How to Read Loudness Meters
Sunday January 28, 2018. 12:33 PM , from Audiotuts+
“Beautiful Loudness” by Orin Zebest is licensed under CC BY 2.0It's no wonder there is much confusion on the web with the subject of loudness meters and the acronyms associated with them. They are are all in fact supposed to do the same job; that is, indicate a level of loudness. The loudness war has been a hot topic over recent years, especially with mastering engineers. The loudness war, however, is ending. This is due to digital music services such as Spotify that has recently reduced the playback volume to -14 LUFS.In this tutorial, I'll explain the meaning of LUFS and some other important meter readings, including:Peak & true peakdB & dBFSRMSCrest factorDRK-SystemPeak & True Peak A peak meter is often the one we're first introduced to. A peak meter instantly indicates the level of the audio passing through it in dB (decibels). Audio above 0.0dB will often result in clipping. A true peak meter indicates the peak that occurs once the signal is converted from analogue to digital. These are known as inter-sample peaks, that are often higher than peak volume (around 0.3db higher). This is why mastering engineers often ensure that no peaks are created above -0.3dB so as to avoid clipping when the signal is converted from analogue to digital. dB & dBFSdB is the acronym for decibels, while dBFS stands for decibels full scale. Both are used for measuring peaks. The term full scale is not used when indicating the difference between two measurements. For example, -20dBFS and -15dBFS is a difference of 5dB. RMS (Root Means Squared)While peak is the highest level, RMS is the average peak. It's the average peak that that the human ear associates with the loudest section of a piece of audio, rather, than a sudden peak. RMS meters are, however, far from perfect as different meters give different readings. This is because RMS isn't just a function of level but a function of time too. RMS meters often have different response times from each other, resulting in different readings. Crest FactorThis is the distance from the average peak to the highest peak. For example, if a song has an RMS of -14dBFS, it peaks at -1dBFS, it would have a crest factor of 13dB. DRDR, or, dynamic range, is the difference between the loudest (highest RMS value) and the quietest (lowest RMS value). K-System The K-System was introduced by Bob Katz in an attempt to standardise levels to overcome the loudness war. The K-System is an RMS meter that displays a target level that is displayed as 0.0dB. There are three different scales: K-12, used for broadcast materialK-14, commercial recordingsK-20, recordings that require a wide dynamic like classical musicIf you play a constant sine wave that is registering -2dBFS on a K14-meter for example, this would give you 16dB worth of headroom. Headroom is the distance from peak to 0.0db.LUFSLUFS is the new standard for measuring loudness and considered the most accurate. LUFS standards for Loudness Units Full Scale. There are three types of LUFS readings: 1. LUFS—MomentaryLUFS momentary gives the current loudness level just like RMS. The difference, however, is that the LUFS takes into account the Fletcher-Muson Curve by applying a slight high-frequency boost giving a reading more suited to humans then computers. For example, a track that has a loud low-end versus a track that has loud high-end, but reads the same on a RMS meter, is not going to be the same to our ears. The one with the louder high-end will be louder. LUFS recognises this and compensates for it. 2. LUFS—Short-TermShort-term is an accumulation of 3 seconds that represents the average loudness level within that time. 3. LUFS—IntegratedIntegrated is a long-term measurement that gives you the overall level of the complete audio track or until you press reset. You do need to employ a little common sense with long-term measurements. For example, using LUFS integrated for a song that has 4m of quiet audio and 30s of ear piecing volume, and then trying to match that meter reading with audio which has 4m and 30s of uninterrupted loudness is just pointless. My tip would be to meter 20s of the loudest section and 20s of the quietest to give a more workable reading for comparison. Final ThoughtLoudness meters should act as reassurance, like asking a friend, and should never take place of the ear. Try importing a few songs into a digital audio workstation and loudness matching them with your ear. Check your judgment against a LUFS reading. Consider the accuracy of your ear. Or maybe the consideration should be the accuracy of the meter.