Answering that question would take a fair-sized book, but you might get an idea from the latest podcast on The Mastering Show by Ian Shepherd and Jon Tidey. In fact if you go right back to podcast 1 and work your way through all 35 episodes, you won't have many questions left about mastering.

Hi Gary, well yes I appreciate its complex, but my main question was I seem to be looking at the same type of on screen result for either a single song or a multitude of songs when in project, so in a single song I get it, but whats it showing me in multiple songs, an average?

Thanks

Loudness is a different measurement than for example a VU meter. A VU meter is telling you how loud the music is for example in a much smaller time window. Still a useful measurement but very different to loudness. Loudness works out the average level over the whole song in fact rather than just a smaller time window.

Loudness readings depend so much more on the type of material that you are monitoring. For example lets take a Steely Dan tune that might be sitting at pretty well the same volume all the way from start to finish. It might be a Loudness reading of say -10 LUFS.

But lets take another tune that is for the most part sitting at a certain level for most of the tune but for a section in the middle it drops way down quiet for a section of it. What will happen here is the LUFS reading will be lower. e.g. It might be -14 LUFS and that is because the music went quiet for a while. The Loudness reading has factored that in. Does this mean then that this tune is 4 dB quieter than the Steely Dan tune. The answer is no! It just means it went quiet for a bit in the middle that is all. It can be a bit misleading like this can you see. You still have to concern yourself with how loud the normal fully reached levels are, and in fact a VU meter is still better at this. Because it is not concerned with the quiet bit in the middle. Although the VU meter will still tell you how quiet the middle section is compared to the rest of the song though.

If you are taking Loudness readings over a bunch of songs then it is doing that to all of them. LUFS readings may be a bit all over the place.

So firstly what is the material like you are checking. What I find more helpful for me is I tend to check with the VU meter first and only on the sections that are up there for most of the time as well. When I am mastering, I factor that in so that over a group of songs I just make sure that the levels that they all reach for most of the time are very similar e.g. on the VU.

You don't have to get that concerned with Loudness when you do things this way because any Loudness readings will always be at least a little lower than the VU readings and if the song goes quite for a bit then the LUFS readings will drop down further. Classical music will have low LUFS readings for example.

LUFS readings tend to get a little more important when you are sending out soundtracks to say a video production and they want the Loudness readings to be within a certain range over the course of the whole soundtrack. e.g. for broadcast.

What I tend to prefer these days is the Dynamic Range readings of your music. There a few free DR meters around and I find these are actually bit more useful. They tend to stop you from going for super loud masters with very poor DR. DR readings should be 10 or more at least.

If your music is at a consistent level for most of the time then you don't have to sweat LUFS readings too much. Use a VU meter instead and keep all the max levels the same. LUFS readings are usually about 2 dB lower than VU readings. (for music that is consistent that is) e.g. If I master a tune at -14 rms on a VU meter the LUFS reading is around -16 which is perfect for iTunes for example. The Apple Soundcheck mode will tend to leave a -16 LUFS reading pretty well alone.

Well said, Jeff. A more complete explanation than I could have given.

Tim: You might like to take a look at the (free) Youlean Loudness Meter as a tool that gives you a lot of choice, and includes dynamic range.