quality.

As I said, MP3 files at "CD-quality" are 11-12 times smaller than their original .WAV counterparts (assuming the .WAV is using the CD audio format). If you add in these two additional assumptions, you can now get a reduction of 44-48 times smaller than the original. Put it another way: these 32kbps MP3 files are 44-48 times more efficient in the use of disc (or CD) space than the CD "audio format" is.

In practice, the people that encode OTR shows to MP3 don't always follow this "accepted standard", and will do other things. They might reduce the bitrate to 24kbps, or even 16kbps, or even (in a few cases) 11kbps. The more you reduce the bitrate, the more you reduce the quality. The sound becomes harsh and metallic, and you can begin to hear "digital artifacts" in the audio. At each of these bitrate settings, there are different frequencies that can be used as well - you wind up trading a higher frequency-response for slightly-higher amounts of digital artifacts. Generally, 32-22-mono is the best way to go for OTR, but it's hardly the only kind of file you'll ever see.

OK, I've spent a lot of time explaining just what an MP3 is, but I now need to wrap this back into your original question. An MP3 file, as cool as it is, is just that: a file. It's a way of storing data that, if played back using a specialized player, will produce some audio. Until recently, the only way to play them back was on your computer, using an MP3 playback program. Nowadays, you can buy MP3 players. These include:

  1. "Solid-state" players. These are usually very small little units with 32 or 64 megabytes of RAM in them, with no moving parts. You connect them to your computer so that you can "load" them with an assortment of MP3 files. You then disconnect it, connect a pair of headphones, stuff it in your shirt pocket, and go jogging. :-) At CD quality, a 64 meg player can give you roughly one hour's worth of music. At "OTR quality", call it about 4 hours.
  2. "CD/MP3" players. These are just like a regular portable CD player (ie: discman), but can also play discs loaded with MP3 files (see upcoming discussion) if that's what you put in it.
  3. A growing number of home CD units, and especially DVD players, are now also able to play CD's that have MP3 files on them. I have a 5-disc DVD player in my living room that I can get about 60 hours of playback with before I have to change discs (at CD quality).

So what is an MP3 disc? Well, remember our discussion of CD formats at the beginning? There is "audio format" and "data format". "Audio format", despite all these advances, is still only capable of storing and playing audio in the one single datarate that was originally conceived, and that limits you to a maximum of 80 minutes per disc. "Data format" discs, however, can accept any type of file - you can put your spreadsheets on it, or your Word documents, or whatever you like. You can also burn a disc filled with nothing but MP3 files.

If you place a disc that contains MP3 files into one of the "CD/MP3" type of players listed in #2 and #3 above, the player will sort through them (discarding anything that isn't an MP3 file) and allow you to play them back. They have onboard circuitry capable of decoding the MP3 and turning it into audio. The disc is still a "data format" disc, but these more-advanced players are capable of dealing with them. A standard CD player still CANNOT deal with data-format discs, only audio-format.

When you order a customized CD-ROM from the Old Time Radio Vault, what you are getting is a data-format disc loaded to the brim with MP3 files (as close to 701 megabytes as I can manage to shoehorn onto it). But it is still a data-format disc, meaning your CD player can't play it. Your computer can, providing it has an MP3 player installed, and so can CD/MP3 players. You can also put it into your computer and then push the files from the disc into your "solid-state" MP3 player.

Or, you could even go so far as to "convert" the MP3 files back into CD-quality .WAV files and burn them onto discs in "audio-format". Again, though, you are LIMITED to 80 minutes per disc. Essentially, you are WASTING a lot of space, but if it's your only option, it will certainly work.

The most common CD-burning software (it comes with most burners) is something called "Easy CD Creator", by Roxio. It will burn your data-format discs for you (it's what I use to make the customized CD-ROM discs). It will also create audio-format discs, and it's really cool because it does not CARE what format the original source files are in. They can be MP3, or .WAV, or whatever, and they can be at any bitrate, frequency rate, or whatever. Easy CD Creator, when burning audio-format discs, will CONVERT your source files into the correct audio format (and thus limiting you to 80 minutes) before burning them onto your new audio-format CD.

Aside: Easy CD Creator also comes with something called "DirectCD", which allows you to preformat a disc and then "drag and drop" files to it. This can be very convenient, but you need to know that this, too, is a "data-format" and not an "audio-format", and thus these discs won't play in a CD player. *SOME*, but not all, of the CD/MP3 and DVD players capable of MP3 playback will recognize this format, but some will not. The documentation for your player will tell you if it will. (The AVC Soul Player, aka the RioVolt, will do so: my 5-disc DVD changer and my car CD/MP3 player will not).

One last thing you need to know before you merrily begin burning discs. There are two types of recordable media. "CD-R" and "CD-RW". The first means "Compact Disc Recordable" and the second means "Compact Disc ReWriteable". The difference is that:

  1. You can write data (or audio) to a CD-R only once.
  2. You can write data (or audio) to a CD-RW many times - you simply "erase" the disc and then write to it again. Also, there are now two different kinds of CD-RW discs: low speed and high speed, which can make a difference to you on playback (see next section).

The material that makes up the recording surface of these discs is markedly different in both cases, and you need to be aware of the following:

OK, nearly finished. Two other things to mention.

I've mentioned "low-speed" and "high-speed" CD-RW discs. To explain that, I need to back up a bit and talk about speed ratings.

As mentioned right off the top, a CD player needs to be able to read data from an audio-format disc at 172.27 kilobytes per second. Similarly, a CD-ROM unit in your PC running at the same rotation and playback speed will read a data-format disc at 150 kilobytes per second (after accounting for redundancy data).

Call this speed, 150KB/sec, the "baseline" speed.

Now, whenever you buy a CD-ROM or CD burner, you will notice that they always advertise the speed of the unit in terms of "x". For example, the most recent CD-ROM unit I purchased is an LG "52X MAX" drive.

The "52X" part of the name means that it is capable (at best, under ideal conditions) of reading data at a speed 52 times greater than the baseline speed. Think of "x" as the multiplication sign in a simple math equation. 52 x 150KB/sec = 7800 kilobytes per second, which is screamin' fast.

When you buy a CD burner, though, it gets a bit more complicated. Instead of getting a single "x" number, you'll get three, and sometimes they won't even give you the "x" part, they'll just assume you know that's what they mean.

My first burner was "2x/2x/6x" and it was painfully slow. These numbers mean that:

  1. It could write to a CD-R disc at 2 times the baseline rate (so 300KB/sec)
  2. It could write to a CD-RW disc at 2 times the baseline rate (so 300KB/sec)
  3. It could read data at 6 times the baseline rate (so 900KB/sec)

My next burner was a "6x/4x/24x" unit, and thus similarly:

  1. It could write to a CD-R disc at 6 times the baseline rate (so 900KB/sec)
  2. It could write to a CD-RW disc at 4 times the baseline rate (so 600KB/sec)
  3. It could read data at 24 times the baseline rate (so 3600KB/sec)

My current burner is something called a "PlexWriter 24/10/40A" (you can disregard the "A" part, it means "ATAPI" and describes the way it connects to the computer - it contrasts with "S" which stands for "SCSI" and is a different interface). Note they no longer bother with the "x" but the numbers still come in the same order, thus:

  1. It can write to a CD-R disc at 24 times the baseline rate (so 3600KB/sec)
  2. It can write to a CD-RW disc at 10 times the baseline rate (so 1500KB/sec)
  3. It can read data at 40 times the baseline rate (so 6000KB/sec)

Hopefully that's clear. What you've probably noticed is that writing to CD-RW discs is slower than writing to CD-R discs, and this has to do with the way the recording takes place and differences in the recording material itself. CD-RW disc burning was limited, for a long time, to 4x. Recent advances have allowed the jump to 8x and even 10x, but this meant going to yet another different type of media itself. When you buy CD-RW discs, the front of the package might say something like "1x-4x", which means this is a low-speed disc designed for use with drives capable of burning up to 4x. Newer CD-RW discs might say "4x-10x" or sometimes "8x-10x" or even sometimes only "10x". Some of them include this warning (and ALL of them SHOULD): use only with newer high-speed disc recorders. Using these discs with the older, slower CD-RW burners won't work: the media is incompatible. On the other hand, you can safely use the low-speed discs in high-speed drives.

And that about wraps up this discussion. I hope I've been of some help to you.