Power Supplies and Computer Audio: Do they make a difference?

No matter how good, how expensive, or how many claims of magic the back end components in your audio system may boast, the best signal they will ever see is what comes from your source.

Ideally one wants to be in a position of having a perfect source so that the rest of the system can focus on maintaining this perfection. The truth is, in most cases, the audio system spends more time attempting to correct for a bad source that passing the signal unharmed.

We have filters, buffers, DSPs, FPGAs, and other programmable interfaces that allow the system to supposedly correct errors. Here’s my question… how does this device know there’s an error? The simple answer is that it’s based on algorithms, or made up math/guesswork.

The truth is that no DSP, filter, or other device can correct errors in audio. All it can do is distort it. Sometimes this distortion is attractive, but at the end of the day it is still distortion. For those seeking the very best sound the only solution is to remove that distortion rather than covering it up.

“But but but… this company says their device is bit perfect and corrects for errors between the input and output”.  Rather than arguing with you, let me just tell you how the audio system works. It’s really not as complicated as the gurus may want you to think.

Every signal in the audio system is created by the power coming from the wall. It doesn’t matter if you’re running a computer, turntable, amplifier, or have a wet chipmunk plugged into the wall, every signal is created by this power.

Think of the power coming from the wall as the source for your source. And every component in your system is just a copy machine. But instead of making exact copies it is making altered copies. Copies that are larger, smaller, or different pictures entirely.

Let’s break it down a little further with an analogy. If you think of your system as a copy machine, it’s just making copies of copies of copies. That means your system is not taking the original file and duplicating it, rather it is taking the file it receives and making a copy each time. To make this copy, a bit of voltage is applied to generate the ones and zeros or to make the signal larger or smaller.  Have you ever tried to make a copy of a copy of a copy? Each time you do this the ink gets greyer and greyer until you have very little left to copy.

Think of the power coming from the wall as the ink. The cleaner and faster the power, the blacker the ink will be. So each time a copy is made it is a much more accurate copy or enlargement.

With a computer there is always the argument that there’s error correction. This is true for data stored on a hard drive or for software applications themselves. If there were errors, the operating system would crash. These processes are sent in packets, which are buffered and checked for errors. If there are any errors the data is just resent. With audio it’s a little different. Because audio is in real time there is no error correction that can be done.

Some companies will mention integer mode and memory playback. These are great functions that can improve the sound in the computer, but all they are doing is allowing the FIRST copy to be a better copy.  Once the data is processed and the data leaves that piece of software there is no error correction that can be done. The signal is left up to the mercy of the software processing it, the operating system, and the power used to create each new signal.

At every stage in the computer thousands of copies are being made. You have a copy made to take the data off the hard drive, a copy made as the data is moved into a piece of software, copies made by the software every time a dial is turned, every DSP plugin makes a copy, and every time the signal changes pieces of software more copies are made all the way up to the output of the computer. Each copy requires a certain amount of voltage from the power supply.

Computers as a source make more copies of the audio signal than any other device in the system. Unfortunately, they too are plagued with bad power supply noise. It is for this reason that they can be so significantly upgraded by using a better power supply.

Even amplifiers and preamps are hugely affected by the power coming from the wall. Each gain stage a copy is being made and an additional amount of voltage is applied to make the signals larger. For resolution and dynamics the power supply must not only be clean, but extremely fast in order to react more quickly than the demands of the amplifier circuits.

A cheap component with a killer power supply will always outperform a high-end component with a mediocre power supply. The better the power the less distortion and noise the components will see. If they see less noise and distortion they don’t have to work as hard to make accurate copies. Better in equals better out. It’s that simple.


Leave a Reply