Don't confuse recording level with the levels of your final production. It's best to leave headroom* and adjust the levels later.
VU meters were a good compromise.
With digital recording a peak meter is much more useful. (I'm sure most analog recording engineers would have loved to have a peak meter and a VU meter.)
0dBFS (zero decibels full-scale) is the "digital maximum". It's as "high as you can count" with a given number of bits. The numbers are bigger in a 24-bit file than in an 8-bit file, but everything is scaled during recording/playback so a 0dB 24-bit file is not louder than a 0dB 8-bit file.
Analog-to-digital converters, digital-to-analog converters, "regular" (integer) WAV files, and CDs are
all hard-limited to 0dB and if you try to go over you'll get clipping (distortion).
Analog tape tends to "soft clip" as you go over 0dB and it was normal to go occasionally "into the red". Digital is unforgiving.
REAPER (and most audio software) uses floating-point internally so there is virtually no upper or lower limit and REAPER itself won't clip.
Quote:
|
I am working on a recording project now that would benefit from having consistent levels in the final product.
|
Are you trying to match the perceived loudness of multiple recordings?
The easiest thing is to
Normalize for "maximized" 0dB peaks.** It's a simple mathematical process and
you don't need meters to normalize.
Most commercial CDs/recordings are 0dB normalized (and compressed and limited to beat-out everybody else in the
Loudness War).
But, peak levels don't correlate well with perceived loudness.
Or, you can scan and adjust the RMS levels. RMS is a better approximation of "loudness" and if the recordings are similar, matching the RMS levels should match the loudness.
Just as an example, the
audiobook specifications call for an RMS level between -23 and -18dB with -3dB peaks.
Or, you can get an
LUFS plug-in to measure the loudness (and then adjust).
If you adjust for a target RMS or LUFS level, you have to be careful not to clip the peaks.
A plug-in that
analyzes the peak, RMS or LUFS of the entire file is more useful than watching a meter.
* Headroom is a funny thing... If you need headroom it's not headroom anymore, but you don't know until after you're done recording. With "live" recording the peaks are usually unpredictable so you should leave lots of headroom. If you are digitizing analog recordings the levels are pretty-well defined so you don't need as much headroom. But as has been said, you've got a TON of dynamic range with digital recording so there's no reason to record "hot" and there's nothing wrong with leaving lots of "unused" headroom.
** Some people worry about "inter-sample overs" and they normalize to around -1dB. I don't worry about it. There is no inter-sample information in the digital audio data so any overs would only exist on the analog-side of the DAC, and there is nothing stopping the analog-side from going over 0dB.
Also MP3 compression will change the wave shape making some peaks higher and some peaks lower so it's not unusual for the MP3 to go over 0dB, so again some people like to normalize to around -1dB to allow for that. MP3s can go over 0dB without clipping, but you can clip your DAC if you play the MP3 at "full digital volume". But again, it's not something I worry about. I've never heard of a case where that slight clipping was audible, and if you are making an MP3 you've already accepted that it's imperfect.
Setting recording levels for consistency
Linear Mode
