Is your audio interface fast enough?
Q: Can I use a low-pass filter to remove noise from my recording?
Your mix sounds good in your car. But does it sound good in ANY car?
Setting microphone preamplifier gain to achieve both adequate headroom and a good signal-to-noise ratio
Recording acoustic guitar in stereo - should you use spaced or coincident mics?
The new battlefield in the loudness war?
Three types of musician you'll prefer to work with in the studio, and one type that you won't
How to record or amplify the melodica or any unfamiliar instrument
A brief introduction to soundproofing
Demonstrating the Waves J37 analog tape emulation plug-in and comparison with a real tape recorder
Subscribe to access our latest, up-to-the-minute articles with hints, tips and adventures in audio in the weekly Audio Masterclass Newsletter.
Let's take the compact disc as an example, simply because it is so common. And we all aspire to get our music onto CD and in the shops, don't we?
'Pure, perfect sound, forever'. That was the slogan when CD was first introduced. But it wasn't so perfect, and still isn't.
Every CD ever made has errors in the digital data it contains. And even if it didn't have errors when it was new, it will soon have some when it is scratched and fingerprinted.
But CD has a powerful error handling system. Any error handling system consists of two components. The first detects when an error has occurred. There is additional information on the disc that is used to identify whether the data that is retrieved from moment to moment is good data or bad data.
The second part of the error handling system decides what to do about the error.
If the error is small and doesn't affect much data, then it can be corrected. There is additional redundant data on the disc from which the digital signal can be completely reconstructed. So after error correction, the signal is exactly the same as it should have been, had there been no error.
But if the error takes out a larger chunk of data, then some of the redundant data might be lost too.
In this case the error is concealed. This means that the system looks at data surrounding the error, and guesses what information the lost data contained. Clearly the digital signal is not likely to be the same as before. Merely similar. This is one of the reasons why CD players sound different to each other - how well they conceal errors.
If the error is really bad, then there is a third and final option, and that is to mute the output. Better to mute the output than deliver random data to the loudspeakers, which could be very high in level and possible take out the tweeter. Unfortunately some CD players don't mute as effectively as they should, as I'm sure you know.
It is possible to take a cynical view of error correction and say that it allows CD factories to turn out substandard product.
CD-ROM uses error correction too - more powerful because although audio might be amenable to error concealment, computer software - for instance - is not.
DVD and digital television both rely on sophisticated correction and concealment. Digital television is in a state of error pretty much all of the time and is designed to degrade gracefully as errors increase.
Hard disks suffer from errors too. But since hard disks are used for all kinds of purposes including storing software and financial data, errors are completely corrected. When we start using error concealment on hard disk data, interesting things might happen...