JUSTIFY MY BITS
As analog circuitry is the weakest link in the recording chain, without good equipment such as quality preamps, mics, and even clean sources of power, the benefits of a 24-bit recording will be compromised. Unbalanced analog connections are susceptible to inducing noise from the environment, and can also promote the creation of ground loops. Also, running a signal with an unbalanced stage in it will likely negate the benefits of 24-bit recording by lowering the signal-to-noise ratio of the total analog stage; devices with unbalanced outputs, that can’t provide levels above +12dBu, typically do not offer great enough dynamic range to justify 24-bit A/D conversion.
It’s kind of ironic that sometimes, using good gear at 24 bits will actually sound “cheaper,” as your recordings will reproduce unwanted noise and crud from the environment far more faithfully. So be it. Just don’t take your best mics to all of your gigs; if there are noisy environments, leave the U-47s at home.
Of all the reasons we may have to record at 24 bits, headroom is one of the greatest justifications. As every bit offers approximately 6dB of dynamic range, with 16 bits we achieve 96dB and with 24 bits, we go all the way up to 144dB. A recording with peaks hitting –12dB maximum will still take advantage of a theoretical 22 bits’ worth of dynamic range and resolution. The result: greater quality for lower level signals, without any clipping.
Once you realize that there are some technical considerations involved in taking full advantage of those extra 8 bits, and that a 24-bit recording takes more space on your hard disk (see Figure 1), I still believe it’s well worth recording with 24-bit resolution. You can record at lower levels, quiet passages won’t be struggling to stay above your system’s noise floor, and you’ll give your converters some breathing room. Not a bad deal . . . now let’s move on to selecting a sample rate.
WHAT ABOUT SAMPLING RATES?
It makes no sense to record a politician’s three-hour speech, or an interview with a wasted rock star for the local newspaper, at 24/96 just to transcribe it to print later on. Built-in microphones on your portable solid-state recorder will do the trick, and recording at 128mbps MP3 will save tons of space on your memory card.
Some instruments with delicate high frequency content may benefit from being recorded at 88.2 or 96kHz instead of 48kHz, but the differences will be subtle. It’s often better to use a higher-quality mic/preamp combination and record at 44.1 or 48kHz than record at 96kHz with your interface’s onboard preamps.
Without getting sucked in to those endless discussions about whether an average human can really hear — or even care about — the difference between a 44.1 or 96kHz sampling rate, we can probably all agree that our choice of sampling rate will depend on the delivery media of our recorded material. If it is audio for commercial CDs or data-compressed formats (e.g., MP3, WMA, AAC, ATRAC), go for 24/44.1 or 24/48 (what I use). Some audio for video editors require 48kHz audio, and DVD-Audio (as an example) demands 24-bit/96kHz audio files, so that may also help you to decide.
I believe that the difference between 16 bits and 24 bits is more significant than the difference between 48 and 96kHz recordings, and given the required extra space and computer resources, sometimes it just does not justify going to a higher sample rate.
Lower sample rates also require fewer computer resources and less recording space per minute, although one tradeoff is higher latencies. Storage media is getting dirt cheap nowadays, and we can get portable disks with over 250GB for peanuts. However, saving some computing resources could mean being able to record/edit more simultaneous channels in your computer and add a couple more of CPU-intensive plug-ins; even better, if you’re using an audio interface with ADAT ports, you could get twice the channels at 44.1/48kHz if you do not give in to the temptation to go all the way up to 24/96 just because you can.
By knowing in advance what you are going to do, you could save some time on setup, bring less equipment for live recordings, save some computer resources, and get higher track counts — yet still achieve professional results.
When not touring as the frontman of the band WoM (www.wom.com.mx), Gus Lozada hosts conferences and clinics in Latin America about music production. He also moderates two forums at Harmony Central, “Nuestro Foro” (Spanish language) and KSS - Keys, Samplers and Synths. His e-mail is firstname.lastname@example.org.