(Editor’s Note: This informal article by Sam provides useful advice on hard disk recordings of live performances. For more info and advice, go to Sam’s World forum.)

My interest in A/D converters and recording is primarily focused on measurement purposes with SIA SMAART , and the tapes I make that are almost exclusively from matrix outputs of mixing consoles mixed with live microphones and multi-channel recordings stored in various multi-track formats.

To these ends I have been interested in computer storage of audio and portable A/D converters/sound cards. So here are some of the tests I’ve done with various sound cards:

1.) I ran a sine wave into 32 channels of my computer in order to test several issues, including the number of bits toggled, the ability to continuously write to a disk, accuracy, and ability of computers to multi-task while writing audio. I found that after filling two Gigabytes, most Windows applications get funky.

There are “workarounds” for this, but most applications appear to want you to limit file size to two Gigs. I used Mathcad and some custom routines to test the accuracy of the sine waves, comparing the two channels looking for glitches, and noise levels. Simple mathematics can be used for these tests, and I was happily surprised at what I found.

2.) To test digital accuracy, I used Mathcad to write 16- and 24-bit test signals, specifically, using Sine Waves, Swept Sine Waves and Maximal Length Sequences functions. This mathematically generated data was stored as .WAV files and used various sound cards to output from one machine to another.

I then ran mathematical routines on the transfer files to see what level of errors were generated. The results were STUNNING in the low number of errors, on order of a few bits per 10 hours.

3.) The company I work with (Walters-Storyk Design Group) provides staff to a summer concert series here in New York City. We ran direct-to-disk recordings on many shows (with permission only) and in subsequent listening, found very few errors. Having now done more than 200 hours of live recording, only once was a recording trashed, and this only when some idiot reached into the FOH area and pulled cables out of my rig.

4.) For me, the big issues regarding A/D converters have always been jitter (which is simply described as the time variance between samples), quantization techniques and dither/noise shaping. In my opinion, these factors account for most of the sonic differences between converters. And, they are all less easily measured, as they require very, very stable measurement systems, such as Audio Precision System I or II, in order to see the noise levels and spectral differences.

5.) I use Wavelab software as my two-track recording / mastering platform. It was recommended to me by my friend Don Pearson (one of the best live audio system engineers ever). It is good program with TERRIBLE support and documentation. It does the basics very well, but more complex features are harder to use without the provision of tools for further understanding of HOW to use them!