am i hitting -18dbfs

D

djclueveli

New member
when i recorded vocals in soundforge, my signal is between -12 and -18 but then when i got to statistics in it reads somewhere around -28 to -30 rms level. is this a good level or should it be higher?
 
the first value you are seeing is peak, rms is kind of like an "average" level, and it's probably including the periods of silence on the track. Your levels seem fine if they are peaking at -12dbfs.

-Timedog
 
i always read the rms should be somewhere around -18 and -20. how come when i'm around -12 and -18 dbfs the rms is so low when people recommend record at those levels and then says "the avergae level should be about "-20dbrms"
 
djclueveli said:
i always read the rms should be somewhere around -18 and -20. how come when i'm around -12 and -18 dbfs the rms is so low when people recommend record at those levels and then says "the avergae level should be about "-20dbrms"
Click to expand...

You are getting 2 conflicting sets of views. If you are peaking between -18 and -12 you are fine.

-Timedog
 
Well, until someone (like probably Glen or Farview or someone else) comes in explaining all the dbFS, dvVU terms, etc. don't worry about it so much.

If the meters in your software are averaging around -18dbFS on your software meters then you're fine.

Some people do gain staging differently for different instruments. I've heard of people getting as close to clipping as possible on drums (even for 24-bit). But really, if you're averaging about -12 or -18 then I wouldn't worry about it too much.

If you want to know the specifics, they've been covered a million times over and a simple search for gain staging or 'dbFS' will probably yield more than enough results. ;)
 
danny.guitar said:
..If the meters in your software are averaging around -18dbFS on your software meters then you're fine.
Click to expand...
In other words, if the avererage levels (the sustaned parts for example) are at -18 on peak reading meters, you have another 18 db of head room below digital full scale.
For drums and such where there is little 'average to be had, just make sure the peaks are still safely below full scale.
 
djclueveli said:
i always read the rms should be somewhere around -18 and -20. how come when i'm around -12 and -18 dbfs the rms is so low when people recommend record at those levels and then says "the avergae level should be about "-20dbrms"
Click to expand...
The thing to remember about average or RMS values is that they depend entirely on the density of the audio signal itself. Think of it this way:

If you recorded a metronomic click track that clicked twice a second, evem if the click value was all the way up at 0dBFS (maximum) the average or RMS value would still be waaaaay down in the cellar. I'm not sure of the exact value (anyone care to try it for us?), but it'd probably be below -20dBFS, even though the peaks are all the way up at zero. That's because the average value is actually closer to silent, because most of the time the signal actually is silent.

On the other hand, if you have a wall of guitar distortion at feedback run through heavy compression, even if you're only peaking at, say, -12dBFS - 12 dB lower than the click track - the RMS value could easily be above -20dBFS, and possibly as high as -14 or -15dBFS. Because the average signal level is much higher.

I think those that told you that a peak level of -12 to -18dBFS would yield an expected RMS level of -20dBFS were probably used to seeing headbanger-class dynamics; i.e. flat as a pancake. If you're doing a different genre of music (and if the "dj" in your handle means anything, it's probably hip hop, but it would be just as true for many flavors of jazz, folk, pop, classical, country, etc.), the difference between your peaks and your average* are probably going to be much greater, and that is normal.

*that difference between the peak and the RMS values is called the "crest factor", BTW.

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