Tech Tip:Mastering Your Music: Secrets of the Pros

Get signed now with




The Myth of Digital Sound
©2000 John Vestman


Myth: Digital sound doesn't loose quality when transferred. There has been a long-held thought that digital can be transferred, cloned, copied, backed up etc. and the sound stays the same. Numbers are numbers, right? They are... if they stay exactly the same, but in the world of moving media/disks/tapes, they don't stay exactly the same.


Here's proof. Computers crash. Hard drives lose data. Files get fragmented. Norton and other utility programs are popular ways of "recovering" corrupted files. If all the one's and zeros worked perfectly every time, none of these problems would occur, since each file would do exactly as it was supposed to do each and every time.


Set up: We are "imprinted" with the experience of analog music media. In other words, we are accustomed to an analog signal remaining the same when an analog volume control is moved. We are used to analog equalizers that shape the waveform vs. digital processors which internally recalculate numeric data and create an entirely new file.


Granted, digital does what it does well. It just has a different set of limitations that we weren't expecting when the technology came out. This is the same thing that occurred when we discovered that solid-state electronics didn't have some of the cool sonic qualities that tubes did.


SHOCK: Dat safety "clones" don't sound the same. They sound more brittle in the highs, and less depth in the reverbs and room sounds. Cutting cdrs at high speed (like 2X, 4X, etc.) also adds hardness to the highs and mid-highs. As for cdr copies... well, as my Sonic Solution dealer says, "There is no such thing as a cdr clone."


Different cdrs sound different. Even within the same brand. I've listened to Kodak's gold-on-gold "Writeable" cdrs, their gold-on-gold "Digital audio" cdrs, their gold-on-gold "Recordable" cdrs, their platinum-on-silver cdrs ... they all sound different! (Gold-on-gold "Recordable" is my favorite... they even sound better than the "audio" cdrs.) TDK's sound different from the Mitsui's, and the bargain-basement cheapies are lucky if they even play!


Even transferring a sound file from one hard drive to another changes the sound slightly. Never mind "saving" files to cdr or dat backup and then reloading back into a hard disk system. With all this technology available, musicians still have to balance (1) the amount of time it takes to learn these details (2) their time and budget and (3) the craft and passion of their music.


Cutting cdrs (ok, burning cdrs...) at 2X sounds different than 1X. I invited a professional engineer and a stereophile guy to listen to the same album on two different cdrs... one cut at 1X one at 2X. The engineer preferred the 1X, and thought the cdrs had different mixes on them. The stereophile guy simply felt the sound on the 1X was sweeter and wider.


My recommendation: I make sure that I include the instructions, "Cut glass at 1X ONLY" when my clients send their cdr master in to be replicated. Even though some pressing plants will say that cutting their glass master at 2X or more creates fewer errors (and saves them valuable studio time), I insist on 1X, just as all the major mastering guys do.


Here's another example:

Q) John, there is an over all harshness to my cd, especially in the vocal sound. I used a Neumann 149 mic into a tube pre, a DBX 160 compressor with no eq into my Roland VS 1680. I know I made the mistake of having the cd replication company handle the mastering, but if everything is digital, why am I getting that harsh sound? The level of my cd is softer and not as full sounding as most commercial cds. -Paul


A) I would recommend a de-esser to soften any peaky sibilance which can be a source of harshness... but I suspect more. Just to be sure, tell me the other recording events that led up to the pressing.


>I eq'd and compressed everything in the 1680, then made a dat mix. I transferred the dat mix back to fresh 1680 tracks to finalize my levels, sequence, and overall eq. I mixed that to another dat and sent it to the factory for mastering and pressing.


Whenever you change eq, compress or change levels in a hard disk system, you are making calculations in the computer. The actual word length is changing and the original "analysis" made by the A to D converters is being changed. The sound is being altered, and generally this means that some resolution of the tone is lost. This translates into a harder, shallower, less detailed sound. Smoothness, width, depth, and openness are some of the first qualities to go whenever these calculations are made.


Key: Transferring your first dat mix back to the 1680 means a loss of 2 digital generations. Dat machines and hard disk systems have moving components, moving tape etc. that are subject to errors, especially when handling huge sound files. Cable problems and word clock issues can cause jitter which affects solidity and resolution too.


Granted, there's no tape hiss. But if you eq'd a lot in the Roland, transferred back and forth, eq'd some more, and didn't have a high-end mastering engineer to refine your master, it's no surprise that the sound is harsh.


Solution: Use a de-esser, use an analog eq on your vocals going into the 1680 and through an insert when mixing. Then, either bring the 1680 to the mastering session (which gives you the flexibility to actually make mix changes on the spot!) or mix to analog 2 track. If you can't get an analog machine, rent some high-end A to D converters going into your dat machine or cd burner. Don't use TDK or Quantegy dat tapes - I prefer Maxell or BASF. Finally, use pro mastering, and you'll be much happier with your final cds!


Q) I am pre-mastering to cd. When I "normalize to peak", is the subtracted headroom helpful to the mastering engineer? Should I not normalize to peak? -Larry


A) Normalizing is a commonly misunderstood function, and it's not a huge benefit for the mastering engineer. Once a musician gave me his cd with a song on it that was 10 db lower than the "standard" volume today for that kind of music. When I showed him how low the level was, he replied, "But I normalized it!" What he learned is that it takes serious mastering to make a hot cd that retains the dynamics too.


Normalizing only does one thing. It brings up the level of the entire song or file to "digital zero"... the loudest possible level without overload errors. This level, however, is only as loud as the highest peak point within the file. So if your song hangs at around -20, but then one peak snaps out at +5, everything will base from that peak, and go down from there. Everything stays proportionally the same.


Headroom is the amount of available signal prior to clipping or distortion of the sound. If a mastering engineer is loading in your music using analog processing first, then it doesn't make any difference whether or not you normalize, because he/she will simply adjust the analog processing devices to accommodate whatever level is there.


However, if the mastering engineer is taking a digital signal into the mastering system and processing it in the digital domain, then the answer to whether you should normalize is based on these questions: Who has the better digital system, you or the mastering engineer? What over all volume level are you looking for on your final cd?


If you normalize your song/file, your computer is going through calculations that change the one's and zeros of your original signal. Huge amounts of level change will add some hardness and shrink the "sound stage". If you want more level, decide if you think your computer is better than the one the mastering house uses. If you're using Pro Tools and the mastering house is using Sonic Solutions, let the mastering house do the level changing. Optimize your gain structure within your system to get a good level on your cd, but don't compress the stereo mix buss or "brick wall limit". Let the mastering engineer take it to the next level.


©2000 Exclusively for All Rights Reserved.