How do you find db level of a song?

[csshakka]

Before anyone flames me for this, I know music is dynamic - the db level per frequency varies along with what's playing in the song, etc. What I'm interested in, how do you find the average db level of a song that would be safe to tune to?

Example: Tune amp to -7.5db 40hz and 1000hz test tones for basic setup using DD-1. Assuming max volume being played, would a song like Skrillex - Recess introduce clipping? How do you determine this?

From another thread....

Highlight 1sec or less, no more
Go To Analyze
Then Plot Spectrum
Change Size to 16384
Change Axis to LOG frequency
Scroll cursor until it aligns with the vertical line designating the peak dB level
Look at the info below the graph and it tells you what Hz, the musical note and dB level.

Doing this on that skrillex song, I'm getting weird results and varies hugely depending on where that 1 second or less is taken from (which makes sense). I also used the wave stats plugin for audacity, which gives reasonable results of RMS -6.2dBFS and RMS (a-weighted) -10.8 dBFS. Given an a-weighted curve, I would assume -6db'ish is safe for mids/high and -10db'ish is safe for subs. Wanted to check with everyone else on this.

[csshakka]

anyone?

[philrab]

Not sure if its exactly what you're looking for, but MP3Gain analyzes and reports db's for selected MP3's, though I'm unsure if its max or average that's reported. It also displays a warning if the mp3 is clipping (varied opinions on whether or not its accurate) and lets you normalize gain across all your mp3's.

[csshakka]

Not sure if its exactly what you're looking for, but MP3Gain analyzes and reports db's for selected MP3's, though I'm unsure if its max or average that's reported. It also displays a warning if the mp3 is clipping (varied opinions on whether or not its accurate) and lets you normalize gain across all your mp3's.

Hmmm good thought, but how would you interpret that? Like when it says track gain -10.5db or whatever, how does that relate to 0db reference recording point?

[philrab]

Got me. Hopefully a more experienced member drops in.

[Weigel21]

Audacity?

[Noel989]

So your planning or tuning and retuning for every song?

[Weigel21]

From the quote the OP put, I'm thinking Audacity was used, which that can give the dB level of frequencies throughout any particular song. Finding the average makes little sense to me, as that would still allow for much of the frequencies recorded in a song to play at levels inducing distortion and I thought the whole reason for using a DD-1 was to set the gains for minimal distortion.

You could plot, for instance, the song you listed and find, for example, that it's average recording level is 9dB, however, that doesn't change that much of the bass peaks in the song are recorded at say -6dB with a rare -5 or -4dB bass hit. So are you content with some of the song, the frequencies that tend to really shine, playing with distortion? If so, then going about with what you've said makes perfect sense.

[hdorre]

Yes, audacity was used

[csshakka]

So your planning or tuning and retuning for every song?

No, it's just more a curiosity and prevention thing. Let's say an amp's gain is set with a -7.5db test tone and a SMD DD-1. If most songs can play max volume with no clipping, great. Then there's the few songs that are louder than others. It'd be nice to open up that song and confirm that many frequencies/parts of the song were recorded somewhere around let's say a -4db level. Then you could either use a program to normalize the volume of that song, or know for sure that at max volume you will be clipping and to turn it down a bit.

From the quote the OP put, I'm thinking Audacity was used, which that can give the dB level of frequencies throughout any particular song. Finding the average makes little sense to me, as that would still allow for much of the frequencies recorded in a song to play at levels inducing distortion and I thought the whole reason for using a DD-1 was to set the gains for minimal distortion.

You could plot, for instance, the song you listed and find, for example, that it's average recording level is 9dB, however, that doesn't change that much of the bass peaks in the song are recorded at say -6dB with a rare -5 or -4dB bass hit. So are you content with some of the song, the frequencies that tend to really shine, playing with distortion? If so, then going about with what you've said makes perfect sense.

I think you're right about the average thing - it could be misleading. But setting gain with a DD-1 just allows you to see when the speaker output begins to distort with a reference signal. If that reference signal is a -15db signal, you will minimize distortion as long as the source is a -15db recording. As soon as the source goes above that, which most songs will at some point, then it will clip. What I'm after is taking a look at the music I generally listen to, to find an optimal recording level to set my amp gains at. I get what you're saying though about how each frequency at each point of the song will be different.