OTR Server Technology Page


The purpose of this page is to cover some of the technologies, both hardware and software, that it used to put these Old Time Radio programs on the World Wide Web.

As this site is a service of the Technology Education Department at Billings Senior High, I feel that it is a good idea to offer a small "course" in how various technologies, old and new, are used together to do something that in the early 1990s was unthinkable.

This page is version 1.0, January 1, 1999



Table of Contents

The Path from Airwaves to Binary

Collecting the Original Source Material
Keeping a Catalog
Stepping into Digital

World of Digital Sound

Wiggley Grooves to Zeroes and Ones
The Conversion Process
Sound File Sizes
Sound File Formats
Software for Creating Sounds
Wrapping It Up

Getting it on the Web

Why Move to Digital?
Hardware and Software
Getting the Shows Online
Going on the Air

Where Do We Go From Here




The Path from Airwaves to Binary

To study the path of sound as it travels from an analog signal delivered through the airwaves to a digital form downloadable from the Internet, is to study the history of technology itself. The story is full of famous and infamous inventors, accidental discoveries, and the claiming of wartime booty.

The entire history of this subject is to be addressed elsewhere on these pages, but here we want to look at how I have used these technologies to deliver the content of radio broadcasting from over 50 years ago to the desktop of today's computer users.

Collecting the Original Source Material

I have been collection Old Time Radio shows since the early 1970s. My primary source has been to tape rebroadcasts of programs from the radio. I did most of this in Los Angeles from 1971 to 1991. Some recording was made from AM radio, KRLA and KNX, but most was from FM stations such as KPCC, KCRW, and KCSN. Some shows were taped from phonograph records that I either purchased or borrowed from the public library.

The quality of the individual shows varies. Those recorded from AM suffer the usual problems of static and other noise. Generally, these shows have also been edited by the producers, usually to fit in contemporary commercials. Programs from FM stations, which make up the bulk of my collection, are in good quality as they were played by other collectors over non-commercial stations.

Akai 202I use 4-track reel-to-reel machines to make the off-the-air recordings. I have used quite a few machines over the years, though my real workhorses have been an Akai 202D-SS and a Tandberg TD20A. The Tandberg has 10-inch reels and was used to pull several hours at a time off the air. I have since retired the Akai, replacing it with another Akai, this one a GX-260D. I use no noise reduction or equalization in making the recordings.

TandbergMy usual procedure was to capture the shows on the weekends, when most of them were played, and then transfer them to my archive tapes during the week. I have used a variety a tapes, for a long time favoring 3600-foot tapes on 7-inch reels. This was because of cost and availability at the time. In retrospect, this might not have been a good idea due to issues on tape life, but what's done is done.

Akai 260The first dozen or so reels were recorded in 4-track mono, which puts one program on the left channel and one on the right. While this certainly maximizes the cost of tape, there is a problem that can occur called crosstalk. This is where the sound on one track can be heard slightly in the other. While I can retrieve shows from the tapes that I made that way with no crosstalk, I became pretty paranoid about it and switched to recording in stereo, placing the same program in both the left and right channels.

In the early 70s, when my budget was non-existent, I recorded at 1-7/8 IPS. The results were listenable, but not worth keeping. I have since used 3-3/4 IPS on all my recordings.

In January 1988 I made the decision to keep my final copy of shows on cassette instead of reel-to-reel. This was because the cost of reel vs. cassette made the smaller format more attractive and I also believe I will be able to find cassette players longer than I will reel-to-reel machines. I usually use TDK DC90 cassettes. I have had good service from these tapes and, while these are a "standard" type, they provide more than adequate sound quality. Because the original source material has a limited audio range, and the fact that I am a collector, not an archivist, I could not hear a substantial difference between the "standard" and "premium" tapes.

Keeping a Catalog

Part of collecting anything is keeping a list of what you have so that you know what you have and where to find it. During the first couple of years of collecting, index cards seemed to work well, however, once I had over 100 shows, I began thinking of ways to make this job easier. It was about 1978 and I had just purchased an Apple ][ computer. One of my first projects was to write a database program, which I did. It worked very well and was, if I do say so myself, a pretty slick package. It stored the show name, the episode title, the broadcast date, and the reel location. It could sort on any field and print in a variety of reporting styles. It also featured an Old Radio Trivia Quiz, which has since been resurrected on these web pages. To try out the quiz, click here.

When I switched to the DOS world in 1986, I knew that it was time to upgrade my database as well. Rather than port my Apple ][ software to the IBM world, I picked up a copy of dBaseIII, which is very powerful and certainly did everything that I needed. I decided to expand the data that I collected on each show, providing fields for the cast, the director, writers, sponsor, network, etc. The big task is to listen to each of the these shows, collect the data, and enter it into the database, a project that may never be completed.

Along comes Windows, more powerful computers, and the Internet. Rather than continue to mess with dBase, I moved the whole database in MS Excel. It works great, though I'll have to come up with something else when I hit Excel's limit of 16,000 and some entries. Since I use Excel for other applications, such as a gradebook and for financial tasks, I am very familiar with how to use it and have become adept at generating web pages with it as well.

Stepping into Digital

My first venture into moving OTR into a digital format was before we had general access to the Internet and the World Wide Web was in still in its infancy. While learning HyperCard on a Macintosh LC II, I was exploring the possibility of using that application to teach a small unit in the History of Broadcasting. I scanned images of radio stars from books that I found at the public library and used the sound recorder, built-in to HyperCard, to digitize samples of programs from my collection by connecting the Line Output of a reel-to-reel deck to the audio input on the Mac. The stack introduced the user to some of the better known programs, showing a different program on each card, a text field with a description of the show, and buttons that allowed the user to play the sample or to advance to the next card. Little did I know then that I would be using a similar layout on many of my web pages.

I eventually moved from HyperCard to PowerPoint, seeing that a presentation program would meet my needs better than HyperCard would. I built the PowerPoint file using more images and sounds than I had before. I recorded the sounds by connecting the Line Out of my reel-to-reel to the Line In on the sound card of my computer. The computer is a 486-66 running Windows 3.1. I used the sound recording software than came with the soundcard. The entire PowerPoint file is about 18 megabytes in size. To make it more portable, I moved it, via backup tape, to my web server. From there I used FTP to move it to a Macintosh computer that has a CD-ROM recorder. I burned my first CD with that PowerPoint file on it. I use this PowerPoint file in a oral presentation that I make in my Technology Education class from time to time. I also use it when presenting to our senior Media Analysis classes where I am a "guest speaker" several times a year.

World of Digital Sound

This portion of the page is reprinted from my article that appeared in the MCCE News, December 1998.

Most of us are accustomed to and downright expect our computers to make sounds these days. Gone are the times when our machines made the occasional "system beep" to prompt us or to alert us to an error. Now we can hear full surround-sound effects emanating from our desktops as we surf the web or do our word processing.

The path into multimedia has created almost as many sound file formats as we have graphics file formats. Each has its advantages and disadvantages and in this column I would like to clear the static, as it were, and look at some of the more popular and useful types of sounds that we can listen to and create on our own.

Wiggley Grooves to Zeroes and Ones

Ever since Thomas Edison invented the first sound recording device, until the advent of the digital world, sound was recorded using the analog method. The first of these devices used a vibrating needle, or stylus, to cut a groove into a soft material, the groove being "analogous" to the shape of the original sound. Analog sound has also been recorded using an optical method on motion picture film and on wire or tape using magnetics. The primary disadvantages to analog methods is that the media used can easily be damaged and imperfections in the playback are common, such as pops from a record or hiss from tape.

Once computers became available, and less expensive, the world of digital sound was born. While sound still starts as an analog signal, it is converted into digital data using a DAC, which is an analog to digital/digital to analog converter, a piece of hardware embedded into the chips of the computer’s sound card. A small understanding of the conversion process will go a long way to improving the quality of sounds that you might make on your own computer.

The Conversion Process

As an analog sound plays, it appears like waves rolling in from the ocean, rising and falling with the amplitude of the original sound, each wave having its own height depending on the loudness of the sound. The frequency at which these waves approach us is also variable and represents the pitch of the sound. The challenge for the computer is to convert these continuous waves into digital data, that is, zeroes and ones. It does this by looking at the sound wave for an instant and recording what it sees as a piece of numerical data. The rate at which it looks at and records the sound image is referred to as the "sample rate" and can be from 11,000 times a second (11kHz) to 44,000 times a second (44 kHz). Other sample rates are also used.

The more often the computer samples the sound, the more accurate a "picture" it will get of the original. As you might guess, going from analog to digital is a "lossy" operation, as some of the original data is lost because it is not sampled, however, the idea is that the human ear cannot detect the missing data if the sample rate is high enough.

One more factor that can affect the quality of the digitized sound is the sample depth, which can be 8-, 16-, 24-, or 32-bit. As the computer samples the sound, it needs to record a number that represents what it heard at that instant. The bit size determines the range of possible numbers that the computer could record: with 8-bit there is a possibility of 256 numbers and with 16-bit there are 65,536. The higher the bit depth, the better quality sound that is produced.

Sound File Sizes

Just like in anything else, sound quality has a price, and when storing sound files this price is paid in hard drive space. If you record at audio CD quality (44kHz, 16-bit, stereo) you will consume your storage space at the rate of about 10 megabytes per minute. Even with the falling prices of hard drives these days, that still adds up to one big file if you let the recording roll for more than just a few minutes.

Cutting the sample rate to 22 kHz will reduce the resultant file size by half. However, the quality of the sound will be reduced as well, heard primarily as a loss of high and low frequencies. Switching from stereo to mono will reduce the file size by half again, a reasonable option when recording sounds such as voice.

Sound File Formats

Once we have the analog sound converted to digital it needs to be saved, just like any other computer file. The format, or file type, in which the sound is saved depends both on the capabilities of the software you are using and the intended use of the file. If the sound is to be used as a system beep on your computer you would use a different format than if the sound is to be delivered as streaming audio on the Internet.

As mentioned earlier, there are many formats available and I will make no attempt to cover all of them. What I would like to look at here are several of the everyday sound file types that you are likely to encounter in your classroom.

AIFF The is the Apple Interchange File Format and is native to the Macintosh world. If you are making sounds files on the Mac and don’t know what format they are in, this is probably the one. They can be transferred easily to the IBM world.
WAV This is the PC equivalent of the Mac AIFF. Developed by Microsoft and IBM, it is the standard sound file format used on Windows systems. It is also quite popular on the Internet, though compression technologies are superseding it.
AU One of the oldest sound type around, it was one of the first file types to be widely supported on the Internet.
RealAudio This format has taken the Internet by storm. A special player, free from http://www.real.com, is required. Unlike the three previously mentioned formats, it is not necessary to download an entire RealAudio sound file before you can start listening to it. RealAudio files "stream," that is, they start playing as they are downloading. The price paid for this technology is that the original sounds are usually sampled at a lower rate and the file is then compressed with an algorithm that throws away some of the sounds that it "thinks" that the listener might not care about anyway.

At best, the result can sound like listening to FM radio. At worst, it can sound like listening to an AM radio played over a telephone. Despite these shortcomings, it also allows the streaming of live content, a feature that lets thousands of radio stations feed their signals into the Internet. If you are interested in hearing these radio stations, point your web browser to www.broadcast.com.
MPEG Developed by the Motion Pictures Experts Group, the sound file formats that have come out of this organization are quite amazing, especially the latest MP3 variety. Once a sound has been recorded in a standard format, such as AIFF or WAV, it is compressed using a proprietary algorithm that results in a better quality, though slightly larger file than that found in the RealAudio method. The advantage here is that the original file size can be reduced by 90% with virtually no loss in quality. A selection from an audio CD, which might easily be 50 megabytes in size, can be squeezed down to 5 megabytes, making downloading almost a reasonable operation. A special player is required to listen to these files and free ones are available on the Internet in many places for both Windows and the Macintosh (www.mp3.com).

It is this technology that has the recording industry up in arms, and rightly so. Using a piece of software called a CD-ripper, the contents of an audio CD can be moved to the hard drive in a matter of minutes. The resulting files, usually in AIFF or WAV formats, can then be compressed in a batch operation that yields the individual MP3 files. These files can then be placed on a Web page for others to download, a clear copyright violation.
SoundVQ Similar to MP3, this competing compression method was developed by NTT Corporation and is supported by Yamaha. Like MP3, a special player is required, which can be download from the Yamaha SoundVQ website (www.yamaha.co.jp/english/xg/SoundVQ/) in either Mac or Windows formats. As of this writing I have not used this file type but have read that it is superior, in some ways, to the MP3.
MIDI While this is not a sound file format that is used to record sounds from the real world, it does play back sounds, but of the computer-generated musical variety. MIDI (Musical Instrument Digital Interface) is the world standard in how we connect musical instruments, such as electronic keyboards to our computers. Rather than storing digital information about the original sound, a MIDI file contains data about what notes were played, their duration, and the instrument that should play them. Compared to other sound formats, MIDI files are very small, typically less than 100,000 bytes.


Software for Creating Sounds

To record an analog sound you will need, in addition to the microphone or other input device, a program that will let you edit and save the digital sound in the desired format. Like all software, some is better than others and some is more expensive than others. There is some good shareware out there, however, if you want to set up your own professional recording studio, you will need to invest in some of the higher-end packages.

It needs to be said right up front that Windows users have a distinct advantage here. All Windows computers that have a sound card, and that means most machines built in the last 3-4 years, come with sound editing software. This bundled software has limited features, but for the most part will record a sound in a variety of bit depths and sample rates and save the file in WAV or several other formats.

For more features, Windows users might look at CoolEdit 16, a shareware program that is highly rated among the audio heads. It is available from

www.download.com


For Mac users there is just not a piece of shareware that I have found that will do what the most basic audio software can that comes bundled with a Windows system. Mac users will need to spend a few bucks, but the software available is very good. The two packages that will take care of most user needs are SoundEdit 16 from MacroMedia (www.macromedia.com) and Sound Forge from Sonic Foundry (www.sfoundry.com). These programs, also available for Windows, are extremely powerful and have more features and options than I can cover here.

Wrapping It Up

To think that the audio file formats that we have today are all we will ever need is unrealistic. The compression techniques will continue to improve, allowing us to put more data into a smaller space. The ability to push more data through the Internet will enhance our Web surfing. So hop on board and give sound creation a try.

The above portion of this page was reprinted from the MCCE Newsletter, December 1998. If you wish to see the online version of this and other issues, click here.

Getting It on the Web

Why Move to Digital?

OK, so now we have a collection of Old Time Radio programs in an analog format and along come advances in digital technology that provides additional options for storing, trading, and broadcasting programs. The development of the World Wide Wide gave the general public access to the Internet and the continual increase in bandwidth is slowly removing the obstacle of long download times.

I set up my first web page in 1996 (I think), and included some OTR data in it. However, I have a limited amount of space with my ISP and it wasn't until I was given a direct connection to the Internet in my classroom that I could entertain the idea of expanding my capabilities. I set up a web server for the school in January 1997 and began developing content for it. In 1998 I started playing with making MP3 and RealAudio files and decided that I needed to get into this in a big way. In the summer of 1998 I started going through my reel-to-reel collection and moving shows from the analog world to digital.

My reasons for doing this:


Hardware and Software

As my old, but still worthy, 486-66 was a bit underpowered to handle this task, I decided that I needed a new computer. I spent the summer shopping it out, ending up with a Pentium II-350. In the meantime I borrowed a computer from my classroom and made the first batch of programs using it. The computer is a Pentium II-233, with a 4 gigabyte hard drive and an Equisonic sound card. For recording software I use Sound Forge, Version 4 for Windows. I experimented with several sample rates and for what I am trying to do, the 22,000 Hz, 16-bit option is acceptable.

Each show is recorded in mono, resulting in a .WAV file that is about 70 megabytes for a 30-minute program. I avoid the use of the equalizer if at all possible. On occasion I have used it to slightly sweeten up the highs, but only slightly as been overly aggressive with that option in Sound Forge can introduce flanging and other artifacts that take away from the quality of the show.

On a few shows it was obvious that they were not playing back at the proper speed. Sound Forge has a pitch control that can speed up, or slow down the file. Another tool that I have found handy is the time-compress/expand option. I don't use it except to edit shows for my television program, but it can change the playback time of a show without changing the pitch.

The only other editing that I do to the .WAV files are to clean up any content problems. Since I recorded most of these shows from the radio, I might have to trim out a back announcement, a station ID, or other data that was not part of the original show. I also remove any skips, a process that is very easy to do digitally compared to the old methods used with recording tape.

My reel-to-reel recorders have an auto-reverse feature that automatically switch the playback from side one to side two. I do not utilize this during recording because during the process of reversing there is a slight loss of content. While I have used this for playback during my casual listening, I disable it during this transfer to digital. When the tape runs out on side one I just let Sound Forge continue recording while I get side two of the tape started. When the show is over, I go back and edit out the gap, pasting the portion of the show from side one of the tape to the portion from side two, creating a complete show with no click or pop where the edit was made.

At the time, I would record about 12 or so shows per day (I don't work during the summer), and let the computer convert them to MP3 files during the night. I have used several MP3 compression programs and right now don't have enough of an opinion to say which is the best way to go with one of these.

I quickly saw that I needed a CD-ROM recorder to archive these shows and picked up one made by Sony. It is a 2X writer, non-rewritable. My research indicated that there is still some backward compatibility issues with using the rewritable drives and, considering that I may need to play these CDs on a variety of machines, I wanted to go with the most stable technology.

When I have about 120 program, I burn them to a CD, all of the shows in an MP3 format. If all of those shows were dumped into the playlist of the Winamp MP3 player, the CD would provide about 60 hours of continuous Old Time Radio. During the summer I was able to transfer about 600 programs, burning 5 CDs.

Getting the Shows Online

Starting on July 1, 1998, I began making shows from my collection available on my web page. I created a page on my web server to handle this and at the time I thought that it would not be any big deal to update the page once a week. However, after doing it a few times I realized that the process needed some automation.

As part of my attempt to keep up with current technology, and to stay one step ahead of my students' questions, I made the decision to make the Old Radio Show download page a dynamic one. That is, the HTML for the page will be generated, on demand, by the visitor requesting it. I have been dabbling in CGI and PERL programming for a while and this seemed like a good project and the first version of the page went online in the fall of 1998. When a user requests the page, a PERL program reads a database that contains a list of shows, formats that information in HTML, and delivers a completed web page to the user. I added a voting booth to allow users to provide some input as to which show they wish to see the following week.

Details about building interactive web pages is being covered in a series of articles that I am writing for the Montana Council of Computers and Technology in Education Newsletter. The first article in the series is in the December 1998 issue. You can get to it by clicking here.


Going on the Air

Related to this whole project was the setting up of a radio station for our school. Because of the cost and legalities involved with going on the air with a station that could serve the entire city, we have going with low-power FM transmitter that serves the school itself. I have a student who is very interested in radio, computers, and computer programming who wanted to do a year-long independent study in setting up and maintaining an automated radio station. Early on we decided to use Old Radio shows for the bulk of our content because:

We went on the air in September 1998 with a FM-25 transmitter from Ramsey Electronics. Most of the content is OTR and if you are within the couple hundred feet of the transmitter, you can hear it pretty well. To give those outside the range of the signal to hear us, we set up a web page with 24 hours of station content available via RealAudio. You can hear these shows by visiting the station's web page here.

To create the RealAudio content for the station, I first make a recording, as detailed above, saving it as .WAV file. Sound Forge has the ability to save in RealAudio format, which I have done, but lately I have been using RealProducer G2, available from the RealAudio website. Once the shows are saved in this format, they are placed on a computer (a Pentium 75) on our network that is running the RealAudio Basic Server G2, also available from the RealAudio site. I then create the necessary web pages and links to these shows on the main web server.

Where Do We Go From Here?

Who knows? If we get more bandwidth on this end, maybe I can make more shows available at one time. I am experimenting with streaming live broadcasts from our radio station. I have also experimented with streaming video content along with the audio but it's a little flaky yet.

Keep checking back, and I'll keep coming up with new ways to combine the old and new technologies.





Back to the Old Time Radio Show Server Information Page
Go to the Old Time Radio Show Server Page




Handmade HTML  Made with Notepad

This page was hand coded by Vince Long

Copyright © 1998, Vince Long