Session Notes 6 - The Vinyl Frontier (Part 3: dynamics, levels & stereo)
Updated: Apr 6
Hello good people. This month I’m going to take a look at three more important factors when mastering for vinyl records: dynamics, levels and stereo width. Interestingly, loud records with a large dynamic range make the stylus wiggle about so furiously that the temperature at the stylus can reach a staggering 240°C. The melting point of record grade vinyl is 248°C. The louder a record is the more movement you get at the stylus. Remember that with a cymbal hit (especially Ride cymbals) you are generating up to 20Khz. That’s 20,000 backward and forward oscillations per second. This will generate an enormous amount of heat considering the tiny dimensions of a stylus which is typically 25μm (25 millionths of a metre). So ... hey ho, let’s go.
1. Dynamic range. This is simply the range of the quietest perceptible sounds to the loudest. Vinyl has an average dynamic range of about 60dB as opposed to the CD which has 96dB. I shall refer to it as “DR” from now on. There is only so much “loudness” you can cut to vinyl and at the opposite end you have the surface noise which will mask the extremely quiet sounds. In a lot of cases vinyl is cut from CD masters that have been limited (by a type of compressor known as a limiter) so that the average DR is quite small across the whole album. The masters that I prepare for vinyl have a much larger DR than CD masters
due to the fact that I allow the peaks to reach their fullest extent without over compressing. The width of the groove is on average 40μm to 80μm. Amazingly, a sine wave at 1Khz cut at 0dB will induce an acceleration in the cutting head that can reach 90 G’s: 90 times the force of gravity. DR is also related to the way humans hear sound. A certain amount of masking will occur. For example you won’t be able to detect a whisper during a Motörhead gig. The whisper will actually exist but you won’t perceive it. There is a “Dynamic Range Database” site that you can visit: dr.loudness-war.info. There are over 130,000 albums rated for DR but in a lot of cases the numbers don’t always relate to the user experience. DR is a perception rather than a mathematical formula. If it sounds dynamic and exciting then it is, regardless of the numbers. 2. Levels. As mentioned, many vinyl records are cut from “mastered” CDs. The trouble with this is that many modern CDs are mastered very loud. In almost all cases where I have mastered CDs loud it is a client or label request. They want their music to sound as loud as the next artist regardless of the consequences. The consequences unfortunately are that the music will suffer. I have cut some of the loudest CDs you will ever hear, where the RMS (average levels) are louder than the peaks. Believe me, I had one request where the artist stated that “I wish to see no movement on the meters”. Now, cutting lathes don’t really like these punishingly loud records. Even if you turn down the gain at the analogue stage you’re still left with the punishing sound. So, when providing cutting engineers with audio destined for the plastic platter keep the levels low: between -8dB to -4dB. Keep the dynamics intact as much as possible and let the cutting engineer work their magic.
3. Stereo. If you look lengthways down the groove of a record there will be a V shape with the groove walls at 45° from the vertical. The wavy pattern on each wall will be the stereo waveform for left and right channels. With excessive stereo panning and out of phase material this can cause the stylus (and the lathe cutting head) to move in undesirable ways. We want the stylus to be able to track the waveform without popping out of the groove or skipping into the previous groove. I quite often narrow the stereo image by about 5% to give the cutting head a chance to cut a useable groove. Very hard panned percussion, for example, does not bode well for the cutting head as does very phasey (fake “out of the speaker” effects) synth sounds. If you have a well balanced mix with nothing too extreme in the panning of L & R signals you shouldn’t run into too many problems.
Next month I’ll look at hacking MiniDiscs ... ooops, sorry, I mean lacquers and pressings.