February 2011

One of the most important steps during the mastering engineer’s work is what’s known as “Dithering.”

Dithering is applied at the end of the mastering process to ready your music to be put on CD, DVD, MP3, Vinyl, or any destination format.

It’s needed because the mastering work is often performed at either 24, 32 or 64 bit (a high resolution). However, most playback devices/destination medias do not currently support anything higher than 16 bit (CD) or 24 bit (DVD).

This means the bit-depth must be reduced. You can think “Bit depth” as the steps of loudness that can be represented. 8 bit only supports 255 “loudness steps” which the music’s volume-dynamic must be fit into. While, 16 bits supports 65535 steps (a much better resolution already).

The ladder then keeps on climbing with increasing bit depths.

However, a problem occurs when you reduce the bit-depth. The loudness steps can’t be evenly reduced because there are more “loudness steps” occupied by the music’s volume-dynamic in the higher bit depth than available in the lower destination bit depth.

Unwanted distortion occurs as a result.

This problem is what’s commonly known as the, “Quantization error”.

To fix this problem the process of dithering is used. Dithering makes the distortion unnoticeable to the human ear by adding noise to the lowest signals to help fill-in the uneven loudness steps.

Simply put, dithering is an essential part of mastering to eliminate low-level distortion (a “grainy” sound) for creating the truest possible sound.

About the Author:
Lorenz Vauck is an Audio Mastering Engineer, Musician, and Internet Entrepreneur from Dresden, Saxony, Germany. He is the Managing Director and Chief Mastering Engineer of XARC Mastering, one of the world’s first online audio mastering studios established in 2003.

Dear Friend,

I hope you had a good first two months of the new year?

I just wanted to quickly let you know that I have again added a new before / after demo to the ever growing big before / after demos section of my website. This time it is a song from Lewis Newman called “Piano Track”. It´s a very dark and ambient song, with a very nice mood to it!

Here is what Lewis replied to me after I´ve sent him the approval master so he could let me know any changes he might require before we go final:

“Hi Lorenz,

I just had a listen to it on a few different systems, and i think it sounds absolutely amazing!…

I am so pleased with it, so much better than the unmastered mix. There’s nothing i would want to change.

I did an A/B comparison, and the unmastered one seems so flat and boring compared to the new version!

I just want to thank you for everything.

The master is exactly what i wanted, and on top of that, your service and communication is absolutely excellent.

I’m doing a demo to send to music libraries at the moment, so i will have more tunes for you soon if that’s ok!…

Thank you so much again, you are great!

speak to you soon,


The before / after demo can be heard at http://www.xarcmastering.com/demos/ (Lewis Newman – Piano Track)

Besides a pretty huge before / after the mastering difference, this demos again shows the massive benefits of my newly installed equipment, including a “new old” tape machine, new passive tube equalizers and a new custom build dynamics processor which again shows its full potential on this demo if you listen to how much details are revealed “out of the dark”.

Especially the tape machine and the tube equalizers, which I carefully select for each and every boost/cut I do on each song, gave the song this massive and fat, airy but still warm analog sound. With all those possibilities at hand, it often takes a few days to finish just one song! But hearing the results and getting such feedback from my clients, I always see it´s worth it to not do it like “common” mastering studios which finish a song in an hour or two usually.

So enjoy listening and let me know any questions you might have, I am always here to help and listen to your thoughts / ideas / questions.

Probably the most important tools in the mastering engineer’s toolbox are equalizers.

Equalization is what makes about 95% of the sound by adjusting the frequency of the various components of a song (such as the bass or treble). It’s essential to bringing-out the vocals or instrument sounds, or to add some “sparkle” and “shine” to the song.

Additionally, it’s vital to making the songs come together and sound coherent. For example, one song might have too much bass, while another doesn’t have enough. You can then reduce the bass on the first one and boost it on the second one so that you have a same level in both tracks.

When applied to vocals, you may use equalization to make them less “nasal” or give them more “body” / “warmth”.

For instruments, equalization is important because many instruments have complex frequency components that are difficult to mic correctly. Equalizing can balance this.

It can also help emphasize certain sounds of an instrument you like and eliminate ones you don’t. For example you can raise the attack / kick of a bassdrum or boost its “body” so that it has more power in the bass area.

Another example might be if you had a whistle sound that’s too loud in a Brazilian song. With equalization, you can reduce the exact frequency of the whistle to make it fit in better.

Inside the mastering studio, there are a wide range of EQ´s available… from clean digital to warm analog models.

Each equalization model can dramatically change the overall feel or “flavor” of the song.

This means that every song varies in the types of adjustment it needs and it can even be the case that several EQ´s (an analog EQ for that “sparkling” highend and a precicse digital EQ for exact low-end adjustments) are used for specific frequency adjustments within one song.

And this is also a reason why you need an experienced mastering engineer to help guide you in making the right EQ adjustments to your song.

About the Author:
Lorenz Vauck is an Audio Mastering Engineer, Musician, and Internet Entrepreneur from Dresden, Saxony, Germany. He is the Managing Director and Chief Mastering Engineer of XARC Mastering, one of the world’s first online audio mastering studios established in 2003.

To get equal volume just by letting the computer do it is still almost impossible nowadays and can’t replace the ears. Reason being is that the perceived volume (so what your ears tell you how loud it is) is completely different from what meters can tell you.

A simple example would be this: take a sine-wave-sound at 50Hz and one at 5000Hz that have the same RMS. Now, while the meters do show the exact volume for both, they have a very different perceived loudness to you, right? The problem is that the human ears reacts differently to different frequencies, and even then, impulse sounds, lasting for short at a high volume on the meter, can still sound lower in perceived volume than a longer sound played back at lower volume. Many factors come in so that only the human brain in connection with your ears can correctly judge what the right loudness relation between the songs is.

However, there is an ongoing attempt (and I say attempt, because it’s far from perfect) to digitally simulate and measure the perceived loudness and tell you how you have to change the volume of each song in relation so they all sound equally loud. This is derived from the ReplayGain that you might know from MP3 Players.

Here is how you go about and initial measurement and then adjusting the volumes accordingly – the results are acceptable for a first start, but still you have to control and adjust the volume of each song by ear afterwards. But it gets you in the ballpark for a good start:

1) Download this and extract it: http://www.rarewares.org/files/others/wavegain-1.2.8.zip

2) Take each of your songs and reduce it to the chorus parts (the loudest parts of the song, or better said: the parts the listener has the most attention to in the end, this usually IS the chorus, right?). Then save each of the “reduced to chorus” versions of each song to a separate WAV file (use 24bit, or better 32bit float).

3) Now put the wavegain.exe from the archive you extracted above to the directory you saved your “reduced to chorus” songs.

4) Open a command prompt and type

wavegain *

…and it will tell you how much you should adjust each song so that they get the same perceived loudness. So you now go back to your original songs and boost/lower so much dB as it tells you for each song.

Normally you can also simply do this on the “full song” instead of cutting it down to a “chorus only” version. But I found that it yields better results this way.

Still you will have to control and manually adjust afterwards, but it’s the best starting point a computer can make out of it, haven’t found any other tool that does it any better than this.

Of course the best point do to this process is prior to the final limiting stage, because if you do this on already limited songs that have no more headroom, it will not work to well, apart from that you will destroy the dithering you (hopefully) applied. So do this at the point where you are happy with the master so far but have not used the final peak limiter yet. After you have fnished this process and did the last adjustment by ears for equal loudness, you can then go to the limiter and use the same setting for each of the songs you’ve previously processed – you do want to do this, because limiting each song with separate values will of course also put off the volume-relations between the songs again.

Hope it helps and good luck:)

About the Author:
Lorenz Vauck is an Audio Mastering Engineer, Musician, and Internet Entrepreneur from Dresden, Saxony, Germany. He is the Managing Director and Chief Mastering Engineer of XARC Mastering, one of the world’s first online audio mastering studios established in 2003.