AES Section Meeting Reports

Los Angeles - March 25, 2014

Meeting Topic:

Moderator Name:

Speaker Name:

Meeting Location:


On Tuesday March 25 the LA Section gathered to hear from mastering engineer Doug Sax on the art of vinyl. Before delving into the technical aspects of cutting records, Sax confessed to his own astonishment at the format's perennial relevance. Even upon opening The Mastering Lab's doors in 1967, in the midst of vinyl's heyday, he didn't imagine that in 2014 anyone would be talking about, much less buying and listening to records en masse. Nevertheless, current market statistics do not lie: last year, 28 million LPs were pressed, and Pro-Ject alone sold 250,000 turntables. These numbers fall amidst a somewhat longer term trend. Mashable reports that since 2011, even in the face of a boom in digital download sales, record sales have increased by 17.7%. Given this context of the format's resurgence, Doug's presentation would constitute a consideration as to why vinyl still exists, and indeed thrives, as much as a considera-tion of the details involved in creating a quality record.

First off, Sax posited that this growing interest, if not preference, cannot be attributed solely to differences in audio quality between vinyl and other available audio formats. The advent of mp3s and other types of digital music has changed not just the way audio is stored and distributed, but also when and where people listen. iPods have long placed upwards of 10,000 songs in one's pocket, meaning an extensive music collection can be taken anywhere. Audio has thus become ubiquitous, almost impossible to escape. The choice to listen via vinyl played back over speakers is a choice, in some part, to terminate other activity, and actually focus on the music. Vinyl therefore offers an altogether different listening experience to younger generations brought up in a time when portable music formats have become dominant.

Getting into the history and characteristics of vinyl itself, Sax asked the audience which invention had to precede the acceptance of records as a viable medium. His answer: the tape machine. While many major studios worked hard to develop different versions of vinyl, such as the 45', the fact was no combination of characteristics of the medium itself would persuade the majority of musicians to sit for a continuous twenty minute, straight-to-disk re-cording session. Obviously unlike with tape, or even with the motion pictures with which records were originally paired to introduce sync sound to the movies (the origins of 33 1/3 rpm lie in the need to sync a record to film with the Vitaphone system), one could not simply splice together pieces of vinyl. It

was the widespread availability of tape recording that enabled the rise of vinyl as a distribution format. That said, speaking to some of the advantages of vinyl over other formats, Sax emphasized the importance of phase. The tape machines which allowed for the rise of vinyl require the use of flattening filters, and even in the case of modern digital audio, recorded at the highest sampling rates and bit depths, the A/D and D/A conversions introduce two stages of filtering, which Sax said effect the way the sound comes at the listener. In the case of direct to vinyl recording and then later reproduction through a turntable, the signal passes through the RIAA curve and its inverse. Plotted against a logarithmic frequency axis, the cutting side roughly runs a line from -20 dB at 20Hz to +20 dB at 20 kHz, hitting 0 dB at 1 kHz. Reduction in bass means that narrower grooves can be cut into the vinyl, thus freeing up space for more music on each side. Upon playback the inverse curve is applied.

According to Sax, because the playback and record curve mirror each other exactly, phase distortion is minimized, if not eliminated. Coupled with this optimal phase characteristic found in straight to disk productions, in the more general case Sax observed that the cutting process brings the input audio in line with the 'laws of nature'. He explained that, essentially,

committing the signal to this physical medium incurs a type of preprocessing. Further down the playback chain, the signal will be better suited for reproduction by drivers, themselves of course physical objects. Getting into some of the finer technical points of the art of cutting records, Sax introduced the idea of variable pitch. Cutting each rotation around the record to the same width, a record might have space for say, 20 minutes of audio on each side. In fact, that figure assumes a fair amount of dynamic range, an assumption likely invalid for much of modern popular music. Keeping with the 20 minute baseline, however, and holding all else equal, employing variable pitch cutting, wherein grooves representing softer passages are cut more closely, five additional minutes per side or more might be gained. Interestingly, dynamic range can play a part in track sequencing when it comes to vinyl. Louder tracks requiring deeper cuts, for example, might be aligned on the flip side of the record with quieter tracks.

At this point Doug introduced Robert Hadley, Senior Mastering Engineer at the Mastering Lab, to describe some of the key problem areas of record cutting. First and foremost, given the considerable boost applied by the aforementioned RIAA cutting curve, high frequency content, specifically sibilant content, can wreak havoc. Hadley said that while many might quickly reach for a de-esser in response to such issues, doing so introduces another piece of gear through which all content, good and bad, must pass. In many cases not every 's' or cymbal crash will cause issues, and so usually he instead prefers to take a few passes with EQ to fix specific spots. On the low end, phase incoherent bass can lead to extremely wide grooves, in turn all but ensuring jumping needles. Blending or monoing below certain frequencies can alleviate such instances.

Sax then returned to offer a few more general observations on both cutting and recording straight to vinyl. Patience, he said, is fundamental. Of course in all types of mixing repeated listening is crucial, but in addition to ensuring the audio fidelity, as far as rendering vinyl is concerned, one must also ensure that reproduction devices, i.e. needles of a range of quality, can play through the entirety of the record without tracking problems. In some aspects this time consuming process stands in stark contrast to the practice of actual recording direct to disk. While the widespread use of editing software today gives musicians and mixers fine control over every note, if not eve-ry last sample, recording live and direct to disc compresses all of that micromanagement into the span of a few hours. To be sure, not the easiest way to cut an album.

Moving on, Sax brought to the podium Carnegie Mellon graduate, violinist, and Mastering Lab engineer Eric Boulanger to provide a look under hood of some of the equipment involved in cutting. Starting with the basics - the cutting head is stereo (thus requiring a stereo amp: mono jobs sum the channels just before the amps )- Boulanger then walked through a more detailed look at the cutting lathe. In identifying the input section, which applies the RIAA curve, Boulanger mentioned as an aside that modern noise shaping techniques, which force energy up into normally inaudible frequency bands, actually come into play here. Typically a lathe can cut flat up to nearly 50 kHz, meaning that energy shifted upwards as noise shaping can come back to burn up the cutter if not addressed.

The cutter head itself acts as both a speaker and a microphone. Two pistons connect to a stylus, and each channel has its own drive coil, the speaker part. Additionally a feedback coil nulls out any resonance, and helps to elim-inate any coloration due to the cutting head, thus constituting the microphone part.

Though Doug mentioned the fact that historically variable pitch had been executed by hand, the Mastering Lab lathe makes use of Jerry Block's Compudisk to do it automatically. Two CPU's work together, one converting the analog signal to digital at a rate of around 1 kHz, and calculating adjustments, the second relaying these moves (lateral and vertical motion, along with incorporating appropriate geometric considerations) to the lathe's peripherals. Unique to the Mastering Lab's setup, and an innovation of Doug himself, the cutting head is actually balanced across a pivot point, with a magnet controlling vertical motion. Conventional lathes will simply press down to create more depth, but in this system, the cutting occurs due to the weight of the cutting head, nothing more. In practice this means that if the raw vinyl material is non uniform, the cutting will actually roll over these imperfections, rather than plowing through them and reducing playback difficulty.

A lengthy question and answer session followed the talk. Many of the questions considered recording to vinyl from the perspective of the artist. Is there a benefit to staying analog throughout recording, up to the cutting stage? In light of the high frequency issues covered, should artists keep the possibility of eventually cutting material to vinyl in mind as they mix? Addressing the former question Sax made the point that while end to end analog might improve the chances of a satisfactory outcome, really what makes a difference on most cases are the mixers (that is, the human mixers) involved. Equipment will not save a mix. As far as the need to keep in mind the characteristics of vinyl and the cutting process in shaping a mix, Doug leaned toward the opinion that professionals should be able to work with a variety of mixing styles. He cited the markedly bright records of Earth, Wind & Fire as an example of extremely bright albums, which, nevertheless proved extremely successful on vinyl.

On some final technical notes, questions arose regarding the value or purpose of making use of 45, rather than 33 1/3 speed, and the differences between record weights (150, 180 gram etc). It turns out that as needles track closer to the center of a record, their velocity decreases. 1.8 seconds around the outside of the record is not the same as 1.8 around the inside. At these slower speeds, there is more weight pressing on the vinyl, and it is more difficulty to run through all of the grooves. Playback at 45 can mitigate these problems, and in fact many older albums are being re-pressed to 2 or 3 45 rpm records in part for this very reason. Doug also contends that equally important to the weight of the vinyl are the metals used to press the records.

The Section would like to thank Doug Sax, Robert Hadley and Eric Boulanger for taking the time to educate us on the art and craft of vinyl mastering.

Written By:

More About Los Angeles Section

AES - Audio Engineering Society