I’m working on a 24-bit multitrack recording in my DAW. I imported some older recordings — 16-bit wave files — into that 24-bit project. Does that degrade the sound quality? Also, when should I apply dither to get the best sound?
– Matt, Elkhart, IN
You should switch on dither when, and only when, you are exporting a 24-bit (or 32-bit) recording to a 16-bit wave or aiff file to burn onto a CD. Dither is very low-level noise added to the audio signal. It helps the signal retain most of its 24-bit sound quality when the signal is converted to 16-bit.
What happens when you import some 16-bit audio files into a 24-bit project? Your DAW software will convert the 16-bit files into 24-bit simply by adding eight zeros to the end of each sample. So even though those 16-bit recordings have been converted to 24-bit, they don’t sound any better, because those last 8 bits aren’t adding any more information. Those 16-bit files sound the same after importing as they did originally.
Now, if you want to export that multitrack recording to a 16-bit wave file for burning onto a CD, add dither so that the 24-bit mix will stay high quality when it is truncated to 16-bits. “Truncated” means the last 8 bits are cut off to change 24-bit samples to 16.
Adding dither to the 16-bit tracks in that project won’t hurt anything — it will just add some very quiet noise that you can’t hear.
Many DAW programs can import all sorts of audio files, even data-compressed ones such as MP3 and AAC. While importing, they get converted to wave or aiff files at the bit depth of the project.
Let’s explain a few terms. When analog audio is converted to digital, the analog signal is sampled or measured several thousand times a second. Each measurement or sample has a certain degree of accuracy called “bit depth”. That’s the number of bits per sample: 16, 24 or 32 bits. The higher the bit depth, the greater the accuracy and the less the distortion. “Bit depth” is also called “word length”.
“Sample rate” or “sampling frequency” is a measure of how rapidly the analog signal is sampled or measured, such as 44.1 kHz, 48 kHz, or 96 kHz.
The digital audio format for a CD is 16-bit/44.1 kHz. That means, when the analog signal is converted to digital to burn on a CD, the analog signal is measured 44,100 times a second, and each measurement or sample “word” is 16-bits long. A 24/96 A-to-D conversion is 24 bits and 96 kHz.
The higher the bit depth of a signal, the less distortion you hear. The higher the sample rate, the higher is the frequency response of the signal.
# # #
Bruce Bartlett is a recording engineer and microphone engineer (www.bartlettaudio.com). He is the author of “Practical Recording Techniques 6th Ed.” and “Recording Music On Location.” If you have a question for Bruce, please send it to the editor at L2PNet.com.
Originally posted 2009-11-07 21:18:52.