I’m mastering a project, and I’m confused about dithering. How and when should you apply it? — Matt, Elkhart, IN

Apply dithering only when you reduce the bit depth (word length). For example, suppose you are recording a multitrack project at 24 bits. Keep dither turned off during the entire project until the last step, when you’re going to export the mastered songs to a 16-bit wave file to put on a CD. So record at 24 bits, master at 24 bits, then turn on dither and export (bounce) the mastered program to a 16-bit/44.1 kHz wave file so that it will play on a CD.

 

Should I use compression during mastering to make my CD hotter? — Scott, Atlanta, GA

Gentle compression on the stereo master bus, plus makeup gain, can make the CD sound louder. But it’s also more fatiguing to listen to, and more distorted. I don’t like the sound of it myself, but it’s a pretty common practice. I’d rather compress individual instruments or vocals that need it during mixdown. I prefer to use only peak limiting and normalization during mastering. You might limit the peaks up to 7 dB; more than that starts to sound distorted. Then normalize the program (raise the gain) so that the highest peak in the mix reaches a “ceiling” of about -0.3 dBFS. Why that level, and not 0 dBFS? Well, digital meters in DAWs read sample values, not the actual analog signal level at the output of your audio interface. Some signals have intersample peaks that exceed 0 dBFS when the meter reads 0 dBFS maximum. So aiming for -0.3 dBFS prevents accidental overs or clipping in the analog signal coming out of your audio interface, or coming out of the customer’s CD player. However, some engineers allow the mastered signal to clip occasionally on short drum hits. That’s because clipping is less audible if it’s of short duration. Again, I don’t like the sound of clipping but some engineers are asked to create a hot CD no matter what. It’s those crazy loudness wars.

Where should the master fader be set? I’ve heard all the way up, 3/4 up, or halfway up. — Nathan, Fresno CA.

Generally the best setting for the master fader is about 3/4 up, at “design center” (the shaded portion of fader travel). That results in the best compromise between noise and distortion in your mixer or DAW. If you set the master fader very low, you’ll have to turn up the channel faders to get enough recording level, but then the signals from the channel faders might be so high that they cause distortion or clipping. If you set the master fader all the way up, you’ll turn down the channel faders to compensate, which sends a weak to the raised master fader and increases noise. Try to keep the master fader at design center, then adjust all the channel faders up or down by the same amount to get a normal recording level. Note: Some DAWs use 64-bit floating-point math, which lets you trim the master fader setting a few dB without introducing any noise or distortion.

Have questions about recording techniques and equipment? Just register for the L2Pnetwork at the top of this page and you’ll be able to post questions directly to Bruce. Any subject is fair game — from miking to mixing, from direct boxes to digital audio. If I can’t answer your question, I’ll point you to someone who can. And make sure to join L2PBandSpace to share your thoughts and experiences with fellow musicians. 

Originally posted 2009-02-01 23:15:53.