Youtube - The Loudness War

[ursamtl]

Just thought I'd pass on the link to this excellent presentation on Youtube about the "loudness war" so many people are discussing these days. This is a clear, simple, and visual explanation of what's been happening to modern music and some movie audio.

The Loudness War (http://www.youtube.com/watch?v=3Gmex_4hreQ)

For more information, you can see :
Wikipedia entry for the Loudness War (http://en.wikipedia.org/wiki/Loudness_war)
Declaring an end to the loudness wars (http://www.barrydiamentaudio.com/loudness.htm)
If the Loudness War struck the art world (http://recforums.prosoundweb.com/index.php/mv/msg/4286/0/0/0/)
Why the new "LOUDNESS CRAZE" in digital mastering really robs music of life.. (http://www.stevehoffman.tv/forums/showpost.php?p=2230655&postcount=1)
The loudness wars: why many CDs sound bad (http://www.itwriting.com/blog/?p=91) (see bottom of this page for more links)

Regards,
Steve.

[check]

It's an interesting debate, but the fact many people now listen with headphones in busy public areas means that they prefer audio with a lower dynamic range. Too many opponents of lower dynamic range forget this fact and assume there is no advantage in what they are opposing. I definately prefer the 'old mixing style' more in the comfort of my own home, but when out an about, a more constant audio level is far easier to listen to.

[raquete]

figure of eight....i'm totally against loudness war,volume allways have my atention....for example:
did you can hear that in the original source the initial drum have more volume(punch) than when the remainders instruments are playing together(no more punch) in the rest of the sample?...or in the rest of the music(i have the source)
the volume of the drums decay to round 50% comparing with the starting point.you can clear hear the difference!
it means that have some "compressor/normalizer" in the drums in the original mixing?..of course something was used and this music is not one good source to be used as sample.
i will that i'm clear.

thanks so much for the links steve,very interesting!

[ursamtl]

It's an interesting debate, but the fact many people now listen with headphones in busy public areas means that they prefer audio with a lower dynamic range. Too many opponents of lower dynamic range forget this fact and assume there is no advantage in what they are opposing. I definately prefer the 'old mixing style' more in the comfort of my own home, but when out an about, a more constant audio level is far easier to listen to.

Yes, but by releasing the recordings already maximized and brickwalled, the industry is forcing us to listen to overly compressed music whether it's at home or in portable devices. The result is that the music ends up squashed. A good example is the Red Hot Chili Peppers' Stadium Arcadium. It's an excellent album but was squashed to the point where the waveforms look like a solid line. The album was recently remastered by Steve Hoffman and released on vinyl. I bought this and the difference is amazing. The tracks are far more powerful and musically interesting.

As far as listening on portable devices, I've been doing so daily now for quite some time. I use Replaygain on everything before loading it on my mp3 player. The levels are always the same and the fidelity is much better. One problem that's been associated with overly compressed music is listener fatigue. I don't get that with the replaygained tracks.

[Mug Funky]

i'm definitely against the loudness race, but understand where it's coming from.

what i'd like to see is compressor/limiters built into playback equipment - sort of like how old tape decks included Dolby NR (a form of compander), or old amps had "loudness" switches that did who knows what.

iPods and the like are more powerful than the computers were 10 years ago. why on earth are they only ever running 6-band EQ's when they could be doing multi-band compression and limiting as well? sure there's battery life to consider, but that's always improving, and it certainly doesn't discourage manufacturers from allowing h.264 playback and other highly CPU intensive stuff.

about listening to music in noisy situations (traffic, etc). sure, it's annoying when the music dips down a bit and you can't hear it, but bear in mind that if you can hear music to a satisfying level when you're out in heavy traffic, you'll be deaf in a year or two - it's way more than your average pair of ears can handle, traffic plus loud music. i would suggest that if someone can't hear reasonably dynamic music, they should go somewhere quieter.

it gives me the irrits when i play a CD in the same device i play DVDs on - my DVD player doesn't apply DRC (and i'd turn it off if it did because i like my action movies actiony :)), and hence the reference level is -20dB. put in a music CD and suddenly i've blown out the windows, as it'll be a good 10-15dB higher than that.

so in short: loudness should be handled during playback, not mastering.

[ursamtl]

Mug, I agree with you wholeheartedly. Let us, the listeners, decide how loud we want the music, and if we want it compressed or not. I remember a rental car I had once that had a compression switch on the stereo. It seemed like a great idea.

Of course, there's another simple solution, just turn it up and enjoy the full dynamics of the music. ;)

[DarkZell666]

Don't forget that compression, in some cases, gives very interesting artistic effects ;)

Electronic music sounds better compressed than not (you don't have the same type of transients to work with in the first place either).

I'm not opposing to what's being said though, too much (of anything for that matter), is just useless :)

[ursamtl]

Compression as an artistic effect applied to instruments or voices and chosen by the performer during the recording or mixing of the track can be tremendous, especially on vocals.

Likewise, limiting as a safety tool to prevent clipping distortion is definitely desirable.

However, boosting an overall mix so that it'll sound louder than everything else by applying compressing, then boosting and limiting ruins makes the music end up sounding like an annoying commercial that blasts on in the middle of a TV show you're trying to watch, or like the endless annoying square wave beep that drones on behind old TV test patterns.

The graphic in this youtube presentation depicts the problem so well. All those red transients the guy shows as missing have been chopped off by the process. In a sense, that renders the music file every bit as LOSSY as a poorly encoded MP3.

I've seen this when comparing the waveforms of quite a few "digitally remastered" recordings with original releases done in the 80's. It's as if they simply boosted the original by 10db or so and then smoothed the resulting square wave clips with a maximizer circuit so the sound isn't quite as bad. Initially if you go back to the old CD, it's easy to mistakenly think it sounds dull by comparison, but then if you crank up the volume on even a reasonable sound system, the performance sounds much more authentic.

Another interesting point is that the problem seems to be format-related. A good example is the Beatles Love release. I've seen graphics of the waveforms from both the stereo CD and the stereo track on the DVD. The CD looks typically chopped but the DVD tracks look much more like a real audio waveform. If one were to crank the two of them on a good system, the difference would be like night and day!


Don't forget that compression, in some cases, gives very interesting artistic effects ;)

Electronic music sounds better compressed than not (you don't have the same type of transients to work with in the first place either).

I'm not opposing to what's being said though, too much (of anything for that matter), is just useless :)

[olyteddy]

For a great 'real world' example of compression at work, just tune across the Analog FM radio band. Some of those stations have spent tens of thousands of dollars on devices with names like 'SONIC MAXIMIZER', 'AURALIZER', 'RECTALIZER'... OK, I made the last one up, but you get the point. The idea is that people will stop at the loudest station, because it sounds 'the best'...:p

[ursamtl]

You're right, oly, and the ironic part is that compressing the original CD releases to sound louder on radio is completely unnecessary because radio does it anyway!

[Mug Funky]

yes especially considering those maximizers typically use "phase rotation" to make vocal more symmetrical (vocals, especially male vocals are quite asymmetrical - they'll peak higher in the positive than the negative. to the point where in a mix you can tell whose microphones were inverted). doing this will turn clipped peaks into zero crossings, and introduce a new (higher) peak than was there before. indeed a dynamic mix often comes out louder than a smashed one after radio processing.

now, on the CD v DVD loudness issue, the main reason is they come from different disciplines, namely broadcast TV and the music industry.

DVD has lots of legacy stuff from broadcast TV that has strict calibration and limits in order to fit the audio into an analog system. reference tones (the annoying ones you referred to) were set in analog to 0 dB VU. to give headroom, when it all went digital 0 dB was set to -20 dB from fullscale, and +10 dB VU was logically set to -10dB FS in digital.

that means there's 10dB room for peaks above the reference level in analog, and in digital there's an extra 10dB on top of that to allow for analog systems not always stopping at -10.

currently, stereo mixes on DVDs are (usually) limited at -10dB, and multichannel mixes aren't limited at all, but share the same -20dB reference. this means you can fit explosions and normal spoken dialogue in the same mix.

so basically, when the engineers who did "love" were outputting the respective DVD and CD-A mixes, they gave them a different treatment.

one common fallacy is that 16 bits isn't enough physical dynamic range and so things need to be compressed to avoid running against the noise floor. wheras on DVD there's 24 bits so you can make things quieter without problem. that's complete arse rubbish, but a lot of people believe it. in reality most DVD audio is in fact provided on 16-bit masters, and there's no problem with quantization noise. sure, digibeta can carry 20 bits, and DA-88 can go to 24 bits, but most often they're not done that way. if it's not truncated on the tape, it'll often be truncated when it's captured.

[raquete]

i don't use but i have a question:
what to tell about Classic master limiter and seamless used in some guides here?

[ursamtl]

When I use the limiter in my guides, it's mainly to block any digital clipping. The process of extracting surround information from stereo can "unmask" transients that were reduced by phase cancellation and accentuate others. Files that may have peaked below zero dB suddenly produce files with peaks above 0dB. By adding a limiter, these transients will not go above 0dB and distort.

The exception is when I suggest using the limiter to boost dialog levels in the center channel or bass level in the LFE. I always caution people to be careful with this because a "little goes a long way." Just like adding EQ, reverb or any effect, it's important to do this kind of thing very slightly. This in fact is closer to the use of compression on individual instruments than on a whole mix.

In contrast, the complaint of those of us who don't like the results of the so-called loudness war is that compression is being applied across the board to entire mixes and squashing the music. I find myself disgusted with some of the results I hear on radio. For example, Gnarls Barkley's Crazy or Nelly Furtado's Say It Right are both great songs but if you listen to them on CD, the kick drum is already boosted so loud that it distorts when run through a maximizer during mastering. These are the kinds of mixes that would never have been released 20 years ago. They'd have been considered defective and sent back to the mastering house for reworking.They sound less bad on radio because of the additional broadcast compression, but if you actually buy the CD and play it at home, you can't turn them up to a decent listening level because it sounds like your speakers are blown!

[raquete]

your answer was one lesson,very clever.
thanks you!