"Official" Audyssey thread (FAQ in post #1) - Page 1657

This is my last try in explaining myself why Reference Level Offset is needed and why I think that music is mixed at lower levels, not at higher level as commonly stated.

Here's a diagram on which I will base my reasoning. So it's crucial to understand what the picture shows. The upper limit of the green area is the average level of a sound recording. The red area is the headroom for peaks.



A) This describes the situation in a dubbing stage. There's a defined reference level that links sound pressure level and the upper limit of the recording medium. 0dBFS = 105dB SPL.

B) shows how loud a modern music mix would sound when mixed and mastered in a dubbing stage. The headroom of such a recording might only be 3dB hence the average level is a whopping 102dB SPL. This is very loud and can lead to hearing damage. That's why in practice most music isn't mixed or mastered that way.
Nevertheless, if this mix would be played back on a movie reference calibrated system at home, the recording would still play at 102dB which would make us reduce the listening level. By doing so the music will sound bassshy because of the equal-loudness contours. But thanks to Dynamic EQ we can reduce the listening volume to a comfortable level without creating a spectral imbalance. Dynamic EQ corrects such a spectral imbalance by boosting bass. It boosts bass when we set the master volume dial below 0 and it cuts bass when we set it above 0.

C) Because mixing and mastering music at movie reference level is so unbearably loud, the mastering engineer will actually reduce the listening level. By doing so the music will sound bassshy because of the equal-loudness contours. He will ultimately correct this spectral imbalance by boosting bass.
At home we listen at similar low levels and the mix will sound more or less balanced. The correct listening level will alway be an approximation because the mastering engineer has no way to tell us about his reference level.
What happens when we play that recording on a movie reference level calibrated system with Dynamic EQ engaged?

D) Let's play the recording from C - that was mastered at lower levels - on a movie reference level calibrated system. Again the mix will produce an average sound pressure level of 102dB but this time the mix will sound bass-heavy when compared to example B because the mastering engineer added it in step C.
If we now reduce the listening level with Dynamic EQ engaged, even more bass boost is added. This will lead to bloated bass.
We need to tell Dynamic EQ somehow to reduce bass with the master volume control at 0. We also need to tell Dynamic EQ not to immediately start boosting bass below 0. This is achieved by setting Reference Level Offset.
We tell Dynamic EQ that the reference level is not 85dB any longer but 70dB. This is done by setting Reference Level Offset to 15 (85-70=15). Now with the master volume control at -15dB, Dynamic EQ would perform no processing because this is our new reference level. If we select a volume setting below -15dB, Dynamic EQ will again start to boost bass and it will cut low frequencies if we set the master volume control higher than -15dB. This is exactly the processing we need for our mix to maintain spectral balance.
We could also simply reduce source input level by 15dB. This would have the same effect as setting Reference Level Offset.

Marc, it isn't helpful repeating "you are wrong" without giving any additional explanation or without going into detail where my explanation might be wrong.

By the way, Roger agrees that reducing source input level has the same effect. Why don't you simply test it yourself? Reduce the output of your CD player or use IntelliVolume.

So someone has to be wrong. I don't have a problem with being that person but you don't present any arguments so I have to stand by my claim.

P.S. I never said that SPL in a dubbing stage is higher than in a music studio. In fact I stated the opposite some posts back. I also never said that recorded level of a movie is higher than on a music recording. It's also irrelevant for the discussion why Reference Level Offset is needed. The important argument is that nobody would mix and master music (with peak normalization) on a system that is movie reference level calibrated. It would simply be unbearable loud. Please read my post again. I don't know how to make it any clearer.

That's actually what is already happening. Everybody can only listen through his own ears. The decision if a sound needs EQ tweaking is based on that. So if Marc's hearing tells him that a sound is lacking bass, he will EQ it. You might have had decided differently.

I don't have the feeling that you really read my posts thoroughly. You put words in my mouth and you also do not really respond to my arguments. This is simply a waste of time for both of us and anybody else that might be interested to get a deeper understanding of the topic.

I don't want to lure you into anything. You can choose for yourself if you want to participate in the discussion or not. It is you that keeps on responding to my posts. All I wish for is a factual discussion.

So to recap, you agree to anything I've said here but you disagree that music is mixed at lower levels than film, correct?
Do you also agree that music isn't mixed, mastered and monitored on film calibrated systems because this would be unbearable loud? If yes, you probably also agree that even if a film calibrated system was used to begin with, the mastering engineer would turn down the monitoring level?