Recording with consumer soundcards

When consumers record from analogue sources on their computer, the sound quality of the soundcard may be forgotten. Five years ago, a soundcard was purchased as a elementary step-up from Chinese soundcards integrated into a computer’s motherboard. Whilst it is rarely used, let’s crudely examine its relative quality.

My sound card

This old Yamaha DS-XG YMF724 ($60) soundcard uses the SigmaTel STAC9704 / STAC9707 integrated circuit which yields a Signal-to-Noise Ratio (SNR) of typically 87 decibels (A-weighted). The minimum figure is 75 decibels. Total Harmonic Distortion (THD) is 0.02 percent. These measurements are made when the soundcard is being used with an analogue line-level input using its Analogue to Digital (A/D) converter. The accuracy of the A/D conversion is paramount whether one uses a portable digital recorder or a soundcard.

This soundcard will be compared to a more expensive M-Audio Audiophile 2496 soundcard (pictured below), based on the AK4528 integrated circuit. The M-Audio card offers a 95 decibel SNR and THD of 0.001 percent. This card costs $108, as does the Zoom H1 portable recorder with uncompressed recording from line-in (or microphone) onto a microSD card! For that sort of dough, the logical choice seems clear.

Premium sound card

How important are these specifications? According to electrical engineer Richard Kuo:

A common way to measure the “cleanliness” of the output of a card is the SNR… SNR is measured in decibels (dB) and the higher the number, the cleaner or better the signal is. Above 90 dB is generally quite good and below 70 dB is getting down there. SNR measurements are usually only useful if you can make sure that all the measurements were done identically. Unfortunately, that’s rarely the case with sound cards. SNR figures from the manufacturer should probably be considered suspect. Your best bet is to figure out what the popular opinion about the soundcard you are considering is.

According to computer scientist Dr Qingkai Ma, inferior cards have a SNR of “30 to 50 decibels” whilst the “High-end range [is] 96 to 100 decibels”. He recommends Audio Win Bench to benchmark a soundcard.

Resampling chart

Unfortunately, the Yamaha soundcard resamples to 48 kHz. Digital artifacts (arguably as problematic as analogue distortion) can be a byproduct of poor resampling. On this card, analogue recordings must be resampled from the fixed 48 kHz to 44.1 kHz (used in the compact disc) for the recording. The Professional Music PC website explains this phenomenon in layman’s terms:

The audio codec or engine of the card is “rate locked” to 48khz. This means that the card resamples everything to 48khz for the dsp/codec chip on the soundcard. This is a big limitation in a card because it means if your recording in 44.1 the audio is at least resampled to 48khz once on the way and and another one the way out through the card to your ears. This also applies to the digital IO as well as the analog IO, in fact the digital output on any [Creative Labs’ Sound Blaster] or ac97 card will be rate locked to 48khz meaning its impossible to output a digital signal to a stand alone cd burner at 44.1 or [Digital Audio Tape] machine or many other products which may need 44.1khz spdif.

There’s many other downsides to this resampling besides just a downgrading of the audio’s quality. Some people complain of timing problems in some software ie overdubs drift out of time and sync errors are just a few. Cards which resample generally speaking have a much higher CPU load on the computer as bus mastering is not used with these cards.

Legacy sound card

iXBT Labs is a Russia-based website dedicated to reviewing all kinds of computer hardware. Tester Maksim Lyadov revealed good audio performance with a substantially-identical OEM variant of this card, the Aopen AW754. This suggests the Yamaha consumer soundcard may provide satisfactory performance.

Whilst no primary reference is provided, Wikipedia contributors also contend that the Yamaha DS-XG YMF724 soundcard was not of poor quality:

Relative performance was good despite the typical low cost. The cards were usually equipped with good quality 18-bit Digital-to-Analogue Converters, providing similar low noise and harmonic distortion levels to those found in semi-professional hardware.

From a casual comparison of the specifications alone, there seems little doubt that audible differences from the original source will creep into the recordings undertaken on any consumer grade soundcard.

FM enthusiasts will be acutely aware of the amount of junk any computer spills into the FM band! Potentially, this flood of noise has an effect on the soundcard too. According to Richard Kuo:

The fact is, your PC is an incredibly hostile environment for a soundcard to be in. All that electromagnetic noise is bad news for a soundcard just trying to get a clean signal in or out of your system. Basically, what this all means is that in order for a soundcard to be true to the original sound, it not only has to have good hardware, it also has to be properly shielded.


In closing, the solution seems simple. Avail yourself of the above medication and you will have full confidence in your recordings! Hmmm… If only life was that easy…

Comments are closed.

%d bloggers like this: