In the early 70s all recording was done on analog tape, then transferred to vinyl records. Digital recording was just beginning to make it’s way into the industry, but was so expensive that only the major record labels could afford this technology. As with computers, the price of digital recording came down until commercial recording studios could afford it. Then in the 90 it came within reach of semi-pro studios as well.
Since the 70s, digital recording has evolved into many different formats. It used to use the same type of tape as analog, but recorded bits and bytes instead of music. Those bits and bytes were series of numbers that were converted into sound. Today, as sound is recorded, it is converted into the same bits and bytes, then back into sound, but there are several types of tape formats and several types of non-tape formats. The non-tape formats allow for a process called “normalizing”.
When recording on to analog tape, a VU meter is used along with Peak meters. The VU meter shows the average level of the signal and the peak meter shows the level that the peaks are hitting…..much higher than the average that is monitored by the VU meter. On analog tape, those peaks could reach various levels before saturating the tape and distorting the sound. The levels would depend on the quality of the tape used, the width of the tape and the speed at which it is running while recording. Therefore, if your peaks hit 0 db on the peak meter, you might still have more headroom before saturating your tape. The idea in professional studios was to reach the threshold of saturation. This allowed for the best signal to noise ratio.
When recording digitally, O db is as far as you can go with your peaks. Beyond that, the sound goes into clipping (part of the waveform is clipped off and the sound is distorted). It is nearly impossible to get the hottest peaks at 0 db during the instrument recording process because the engineer has limited control over the peak variations played by the musicians. So to avoid going over 0 db, it remains safer to stay under 0 db with the peaks. However, to optimize the recording, the hottest peaks in the song should just hit 0 db without going over. This is done through normalizing, after the final mix and during the mastering process.
To normalize your song, the entire song is fed into a computer with the appropriate software. The waveform of the song is then displayed on the computer screen in the form of a graph showing the level of each peak in the song, all of which should be under 0 db. Normalizing once will take the highest peak in the song (more than one if they are at equal levels) and bring that peak right to 0 db. Doing so also raises the level of the rest of the song somewhat. The song is now ready to be transferred to CD.
Some music, such as dance or rock has to be extremely loud to compete with other songs of the same type that are on the market, so normalizing is often done several times, bringing the next peak to 0 db, then again to bring the next one to 0 db and so on. This can only be done so many times before the song begins to sound unnatural. The key is stopping before it begins to sound unnatural. This optimizes the overall level of the song and makes it louder. It can be made even louder yet by heavily compressing the mix prior to normalizing.
For complete private instruction, See contact info for private lessons