 |
||||||||||
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
DXer & Audio
Restoring old audio or SW recording
1. DO YOU NEED TO KEEP A TRACE ?
For many reasons, it's always interesting for a DXer to keep recordings of his better broadcast band catches. Whatever your activity, BCL, utility DX, Ham radio, you need a sound library to keep the trace of a rare transmissions. Many years later, you'll easily find your historical audio clips. Of course, a consequent library must be used with a sound manager software.
CDTree : manage, sort out or search your audio files
2. WHICH AUDIO SUPPORT ?
The standard cassette recorder remains a cheap and easy to use equipment, but today with the digital technology, more and more DXer are using some new audio supports like Mini Disc or CD. Filing audio data on hard disk must be considered as a temporary solution because of frequent crashes of our operating systems. So let's see how to...
3. PLUG AND ... RECORD
First you must connect your receiver (REC OUT) to your sound card (LINE IN) with an appropriate cord (RCA or mini-jack 3.5).
Sound card : Line-input
Receiver : Line-output
One of the inputs is marked as the microphone input either by the word "Mike" or "Mic" or by an icon of a microphone. Don't use it, and search the line-input. Check the markings on the back of your sound card to determine which one it is - there are usually either symbols or text labels back there to help you out. In this case, the Line-in is the first plug on the left and the Mic plug is the second one.
All the recent receivers have an audio output. On the rear panel text labels or markings will help you to find it ( REC OUT, AUDIO OUT, LINE OUT). Avoid the headphones jack because in the recording step you could have to adjust both the audio gain button on the receiver and the recording level in the Windows mixer. Generally, this output is an RCA plug on table receivers and mini-jack 3.5 on portable set.
4. INSTALL AN AUDIO SOFTWARE
All the audio software with a record function can be used. Don't use the Windows Recorder because if you have very long file to record, you'll be limited by your RAM size. Audio editors are mainly designed for music, sound design or audio restoration but as you have to manipulate sounds you can use their numerous commands with benefit. Don't forget that edit functions are the common base of these software but audio editors like or have also more sophisticated functions and effects to clean up or modify your audio. They also present a very good quality/cost ratio.
Scanner RecorderScanner Recorder is a free audio recorder that is primarily designed to record speech. It has a VOX control that allows the user to save disk space when no sound is present in the signal. Scanner Recorder provides a switched recorder that listens to the sound coming into the audio card and if the sound is above the minimum threshold, the recorder operates. If the sound level falls below the threshold, the recorder pauses. The output file format only supports WAV files. Currently you will need a separate WAV player capable of playing very large files. The Media Player that comes with Windows works fine to play back such files.
5. RUN THE WINDOWS MIXER
If you don't adjust the recording level in the Windows mixer, you could get a silent file. So Run the volume control panel - Select Options / Properties / Recording - The "Line-in" must be checked then OK - Back to the mixer, move the line-in cursor up or down to increase or decrease the recording level. During the recording, you can check the level meter in your audio editor.
6. THREE AUDIO FORMATS FOR THE DXer
Many audio formats exist, but for radio applications only 3 are very important :
Wav format (.wav)
Wave is the standard format for uncompressed audio on a PC - as close a copy to the original analog data as possible - it is therefore much larger than the same file would be in a compressed format such as mp3 or RealAudio. The Nyquist's demonstration shows that to code a signal correctly, the sampling rate must be at least equal to the double of the highest frequency present in the signal to sample. Thus for a telephone quality, the sample rate will be 11,025 Khz - 8 bit - mono. That will make it possible to treat frequencies going up to 5500 Hz. It's largely sufficient to make a voice perfectly comprehensible.
If you project to use your audio samples in some audio applications, it seems preferable to record in the higher quality. Later, it will be possible to resample the file if it's necessary.
MP3 (.mp3)
Mp3 is a popular compressed audio format widely used to transfer music over the internet. MP3s are created by taking wave audio data and processing it with a special algorithm. This alrgorithm removes parts of the audio that theoretically cannot be detected with the human ear. Actually, there will be some degradation of quality, but this depends on the quality (bitrate) with which you choose to encode the file. By using MPEG audio coding, you may shrink down the original sound data from a CD by a factor of 12, without losing sound quality. Factors of 24 and even more still maintain a sound quality that is significantly better than what you get by just reducing the sampling rate and the resolution of your samples.
RealAudio - RealMedia (.ra , .rm)
This is a very popular streaming audio format often used by internet radio stations and for posting sound files on websites. RealAudio files can be smaller than MP3 files but are of lower quality if compressed enough to play over a slow connection (such as a 56 kbps modem). A 16 kbps compression can be used for speech with a 8 kHz frequency response.
7. AUDIO FILES... WHAT FOR ?
Which format? Which quality ?
You can keep recordings of historical stations, of your favourite programmes or short IS/ID. If you appreciate tekno music you can even use your samples in your own compositions ... You can also post your samples on websites or email them...To make a choice between compressed or uncompressed format, higher or lower quality, it depends on what you want to do.
If you want to make your own archives, I think you can use compressed or uncompressed format with the higher quality.
If you have to post your samples, your files must be as light as possible to minimize the server space. The band-width occupation during data transferts is also important. In this case, you must choose a compressed format and for the quality you have to seek a compromise which gives you satisfaction.
... When you're listening to the shortwave you don't expect a High Quality. The main point is to catch an unusual DX station. In this state of mind, heavy noise or jamming is not a real problem. If some music is running you know that your receiver is not a CD player, so the voice of the announcer giving an identification is at this time the most important.
8. CLEAN UP YOUR SOUNDS !
There are 2 ways to consider an audio archive :
First you can save the file as it is. I mean with noise, interferences, without touching up the level or the audio spectrum. In the second way, the file is recorded with an audio editor and you apply digital process to remove unwanted noises to make your sound clear. In my point of view, if you want to constitute a serious audio database recorded on CD or MD, clean up your sounds. In that way, each audio clip will be ready for a future use.
As you can see, this website is about sound, SFx, audio restoration and vinyls to CD. Today, many audio samples are used by musicians for their own compositions. In the audio restoration's domain we have to clean up voices, music, as well as an historical speech or some audio mixed with our last video.
In this case, different noises appear :
- The constant noise like hum produced by 50/60 Hz electric current or hiss on tapes. The audio level of this noise is more or less important but it's heard continuously when you play your sample.
- The accidental noise is most of the time very short. It appears when the audio support presents one or more defects as for example an important impulse due to an electric noise.
The DXer will be confronted with a wide variety of noise and interference, high level ambient noise, hum, hiss, atmospheric disturbance, broacast on adjacent channels and a damaged audio quality if some fading occurs. The recording level must be high enough if you want to get the best signal/noise ratio. If your sample is low level recorded you'll have to increase its volume... and unfortunately the noise level too !
For those who are CoolEdit 96/2k/pro users they can take a look at the excellent website Audio restoration with CoolEdit 2000 by A. Reny - All you need to restore audio samples with this audio editor. You'll have to experiment with various adjustments for the numerous audio tools and it's always useful to start with basic tips. Most of the time, the tools have presets to make the tools ready to use. You will find very interesting informations on these pages.
As a general rule, all your radio samples will be cleaned up as follows :
Low-pass : Used as a de-hisser or to eliminate "ssss"
High-pass : To eliminate very low frequencies or mike "pops"
Band-pass : Combination of LP and HP filters. This filter is generally used to increase intelligibility (single voice, interview, talk radio)
Notch : To eliminate a nasty 50/60 Hz hum
For example, choose 50% to compute values needed to amplify the selection no more than 50% of maximum (resulting in a 3dB attenuation from maximum output.) Choose 100% to apply the greatest amount of amplification possible without clipping.
Some recording hardware may introduce a DC Bias, which results in the recorded waveform appearing to be above or below the normal center line. Use DC Bias Adjust to center the waveform on the center line (zero voltage).
If you want to simulate a special acoustic space like studio or particular rooms, you can use a light reverb effect. It also gives a nice presence to a voice. With an Equalizer, you can boost or cut the signal at particular frequency bands, so you can use a light boost 3/6 dB to give more presence to a voice around 5 KHz.
In audio restoration there is a general rule : The suite of audio processing routines that clean up your sounds must not introduce processing artifacts. In that case you get a bad twisted audio. So the result and of course the audio quality depends on your fine-tuning. The more radical the removal, the more is affected the quality ! Sometimes you'll think it's better to keep some noise ! Radio recordings are not so problematic since you mainly work on male or female voices (frequencies up to 5 KHz). When you clean up a vinyl soundtrack the problem is really different because you work on music produced by several instruments. So it's easy to get a "fatal error" if the frequency spectrum is heavily altered.
We have very powerful tools to help us but audio restoration will take time, a good deal of practise and critical listening.
