Everyone who has a PC with a sound card and a radio station has wanted to record a signal at least once. It can't be anything complicated! Yes, it's not difficult, but the result rarely lives up to expectations. How to do it professionally?
First of all, it is necessary to connect the PC to the station so that hum does not enter the signal. I use a direct connection with a shielded cable with two 3.5″ jacks. I have soldered 1nF capacitors in the connector housing as protection against HF ingress. This wiring works fine with the SB128, but not with the second card on the motherboard. After connecting the station with this card, an extremely strong hum entered the signal.
If a similar problem arises, it can be solved with an interface with separate lands. It can be with an NF transformer or with an optocoupler. After a suitable connection, it is enough to look on the Internet or in the literature.
Of course we need some program for recording. However, Windows Sound Recorder is not suitable. I use Creative Wave Studio, which was on the CD with the sound card drivers. The length of the recording is affected only by the free space on the hard disk, and the recording can be edited and cut. However, everyone probably has their favorite program, so they choose it themselves.
The next step is to set the sensitivity of the sound card. Turn the volume control on the device to the usual position and use the mixer to adjust the sensitivity of the sound card so that it is not woken up. On some mixers there is a wake-up indicator, if it is not there, then we set the regulator to about half and record a few seconds in mono mode at the maximum sample rate of the sound card. We save the file to disk.
Let's start the Peak Limiter program. This program is used for optimal recording. If the sensitivity is too high, it will limit the peaks, if it is low, then it will increase the volume by a certain coefficient. The whole program is very simple. We must start by opening the WAV file. Two spectrographs will appear on the screen, right and left channels. After clicking the Analyze Peaks button, the program analyzes the modulation peaks in the recording. If there are many peaks exceeding 80 and 90% of the maximum, then the recording is awake and the sensitivity of the sound card needs to be reduced.
If the recording is good enough, we proceed to the Normalize Peaks button. Above it is written data in percentages, which we can rewrite. All peaks above this value will be attenuated. There is no need to overdo it with restrictions, it is possible to determine the real value from the analysis. Right next to it is the Normalize Volume button, which ensures optimal wake-up. Of course, we build at 100%. At the end, we can play the edited recording with the Play button and then save it with the Save button.
However, such a record takes up quite a lot of disk space. I recommend saving it as an MP3 file. There are again countless programs for compressing WAV to MP3. I use bEncoder. It has several settings suitable for efficient processing of voice recordings. First, select the WAV file(s) and select the output format (MP3). In the top menu, I click on MP3 Format and I can set the compression parameters. I will set MONO, sample rate to 11025 Hz and bitrate to 16 kbit. This is the main setting (GENERAL), in addition I use Filtering for some records. It offers both a high-pass and a low-pass filter. Then I go back to the General settings, click Apply and go back to the main window where I click Convert.
The result is a file about 50 times smaller with an almost imperceptibly degraded quality. Files of a few kilobytes are also suitable for publishing on the Internet. The way it is written, it may sound complicated. However, you have already done most of this, the professionalization lies only in the use of Peak Limiter programs http://www.peaklimiter.com and bEncoder http://technet.idnes.cz/sw/sw_audio-video/sw_audio-video_mp3/audiofilez2003-03.html.