The #1 Mistake With Digital Audio Recording

digital audio recording

Everybody wants to record audio like the great engineers of history. You might read interviews and books about their warm sound and smoothly saturated tape. And then you try to apply those same ideas to your recordings.

Big Mistake

Digital audio recording is a bit different than what you read about in the history books about tape saturation and overloading your pre-amplifiers. Today, digital audio is not so lenient towards overload. Digital clipping is one of the worst sounds an audio engineer can hear. In the old days, engineers liked pushing their levels to saturation to get that warm tape sound. But now, the only thing you accomplish by pushing digital audio to the max is horrible digital clipping.

No Clipping Please

correct digital levelsNow, with digital audio recording we have to record at an optimum level without distorting our preamps, causing that horrible clipping. When you are getting levels into your DAW, make sure you are getting a good enough level to record with before it clips. A nice rule of thumb is to record the loudest part of the signal at around 3/4 before you reach 0dB. If your meter goes from green to RED, then somewhere in the middle of the orange part should be a nice enough level.

So when setting levels, make sure your audio is being recorded like in the area shown in the picture. Not too quiet so that your audio is too close to the noise-floor, but not so loud as to cause clipping.

24 Bit Recording

Also, make sure you are recording at 24 bit since it will give you much more headroom than recording at 16 bit. With 24 bit recording the signal to noise ratio is much higher and you can effectively record at lower levels without introducing too much noise into your signal.

Conclusion

By setting your levels correctly you can avoid a few problems in the mixing phase. You avoid that pesky digital clipping that ruins your audio completely, and by recording at 24 bit you have enough volume and headroom to play around with without the noise floor posing a problem to your recordings. Keeping simple things like this in mind helps smooth out any kinks you might have later. Get it great without digital clipping and your audio will be saved.

Image by: Mr Guep

About the Author

Björgvin Benediktsson is an Icelandic born musician, audio engineer and writer. He has worked in the audio industry since 2006 and is an SAE Alumni from the SAE Institute. He is the man behind Audio Issues. His latest ebook Mixing Strategies tackles the all important aspects of mixing music. Learn more about him and Audio Issues here

7 Comments on "The #1 Mistake With Digital Audio Recording"

  1. Peter Juul Kristensen March 13, 2011 at 12:38 am · Reply

    Hi all,

    One thing that I think Björgvin left out of this (short) article is to to record everything in 88.2 kHz or higher. When operating in the digital realm all your internal processing (simply put) just sounds better and more detailed in the higher sampling rates. I even find that projects recorded in 44.1k or 48k can benefit during mixing to be upsampled to 88.2k or 96k during the mixing process – even though they were only recorded at the lower rates. On a few projects I’ve been able to set my foot down and have everything recorded AND mixed in 88.2k – with great results.

    Stay in tune!

    Peter

    • Björgvin Benediktsson March 14, 2011 at 9:29 am · Reply

      Hey Peter, thanks for the comments. While I do agree that higher sample rates can make everything sound better some people trade off the disk space it requires. Since hard disk space gets cheaper and cheaper it’s not that big of a deal but if somebody that doesn’t have all that space to spare might need to settle for 44.1 or 48.

  2. Peter Juul Kristensen March 14, 2011 at 1:32 pm · Reply

    @ Björgvin:

    Yes, the files are bigger, but HEY!: Hard drives are getting bigger and cheaper every second it seems, as you rightfully point out, so the real question is: Do you have the internal/external DSP power to process 88.2, 96 or even 192kHz audio?

    • Björgvin Benediktsson March 14, 2011 at 1:37 pm · Reply

      That’s exactly the other question! You really need a lot of processing power to mix a 30 track song at 192kHz, so that is also a hindrance to budget recordists.

      It’s about compromise I suppose. The best sound without making you feel like you’re working on a 486 again…

  3. Sean May 3, 2012 at 9:59 pm · Reply

    Samplerates are a whole other bag of worms… it can be debated all night long and has been hundreds of times. While in theory its more information and higher resolution, theres a lot of debate as to the extent humans can hear a difference. My 3.2ghz quad core gets pushed hard already with a full session at 48k so until processor capabilities are increased to easily handle it I think most people with stick with lower rates. Great post and site man.

Trackbacks for this post

  1. Digital Audio Recording – The Nr. 1 Mistake | Audio Issues | recordingdigitalaudio.com
  2. 20 Steps to Becoming a Better Audio Engineer | Audio Issues

Leave a Comment