Whatever Happened To Dynamics?

Whatever Happened to Dynamics? A Note on the Loudness Race Ivar Løkken, 2006 "The average level of popular music CD's continues to rise. Popular CD's with this problem are becoming increasingly prevalent, coexisting with discs that have a beautiful dynamic range and impact, but whose loudness (and distortion level) are far lower. What good is a 24 bit 96-kHz digital audio system if the programs we create only have a 1 bit dynamic range?" Bob Katz, Digital Domain Studio [1]

Dynamics, the very soul and nerve of music, is something that in the modern age of digital audio seems to be disappearing more and more. This is somewhat of a paradox; both digital formats and equipment have higher dynamic range than ever before. The technology is there to create music with pitch silent background, virtually no distortion and a dynamic range just as high as the one we enjoy in the concert hall. Still, what we get from CDs, especially modern pop recordings, is music that is massively compressed and also clipped, leading to distortion and loss of detail. This trend has been dubbed “The Loudness-Race” and an internet search will show that many critical voices are raised, both from consumers and professionals. Still, this seems to be an ongoing trend that is not going to turn anytime soon.

The Loudness Race – Reasons “We aim very much towards radio reproduction when producing and especially when mixing music, because ultimately it is through such media that music is mostly played today.” Hallgeir Rustan, Stargate Studios [2].

Music, especially popular music, is sold through mass media. This means that it is exposed to the public through radios, car stereos and music videos on TV. It also means that these media serve as sales channels. TVs and radios are low-quality playback systems with high distortion and low dynamic range. Thus, low level detail is very hard to notice. Since the units serve as sales channels, the record companies want their product to be noticed first. As stated by Dr. David Robinson [3] “Quite simply, when a listener is flicking through radio stations, a louder station is more likely to catch their attention. Also, it is well known that if two different presentations of the same audio signal are compared sequentially, then the louder one is subjectively preferred.” However, he also writes: “The advantage is transitory: once a listener has chosen a station, they adjust the volume control on their audio equipment to deliver a comfortable listening level.” In the current world of information and mass media overload, it is all about being noticed. When exposed to music through TV or radio, the majority of the audience is not in a position to make any critical assessment of sound quality. Consequently the producers are targeting the most basic of all psychoacoustic mechanisms; if it’s loud, we’ll hear it. However, hearing and listening are two very different things; for a listener the loudness-race will mean distortion and loss of natural dynamics, it simply won’t sound like real music.

The Loudness Race – Consequences “One of the gross areas is level; to get their records noticed, people want them to be louder than the next guy's, and to get level, there's a lot of processing that goes on. The by-product of that is a form of distortion. It usually isn't blatant [distortion], but it doesn't sound natural and it sounds less musical; it sounds almost mechanical.” Bernie Grundman, Berne Grundman Mastering [4].

To achieve the high average volume that is used on modern popular music recordings, two processes are mainly used; dynamic compression and limiting [5]. A compressor is a unit that amplifies the low-level signals and reduces the loud ones. This brings low and high level parts closer together and thus reduces the dynamic range of the source material allowing it to be boosted in level. To avoid harmonic distortion a compressor is usually realized as a time-variant linear amplifier, i.e. an amplifier whose gain varies slowly, dependent on the power of the input signal averaged over time. This is called dynamic compression. A dynamic compressor is characterized by its power gain curve and its dynamic behaviour, the attack and release time. Dynamic compression is something that has been done since the earliest days of music recording and is considered absolutely necessary. A percussion set or an 8x12” guitar rig recorded with near-field microphones has much too high dynamic range to be reproduced by any home audio system. This is equipment designed to fill big halls with very loud sound, few listeners would use such levels when sitting in front of their hi-fi. But with the digital processing available today, there is almost no limit to how aggressive a dynamic compressor might be. In many modern recordings, the SPL is constantly between -6dB and 0dB referenced to the full scale range. When an average hi-fi system is capable of perhaps 60-80dB dynamic range, it is obvious that its potential for reproducing pianissimos and fortissimos is not in any way utilized. As mentioned it’s tailor-made for radios and TVs. When recording, there are bound to be transients that overshoot the input range. A limiter is used to ensure the input is not overloaded. A digital system is self-limiting since it will just saturate when the input is overloaded. However, the engineer can use soft limiters which will produce less distortion than pure clipping, called soft limiters. They can also amplify weak signals and act as a static compressor. Limiters are usually characterized as having “hard knee” or “soft knee” responses. Static compression/limiting and dynamic compression is usually combined in one unit.

Figure 1: Hard knee (left) and soft knee compressor/limiter

So what does this do with the music? Obviously, it adds distortion products, since it is symmetrical odd harmonic ones. Although a soft-limiter has a somewhat gentler distortion characteristic than pure clipping, significant amounts of harmonics are produced. Figure 2 shows a pure sinewave run through a limiter with -15dB threshold and 4:1 compression in the sound editor Audacity. As can be seen, the harmonic content is gross.

Figure 2: Signal run through -15dB 4:1 compressor/limiter, Audacity. Distortion due to clipping is very much something that can be found in commercial releases as well. Figure 3 shows a 10ms measurement of a clip from Jennifer Lopez’ CD “J-Lo” (Columbia Records SME- -5005507). We can easily see very much clipping present.

Figure 3: Example of clipping in commercial audio material, Jennifer Lopez ”J-Lo”

In addition to the above, and as mentioned earlier, dynamics is what constitutes the very nerve and soul in music. The nuances in the playing is what separates a great pianist from a

mediocre one, the punch of the rhythm is what gives rock ‘n’ roll its attack. Any musician can play or sing pure notes, but to deliver it with correct dynamics is what gives the performance emotion. If that is not preserved in the recording, much of the experience is gone. Also, when the music is very loud all the time, the listener quickly gets fatigued, it simply becomes annoying. Few have said it better than Rip Rowan, editor of Prorec.com [6]:

WHY IS THE “LOUDER IS BETTER” APPROACH THE WRONG APPROACH? BECAUSE WHEN ALL OF THE SIGNAL IS AT THE MAXIMUM LEVEL, THEN THERE IS NO WAY FOR THE SIGNAL TO HAVE ANY PUNCH. THE WHOLE THING COMES SCREAMING AT YOU LIKE A MESSAGE IN ALL CAPITAL LETTERS. AS WE ALL KNOW, WHEN YOU TYPE IN ALL CAPITAL LETTERS THERE ARE NO CUES TO HELP THE BRAIN MAKE SENSE OF THE SIGNAL, AND THE MIND TIRES QUICKLY OF TRYING TO PROCESS WHAT IS, BASICALLY, WHITE NOISE. LIKEWISE, A SIGNAL THAT JUST PEGS THE METERS CAUSES THE BRAIN TO REACT AS THOUGH IT IS BEING FED WHITE NOISE. WE SIMPLY FILTER IT OUT AND QUIT TRYING TO PROCESS IT. We notice it at first glance surely enough, which is just what the producers want, but we will not be able to process it, or more importantly enjoy it, over time. If you were browsing through the document the above quotation probably caught your attention first, but would you enjoy reading this document if it was written like that in its entirety? Probably not. The paradox is; the better your hi-fi system gets, the more noticeable this problem is. Thus, the record industry is pulling in the opposite direction of the audio engineering community. As digital recording and playback equipment gets better and better, as new high resolution formats with extremely low SNDR-limits are introduced, the real performance we get from the average hit record, both in terms of distortion and dynamic range, is worse than ever. If listening through a TV-set or via the car-radio when driving, it is acceptable because the quality of the playback is so low in any case, but the ones investing in high-performance hi-fi systems are definitely getting ripped off. The process is not reversible with generic expanders either, as the compression is done track by track rather than on the downmix, and the added harmonic distortion is already there.

The Loudness Race – Status I know that many of the executives that I have spoken to secretly sympathize with the fact that compact discs these days sound so loud and so squashed and so crappy--they don't like that sound. But they can't do anything about it because all of their competitors have loud CDs and they're stuck. Steve Hoffman, Steve Hoffman Recordings [8]

In general, there is no doubt that the loudness-race is ongoing. Figure 4 shows the average (long-term) RMS level of a selection of CDs from different time periods [7]. As we can see, the RMS-level has gone from -18dB relative to full scale in 1985 to less than -2dB relative to full scale in 2002. With the long-term RMS-level at -2dB there is virtually no headroom left for dynamic impact, which many modern recordings are absolutely void of.

Figure 4: Average RMS-level of CDs from various time periods [7].

Figure 5: Waveform, 1m. Sample, Robbie Williams “Escapology” So, is there really no hope? Well, the loudness race is largely an invention of the record industry, but now a growing number of studio engineers along with audiophiles are beginning

to call out warnings [9]-[10]. The European Broadcasting Union has attempted to release a recommendation to record producers standardizing at least18dB headroom between the nominal level and the digital 0dBfs (0 deciBel full scale). Hopefully, the availability of higher dynamic range audio recording and playback system, and an increasing awareness among the consuming audience, will contribute to reversing this trend in the coming years.

References: [1]:

[2]: [3]:

[4]: [5]: [6]: [7]: [8]: [9]:

[10]:

[11]:

Bob Katz: “Integrated Approach to Metering, Monitoring, and Leveling Practices, Part 1: Two-Channel Metering,” Journal of the Audio Engineering Society, Volume 48 Number 9 pp. 800-809; September 2000 ”Music adapted to poor formats” – NRK music report 6/5-05. URL: http://app.nrk.no/musikk/4728287.html David Robinson: “Regulations in Digital Audio Broadcasting”, University of Essex, Comments on the Draft Communications Bill and the proposed powers of OFCOM: The Office of Communications. July 2002. “Quality in the age of good enough” – MIX Professional Audio and Music Production Magazine. URL: http://mixonline.com/mag/audio_quality_age_good/ Lou Rawls: “An Explanation of Audio Compression: Theories and Examples”. URL: http://www.whereismyhead.com/lit/compression_faq/ Rip Rowan: “Over the Limit?”, Prorec.com editorial, September 2002. Bob Speer: “What Happened to the Dynamic Range?”, Pro Audio Rx article. URL: http://www.proaudiorx.com/dynamicrange.htm Interview with Steve Hoffman, April 2005. URL: http://www.netassoc.net/dougspage/hoffmanapr2005.htm “The Big Squeeze: Mastering Engineers Debate Music's Loudness Wars”, Dec 1, 2005, MIX Professional Audio and Music Production Magazine. URL: http://mixonline.com/mag/audio_big_squeeze/ Siegfried Klar, Gerhard Spikofski: “On levelling and loudness problems at television and radio broadcast studios”, Audio Engineering Society Convention Paper 5538, Presented at the 112th Convention, 2002 May 10–13 Munich, Germany The European Broadcasting Union: EBU Recommendation R117-2006: “The use of high level digital audio material in the production chain”, Geneva, February 2006.