M-Audio Torq 2.0 [118/171] Sample rate

Chapter 12: Torq Preferences

113

Torq 2.0 User Guide

Hardware

Audio Device

This preference lets you choose which audio interface to use with Torq from a list of properly installed and

connected devices.

Sample Rate

The sample rate determines sound quality of Torq depending on the source material you are playing. This set-

ting lets you choose a sample rate of either 44,100 (default) or 48,000. Although the higher sample rate offer

better sound quality it does place heavier demands on your computer. It is recommended to sample the sam-

ple rate only as high as you need it to be. If you only play MP3 files or songs from CD's, you should leave the

sample rate as 44,100 since this is the sample rate used by MP3s and CD's.

HW Buffer Size (in Samples)

This parameter can cause a lot of confusion for some users, but is actually not that complicated. To understand

how the buffer size affects the performance of Torq, you must understand a little bit about how your computer

processes audio.

Multi-tasking is a term that refers to doing multiple jobs all at once. It’s what allows your computer to run

more than one program at a time (i.e. listening to iTunes while surfing the Web). While it looks like the com-

puter is doing two separate things at once, it’s actually not—it’s still doing only one task at a time, but chang-

ing between tasks faster than you can see.

This provides for a streamlined computing experience, but creates a problem when using audio applications.

Audio is non-stop—a 5-minute song will play for 5 minutes without interruption. So how can the computer

keep audio playing while it’s jumping around to do other tasks? The answer is buffering. An audio buffer is a

temporary “storage tank” that can hold a brief moment of audio. The computer will fill the audio buffer with

music then let the buffer play while it does other things (like update the clock on your screen, check your net-

work connections, monitor RAM usage, etc.). When all things work properly, the computer will complete its

other tasks and fill the buffer with more data before the buffer empties, thus resulting in perfect audio while

multi-tasking.

If the audio buffer happens to empty before the computer can fill it with more data, the audio playback will

stop until the computer can fill the buffer again. These “dropouts” happen very quickly, not sounding like

prolonged gaps of silence, but sounding more like clicks and pops or otherwise distorted audio. When this

happens, the solution is to either lighten the CPU load (by closing unnecessary applications or processes that

are wasting the computer’s time) or by increasing the size of the audio buffer, allowing it to play longer (thus

giving the computer enough time to perform its other tasks).

So why not just go with a large buffer size and avoid dropouts? The problem is that increasing the buffer size

increases the system latency. Latency is the time between when you tell the computer to do something (such

as activating an EQ Kill) and when you actually hear the results from the speakers. If you have a large audio

buffer, the buffer will have to play out its entire contents before you’ll hear any new EQ changes in the audio.

When DJing, this can be a nightmare if you’re trying to do things with accurate timing—all of your actions

will have a delayed effect on the music.