So now that we understand that it's important to compare sounds at consistent playback levels, and that simply adding more effects without adjusting playback for the additional signal level can be deceiving, the obvious question is: how loud to monitor?
For people of a technical bent, the first answer is 83dB SPL (but hold your guns). SPL means "sound pressure level," meaning the actual air pressure of the moving sound waves. There is no way to measure it in within reaper or any other software, you can only measure it in open air, after the sound has left the speakers.
83dB SPL is right about where human hearing is most linear. It is about as loud as city traffic, or a noisy restaurant. Alarm clocks are supposed to ring at 83dB. THX movie mixes are supposed to be calibrated with an average speech level of 83dB SPL, somewhat louder than typical conversation in a quiet room. 83dB sounds "loud," but not painful. OSHA requires no more than 8 hours continuous exposure to 83dB for workplace hearing safety, so it's right on the cusp of where you could spend a full work day without hearing damage. The legendary Bob Katz recommends that mastering engineers master music recordings at an average level of 83dB (actually, he recommends mastering at comfortable levels with a system calibrated to have a certain amount of fixed headroom above 83dB playback, but that's getting ahead of ourselves).
As it happens, 83dB is not only where hearing is most linear, it is also right about the average level where average listeners tend to set the playback volume when listening to music on a capable system. Just before "too loud." (what a coincidence!)
So, 83dB seems like an obvious level for monitoring, but not so fast, partner!
Remember what we said above, that louder always sounds better. We can make this rule work for us as well. As it happens, almost anything that sounds good quiet will sound even better loud, but the reverse is emphatically not true. Cranking up the playback speakers (or just adding more gain with piled-on effects) makes shitty mixes sound great. By the same token, turning something down makes it sound worse.
This effect is especially brutal on live recordings of metal and hard rock bands. When you're standing in the crowd, and hearing a roaring 110dB that shakes your bones and pierces your ears, the effect is massive. But when you record that sound and play it back at workplace-background levels, the huge guitar sounds like nasal fizz, the furious double-kick turns to mushy paper, the churning bass becomes clackety mud, and the screaming singer sounds wimpy and shrill. These kinds of acts require a lot of tricks and psycho-acoustical funny business to achieve the right effect of power and loudness WITHOUT the actual power and loudness (more later).
But the same principles apply to anything. If you want your recording to sound right to every listener, then you cannot rely on high-quality 83dB playback every time. Your records are (hopefully) going to be heard in noisy cars and bars, on crappy speakers at 50dB in shopping malls, and so on. Unless you want them to sound wimpy and limp, it is really important to make sure that they sound good even in worst-case scenarios, because that is often where they will be heard.
So there is a really good case to be made for monitoring at very quiet levels as much as possible. In fact, I think it is safe to say that a majority of commercial mix engineers do a majority of their work at conversation-level or below, occasionally turning up the volume to check the lows and the balances at higher playback volume.
Monitoring at quiet levels has another practical advantage. Even before we hit the levels of hearing damage, our ears get desensitized by loud sound. Listening to 83dB for extended periods is like being in bright sunlight-- it's hard to see when you walk indoors. Keeping the lights dim allows you to occasionally focus spotlights where you need to check detail without dulling your overall vision. So it is with sound.
If you can create recordings that sound good at very quiet playback levels on decent nearfield monitors, they are almost guaranteed to sound better or at least as good in any other circumstances, including headphones and louder systems. But of course, it's always easy to double-check by putting on some headphones or cranking the volume for a few seconds.
There are a lot of schools of thought, but if you haven't already done so, I would encourage you to try recording and mixing at very quiet levels, and see if you don't start making better decisions, and generally better recordings.
Last edited by yep; 12-03-2008 at 08:44 PM.