Some Fundamentals about Sound

Sound travels in the form of waves - pure sine waves, square waves, complex sines, etc.


In Demo B - Exhibit A

we see a simple 440Hz sine wave at 80% of digital full scale.  Below it in that same image is a limited version of that same 440Hz sine - notice the peaks(tops and bottoms) of that sine have been removed, flat-topping the wave.



Please play Demo A   to hear the sine wave represented at the top.

Please play Demo B    to hear the sine wave represented at the bottom.

Hear a difference?  I'm sure you did.  Limiting(or clipping) any sound wave, however simple or complex, adds additional harmonics and changes the sound of the wave!  Now imagine that effect as it applies to real music.  Musicians and artists need to think about this when they demand their recorded mixes be made "louder" by the mix and mastering engineers hired to handle their project.


Next, let's take a look at the effects of "loudifying"(as I call it!) on complex waveforms, such as actual music.

This is an excerpt from Van Halen's Spanish Fly - a guitar solo - Demo F Exhibit E:

On top in that image is an excerpt from the original track on the original CD release.  On the bottom is my recreation of a typical "remaster".   Note that all that has really been done to it is some compression, hard-limiting, and make-up gain to bring what's left up to almost digital full scale.

It should be noted that dynamic range compression, and it's more potent cousin, limiting, do not by themselves make a piece of audio louder.  What they do do is make the louder/loudest portions softer, reducing their amplitude and proximity to full scale.  A final step - makeup gain, must be applied to bring "what's left" of the waveform back up to almost full scale.



This image, Exhibit G, Demo A(as heard above):

 should clarify what most of what is sold to the public as "Digitally Remastered!!" has really had done to it.  Within Exhibit G, you have on top, the original waveform(or your song!), in the middle, limiting and/or compression reduces the amplitude of the selected loudest portion, and on bottom, the remaining waveform, the quieter portion, has gain applied to bring it back up to full scale - LOUDER.

Not only is this "remaster" louder, but it sounds different - even if the effect is more subtle in real life than in my basic demos.




Let's listen to just a snippet of "Spanish Fly" - first, the original:

Secondly, the "remaster":

Now, initially, a lot of you will be impressed - don't try to deny it! - by how much louder the remastered snippet is than that of the original.  This is the case especially if you are viewing this blog on a small mobile device through earbuds.  But if you turn it down to a comfortable volume - or to match the volume of the original - you will notice a lot missing, plus some tonality that is not present in the original.

You might even mistake these additional harmonics as revealing elements previously unheard in the original version. All you are hearing is what I demonstrated, at the top of this section, in my simple 440Hz waveform demos and illustrations - stuff that shouldn't be there.

And remember: Making something louder is not remastering.  The only reason you might more readily hear background instrumentation, or other sounds(fingers on strings, the squeak of the kick drum pedal), is because something was sacrificed - very often the louder transients, hard limited or heavily compressed away, so that these quieter elements could be elevated to make use of the new headroom.  But there is a reason some elements are louder or softer than others in a song, why a chorus or refrain is usually LOUDER than the verses: it's called emotional context.  If everything is made equally loud, and pushed up to full scale, from the lead vocals to the percussion to the melody to the backup vocals, it's not natural, and can get fatiguing after awhile.


As a side note, I need to point out that digital audio, despite it's proven sonic advantages, utilizes a lot of the same terms to convey different meanings.  One such word is "compression".   It can refer either to data compression - which is what lossy codecs such as MP3 perform to make an audio file smaller to fit more of them on a digital player, or, to dynamic range compression(the topic of this blog), where the louder parts of audio are made quieter.  Please make a note of this so you are not confused!  :)