buffer problem

I am new to n-track and I am hoping to use a portable laptop (pentium III, 1000 Mhz) to record some things on. I am using the laptop because it is nice and quiet. I have a 3.6 Mhz machine but it is noisy as ####.

Anyway for the laptop I can’t seem to get the buffer size right. Either I can hear a delay or I can hear static and noise. Is the buffer size a function of the machine, or is it something else? I am worried about the delay once I start laying down tracks, am I going to get a delay in the playback and live track?


I’d set it to use the wdm drivers, crank the buffers up, and use the red button to start recording, and don’t press the “live” button.

But how can you multi-track without using the Live button?

You select which tracks to record either from the timeline, or the recording VU. There’s a red arming button that when you click on, lets you select stereo, mono, or disabled.

As a general rule you don’t need the Live button at all unless you need to monitor with effects (reverb, distortion).


Modern computers are more than fast enough to handle audio data in “real-time” if that was all they were doing. Unfortunately that is not all that they do. At the very least they are drawing the screen and doing some basic housekeeping in the background. If you are running multiple programs in the background (anti-virus, firewalls, IM, spyware etc.) there may be processes which interfere with the audio data. Buffers exist to temporarily store the audio data so that if another process takes over for a while, the data is not interupted. Data flows into a recording buffer at a uniform rate but may be pulled out to store to disk in bursts. The bursts occur at a much faster rate and as long as the DAW program can empty the buffer to disk before it fills up no data is lost. If the interfering proocess takes too long or the buffer is too small, the buffer has to either stop accepting new data or throw away old data, resulting in noise or drop-outs. A similar process happens in playback except that the buffer is loaded in bursts and emptied at a steady rate.

Large buffers reduce the chance of a buffer “over-run” where data is lost but can increase delay. When you use the “Live” function you are listening to the output of the buffer with all of the delay. It is better practice to monitor using an external mixer or other more direct path unless you need to listen to the effect of a plug-in to play properly (a guitar distortion plug-in for instance). If this is the case you need to eliminate as many background processes as you can to allow you to reduce the buffer size.

Many people focus on reducing latency but I choose not to make a hobby out of tweaking my DAW and use large buffers. The basic latency due to buffers and plug-ins is compensated for by the software so everything stays synchronized up to a point. The software has no way of knowing what the “external” delays are in the soundcard and your environment so it will always be necessary to do some “manual” adjustment if you want “perfect” synchronization. With very low latency you may be able to avoid adjustment but there will always be some lag compared with the first recorded track.

Anyway, the first requirement is that you get good audio onto the disk so adjust the buffers to the point where it works and/or eliminate background programs. A “loop-back” test where you re-record a previously recorded track by playing it with the soundcard output connected to the soundcard input will let you identify the amount of latency you need to compensate for on overdubbed tracks.


When I run into this problem, the first thing I do is defrag the drive. Of course, as Jimbob says, all unnecessary programs should be shut down religiously before any recording session.

Willy has the answer. Don’t use LIVE mode for normal recording. That way you don’t even have to contemplate latency.

You need to setup your soundcard or external mixer so you can monitor directly without having to monitor via software.