I guess I don't understand what normalization does under the hood. I always thought 100% was raising the peaks to the edge of clipping. I assumed that was zero.
Part of the confusion which appears to be occurring is due to the misunderstanding of the term "normalization". Normalisation just means raising or lowering different audio files/tracks to the same specified amount of the same specified property. Firstly then, you are saying just "normalization" and haven't specified the property! This is entirely understandable though, because by far the most commonly the specified property was peak level and therefore, when the term "normalisation" was used on it's own, it was assumed to mean "peak normalisation". Furthermore, as there is only one recognised standard in music, the physical limit of 0dB, it is assumed this is the "specified amount". However, that's not the case in other areas of commercial audio, TV for example. Before the new/current paradigm, the "specified property" of normalisation was again peak levels (or quasi-peak levels) but the "specified amount" varied. Most commonly in Europe it was -9dB. IE. In most European TV stations the audio was normalised to -9dBFS.
All this might seem like semantics but it's vital to understand the above, otherwise there's not the slightest chance that you'll understand the new paradigm of "loudness normalization". The "specified property" of loudness normalisation is the human perception of loudness, which is both frequency and time dependant but unrelated to peak level!
I've had those same thoughts after staring at the chart for a while. Also, the first chart he put up was quite a bit different from the second chart, which confuses me even more. My gut feeling is we'd have to read the actual pages in the books and not just look at the pictures to get it, as unfortunate as that may sound. Or have a recording engineer explain it to us.
They are a bit different. The second one is basically demonstrating that with digital we can normalise to the absolute peak (0dBFS) and over time most pop/rock music has become more compressed, higher average levels and typically either no headroom or only a few tenths of a dB from absolute peak. You couldn't do that with analogue media because unlike digital media, which stays perfectly flat/linear until you hit 0dBFS, analogue media gradually distorts more and more before hitting the absolute physical limit. Some of that distortion, in limited amounts, was often desirable artistically (tape saturation being a good example) but even then, there was a significant gap between the peak levels of the recording and the peak physical limit of the media, this gap is the "headroom". The first graph makes more sense, and it's designed to illustrate a point simply, which it does, but it's not entirely accurate. To be honest, it's many years since I read/studied Bob's book and the charts could be more accurate than they appear, depending on exactly how he's defining "headroom".
[1] I understand Bob's charts because I learn better via visualization than via acres of text.
[1a] The left chart simply represents various sources/tracks, of differing dynamic range(or PLR) normalized so that they all peak at or a fraction of a dB below 0dB full digital scale. Instead of being 'commercials' or 'movie soundtracks' or 'news broadcast' or whatever, imagine, for purpose of the discussion, ...
[1b] Now, imagine the right-hand graph representing that SAME album, but all songs on it LOUDNESS normalized - either with the help of a loudness meter plugin, or even by the engineer just using their ears, playing each one at a time, or several simultaneously, until no one song seems to stick out in the ensuing cacophony.
[1c] Mastered that way, such a CD could be played with the listener setting their volume once, during the first track, and perhaps never having to adjust the volume again - except to turn it down to take a phone call, or perhaps because his concentration when driving is needed while proceeding through a construction zone.
[2] Sure, peaks/transients are louder than other sounds. But in the case of music, they last typically less than 1/10 of 1 second, sometimes as little as 10-30milli-seconds.
[2a] Their presence, relative to the average level of a song, make folks want to get up and dance to it, or at least nod their head or tap their foot.
1. Ironically, you've both confirmed what I posted in my last message and explained why you DON'T understand Bob's charts!! Instead of "visualizing" an understanding from Bob's charts, read the text, understand the context of the charts and then you'll have a much higher likelihood of arriving at a true understanding instead of a complete misunderstanding!
1a. Instead of "imagine, for the purpose of the discussion", why don't you stick to the actual facts rather than changing the "purpose of the discussion" to your personal agenda?
1b. Again, this is the "sound science" forum, not the "what thesonictruth wants to imagine" forum! Also, your last suggestion clearly wouldn't work. A song "sticking out in the ensuing cacophony" could just indicate that it's at a particular loud point relative to the other songs, not that it's louder overall. For example, it might be sticking out because it's the chorus while the other songs are in their verse. Again, you don't seem to understand what loudness normalisation is but rather than go and learn what it is, you simply carry on regardless.
1c. Yes, you could and there would certainly be some great advantages to loudness normalisation being applied to music creation but there would also be some disadvantages. Rather than only picking facts (or misunderstandings) which support your agenda, why don't you include all the facts and arrive at a true understanding?
2. True!
2a. Oh dear, completely false! You said you don't want me to state that your assertions are completely wrong/backwards but seem oblivious to the blatantly obvious solution, don't post assertions that are completely wrong/backwards to start with! Surely that's not a difficult concept to grasp? To your point: An acoustic Mahler Symphony obviously has no compression and very high peaks relative to it's average level. How many people have you seen "get up and dance or at least nod their head or tap their foot" at a Mahler Symphony? On the other hand, DJ's in night clubs typically apply huge amounts of compression and reduce the peak/transients to almost no higher than the average level of a song. How many people have you seen "get up and dance, or at least nod their head or tap their foot" in a night club? There are several factors at play here of course, but high peak/transients relative to average song level isn't really one of them!
[1] The whole meme of "loudness wars" is a total misnomer...
[1a] The peak level you can store on a CD hasn't changed...
[1b] Therefore modern CDs aren't any louder than older ones...
[1c] What they are is more dynamically compressed... The quiet parts have been made less quiet - which raises the average level - but the loudest spots aren't any louder.
1. While I agree with the basic message/principles of your post, this part of it is incorrect. You are confusing loudness with level!
1a. True.
1b. False, modern CDs are louder than older ones but their peak level hasn't changed.
1c. Partly true. The average level is higher but the loudest spots are also louder (although their peak level is the same). Loudness is a human perception, while levels are not. The perception of loudness depends on several factors; average levels over the short term, differences/contrasts in average levels over the longer term, the audio frequency/pitch of those levels and a few other factors (size/distance of an acoustic environment for example).
The loudness wars is not just about more compression, it's also about all the other factors that affect the perception of loudness and therefore "loudness wars" is not a misnomer, it's accurate.
The "AVERAGE LEVEL" - which is 90% of what we judge the loudness of something by!! Not the peaks!
If you're going to correct someone, please do so with the actual facts. Average level is not 90% of what we judge loudness by, there are other factors as important or more important. For example, which is louder: A. A sound with an average level of -10dB (RMS) or B. A sound with an average level of -16dB (RMS)? The answer according to you must be "A" but the correct answer is that it's impossible to know from the information given, it could be either. For example, if A is centred around 80Hz and B is centred around 2kHz, B will sound several times louder, even though it's average level is half of A!
G
Edit: I wrote this before I saw Keith's last post. I have posted that document several times and also highly recommend it. I would say that there is only one definition of loudness - "perceived loudness", the actual signal levels are the actual signal levels and not "loudness".