Loudness, rms, peak... A basic question

4140

New member
Im learning what all this means but I had a quick question concerning my mix. OK, I have a mix I'm pretty happy with. I know I'm supposed to leave some headroom. I'm using reaper and on the leval meters, the main one is peak and I'm pretty sure the two on the outside of it is rms (they go up and down in a slower, more steady way). Now, as far as peak, nothing is going passes - 6 to - 12 db. However, on some parts the rms is touching red (-6 I think). I can tell it's when the bass and kick hit. Is this bad? Where should overall loudness be for a final mix, before mastering?
 
If your RMS level was actually the same as your peak levels, it would probably be pretty horrible. Luckily, I think that Reaper's master meter has a display offset for the RMS levels by default. In that case, it's probably pretty ok. If you want to find out or change how yours is set, right-click the meter.

In the real world of modern mixing, there is no good reason to "leave headroom" in your mixes. There's no good reason not to, either, of course. If you're ever in doubt, render your mixes as floating point .wav and there's literally nothing to worry about. Up until you are actually rendering the final distribution files, the actual numbers on the meters - peak OR rms - are arbitrary. It's the difference between then that matters.
 
For real? I thought the point of headroom was to give the mastering engineer room to work, I mean everything I've read says to do this...
 
For real? I thought the point of headroom was to give the mastering engineer room to work, I mean everything I've read says to do this...

I'd say it's also about giving yourself room to work with no worry of anything clipping during the record/mix process. Some headroom is good when it goes to mastering, but 3dB is fine. It doesn't have to be 12dB.
 
That's a common misunderstanding. It made more sense when more of the process was analog. It's not really a thing in the modern digital world.

If the ME needs "room to work", they can just turn the damn thing down. You might want to leave some Dynamic Range (the difference between peak and RMS) in your mix so that the ME can squish it down to your final target for you. If you have smashed your mix into a limiter to get it LOUD, that'll be tough to undo, and the ME might be able to do it better without so much damage. Still doesn't matter what the peak or RMS values actually are, just the difference between them.
 
That's a common misunderstanding. It made more sense when more of the process was analog. It's not really a thing in the modern digital world.

If the ME needs "room to work", they can just turn the damn thing down. You might want to leave some Dynamic Range (the difference between peak and RMS) in your mix so that the ME can squish it down to your final target for you. If you have smashed your mix into a limiter to get it LOUD, that'll be tough to undo, and the ME might be able to do it better without so much damage. Still doesn't matter what the peak or RMS values actually are, just the difference between them.

Just want to second this. 99% of the time I don't apply compression to the entire mix (the master bus), I leave that for the mastering stage. Only use as much compression at the track level as you find necessary to get the instrument/vocal sounding the way you want. Your mix should have some dynamics.
 
RMS value can be pretty high or low from time to time, peak for some synth bass will pass their RMS for around 3db. For the whole track, I personlly think the RMS range can be from -15db to - 9db (depends on which side you choose in loudness war). One thing should notice is, some RMS leve meters shows offset value, for instance the number shown is 3db more than it actually is.
 
OK, so for the 5 songs on this ep, the peaks average around - 12 to - 6 db and the rms usually hovers around - 6, sometimes going up higher to around
- 3. The project was recorded and mixed 24/44.1 which is what I'm going to give the M. E. Unless he will take 32 bit float. Am I in good shape? I really appreciate all the help this EP is the first thing I've released in a few years and I've learned a lot through the process
 
OK, so for the 5 songs on this ep, the peaks average around - 12 to - 6 db and the rms usually hovers around - 6, sometimes going up higher to around
- 3.

Those numbers don't make sense. The RMS level can't be higher (smaller number) than the peak level. If those peak numbers are right, then the average should be more like -18dBFS, which would be about right.
 
If your peaks are at -12dbFS, and you render to 24bit fixed, you're only actually using 22 bits. That's still a lot, but if there's going to be compression and stuff happening later, you might start to notice either noise floor or truncation distortion coming in. I'd strongly suggest printing to 32bit float in order to preserve as much of the dynamic range as possible.

You still haven't told us what the settings on your master meter are set to, so it's hard to correlate your RMS values to any kind of real world numbers.

Edit - Witness bsg's confusion directly above (he ninja'd me!)
 
Last edited:
I don't know what my meters are set to, whatever the default is in reaper. I just know that the meter on the inside peaks at - 12 and the two on the outside, the ones that move slower go up to - 6 but not all the time. Cone to think of it, I don't even know if the outer meter IS rms, I just assumed it was.
 
Back
Top