Decibels, loudness, and mixing

B

BrentDomann

Has a Dedicated Member.
So it is said (argued, with data) that a change of +3 decibels means a sound is physically twice as loud as before the change. +10 decibels is 10 times as loud (but is perceived as "twice" as loud). Many cite 3dB as the smallest increment that the average person can perceive.

My question is this: in your experience, does that relationship maintain itself in a mix? Or is this only the case with a solitary sound?

My experience has been that in a dense mix, 3dB is huge. I will spend three days fighting over 0.5dB with audible differences between settings. This appears even more so when using buss compression in mastering.
 
apparent loudness does not equal actual loudness.

this is where studying RMS values and Crest factors come into play.

best thing to do:
don't wear out your listeners ears, unless you only want to last one song.
 
Three dB is generally noticeable to the untrained ear. For us doing the mixing it's a huge change. I can hear a 0.1dB difference when I'm the one at the controls tweaking the mix.
 
Hi,

Don't mind those "general rules", they seldom are of any practical use when doing the creative side of work (mixing, recording etc). Regarding dB there are so many ways to measure, but I guess you mean db FS (fullscale)?

/MelkerMusic
 
If you're talking about the marks on your meters and faders, then you're talking about voltage, not power, and so it's actually 6db that equals (actually, just really close to) a factor of two, and 20db is a factor of 10, and I think it's 1db that's considered the threshold of perception.

When talking about ratios like this, it doesn't really matter if it's dbfs or dbu or dbVU. They're all voltage scales, and we're not trying to talk about any absolute values. That is the "fs" or whatever tells us the reference level against which all the other values on the scales are measured, but if you say "I'm adding 6db to channel 2", that +6db is relative to wherever the fader was to begin with, and is the same ratio no matter what the meters think they're doing. To put it yet another way - meters should have labels so we can relate them to real world values, but faders don't need them because they are always referenced against the input signal.

Now, I'm pretty sure that this "threshold of hearing" thing is more about the overall level of a broadband source. If you have a song playing and somebody turns it up or down a db you're really just barely going to notice a difference if at all, and it actually might come across more as a change of tone or timbre than an actual change in loudness.

Our ear/brains are actually much more sensitive than that to changes in timbre - the relative balance of individual frequencies in a broadband signal. Those are the changes we're making when mixing. Turning up the guitar by 1db doesn't actually add up to a 1db increase in overall loudness of the mix (you'd have to turn all the faders up a db for that), but it does make a fairly subtle shift in the overall frequency balance, which is pretty easy for us to detect.
 
Last edited:
As far as I am aware the "3dB change is the smallest noticeable amount" notion was done many years ago and with pure tone. It does not really relate to real life signals and equipments.

Take the shelving switches found on the back of most monitors. Many just lift or cut the response by 2dB but, that lift is from say 2kHz to "infinity" and that is a huge chunk of the acoustic energy radiated by a speaker on a complex signal. So yes, you WILL notice that!

"Twice the subjective loudness needs ten times the power" And yet anyone who has swapped a 100W guitar amp for 10W one knows that the former is MASSIVELY louder.
This is in part because it is virtually impossible to find two amps, 90W apart but with identical voicing but also because bigger amps have bigger OPTraff and can deliver more clean LF and we are programmed to find V loud, V low frequencies terrifying!

Then of course as Ash' has pointed out, which "dB" are we talking about? Then in acoustics we generally speak of sound Pressure, analogous to voltage whereas the acoustician might have been referring to intensity, analogous to power. It is very difficult to measure the "power" level of a sound!

Lastly, the world of repro' is moving swiftly to a whole new way of measuring "loudness" and one that should help people like me, h of hearing, at last enjoy dialogue!
Check out Sound on Sound Feb 2014.


Dave.
 
You must log in or register to reply here.
Back
Top