Mastering for Vinyl
LPs are staging a big comeback, which means there’s a lot to know before you send your music off for pressing.
By Scott Dorsey
There’s something of a small renaissance in LP production these days, and we’re starting to see a lot of people in small studios producing material for issue on LP for the first time.
It’s not just in one sector of the industry, either. The guys producing dance music for DJ use have never really given up on vinyl because their customers like the ability to mix and scratch the stuff, but the techno music explosion has also brought in a lot of people intending to cut vinyl. There is an increasing amount of jazz being released on vinyl, and a number of small audiophile labels cropping up that release primarily on vinyl.
This is a bit of a problem, though, for people who want to get into this growth. LPs aren’t like CDs at all, in that a lot of manipulation has to be done to fit your material onto the disc. So there are a lot of things to watch out for that most folks who haven’t mixed for LP release might find a bit odd.
How LPs are made
Now, this stuff is actually more important than you think. When you release on CD, you don’t need to know a thing about the pressing process, because you know that the bits that get sent out to the plant will be the same bits on the disc you release.
This is not at all the case with LPs, and you need to involve yourself in the process a lot more. You also need to know how things work, because the probabilities of something going wrong are great, and you will need to talk with the manufacturing people and understand what they are saying. So a lot of this stuff may sound completely irrelevant, but it’s important in understanding some of the limitations of the medium.
In the beginning, we start out with an acetate blank (also called a lacquer), a disc made of aluminum with a thin layer of acetate plastic on it. It’s placed on a mastering lathe, which is a sort of turntable with a cutting head (basically a huge oversized phono cartridge in reverse) on a bar that is suspended over the disc like a linear tracking tonearm.
The head may be cooled with water or compressed helium because of the huge amount of heat generated by the high torque coils in it that are required to cut smoothly through the acetate surface without distortion. This is despite the fact that the cutting stylus is heated to help it glide through the plastic without tearing.
Modern cutters also have a positional feedback mechanism, with a sense coil as well as a drive coil, and additional electronics in the cutting head amplifiers to handle the negative feedback input; this also helps compensate for any nonlinearities in the cutter itself.
Often you will have two or more acetates cut at one time so that one of them can be returned to you as a reference. You can put it on a turntable and listen to it, but be aware that the material it’s made of is much softer than vinyl and it will degrade slightly but audibly with every playing. If you ask the mastering house for a test acetate before pressing, be aware that it’s only good for very few plays if you are listening critically.
Once the acetate is cut, a fine layer of silver is chemically deposited on it to make it conductive, and it’s electroplated to form a negative relief mold of the original acetate made out of metal, called the metal master. From this, a positive metal relief called the metal mother is made, and then a negative relief called the stamper is made from that.
With each generation, some small amount of detail is lost, and as a result the final pressings never seem to sound quite as good as the original acetate. The electroplating and production of the metal elements is still more of an art than a science. Having it done properly can make a real difference in the tracking distortion and noise floor of the final pressing.
Because of the sheer number of generations involved in the process, a few mastering labs (such as Europadisc in NYC) use the Direct Metal Mastering process in which the acetate is replaced with a copper disc. That disc is used directly in place of the mother to make the stamper from, eliminating several generations.
However, the copper disc is more difficult to cut because of the hardness. It requires special cutting styli that require constant replacement, and (in the opinion of some mastering engineers) it tends to emphasize cutting head resonances. However, it can also produce a quieter surface, permitting you to cut at lower levels which makes all of these factors less of an issue, and of course there is reduced distortion because of the reduced number of generations. Some people love the process, others don’t.
Once you have the stampers, they are placed on a large stamping press, where one stamper is suspended above the other. A small “biscuit” of vinyl is dropped in the center, with the label on either side of the biscuit, and then the stampers are pressed together and rapidly heated with high pressure steam.
The 1/4-inch thick biscuit is squeezed out slowly across the stamper surface like silly putty, and a little knife edge pops out and trims the flash from the edge. After giving the stamper enough time for any gas or bubbles to escape, cold water is pumped through the same channels that the steam was piped through, and the new record quickly solidifies.
Then the whole thing is released and the record is popped out. If it’s not allowed to sit long enough before being cooled you get noisy disks, and if it’s not cooled down enough before removal you tend to get slight edge warps.
The composition of the biscuit has a lot to do with how the final record comes out. There are various religious wars that go on about what the best mixture of virgin vinyl to recycled vinyl is; it’s generally held that a small amount of recycled vinyl helps keep the noise floor down, but I can’t tell you why. I know for a fact that one of the major audiophile labels that advertises “100% virgin vinyl pressings” is only using about 90% virgin vinyl, but I don’t fault them for it because their records sound better as a consequence.
There are also various religious wars about which plasticizer is better and longer lasting, and everybody has their own secret formula.
Now, your pressing house won’t give you an opportunity to do much about this, but they will invariably give you a chance to select how heavy a biscuit (and therefore how heavy a record) you want to use. Thicker, heavier records are more expensive, but they tend to damp out resonances better on playback and there is an audible sonic difference in most cases.
Incidentally, if you are cutting 7″ discs, you may be given the choice of styrene as an alternative to vinyl. Traditionally, most 45s were cut on styrene, which is harder and more brittle than vinyl. Because of this, any disc resonances occur at higher frequencies, and this could give you sonic improvements.
Styrene discs definitely sound different than vinyl pressings from the same stamper, but wars have been fought over which one is better. I generally avoid styrene because when the discs wear out (especially if they are played with worn styli) they take on very nasty distortion that is much more offensive than that of a worn vinyl disc. Again, it’s a judgment call.
In the old days, vinyl pressings were issued to radio stations and for juke boxes, but the styrene injection molded disks were for retail. Occasionally, there have even been some budget 12″ discs on styrene (like some of the Decca and Harmony discs), but these days I don’t think anyone is still pressing 12″ styrenes.
Incidentally, if you didn’t order a reference acetate (as discussed above), you want to request a few test pressings to be made before you pay for the big pressing run, so you know what it sounds like. There are apt to be noticeable differences from the master, so doing one of these is really essential.
I like to get the reference acetate because that way if I want everything redone, I get the problem before I have paid for the plating. Other folks prefer to get a small run of 10 or 20 test pressings so they can pass them around for evaluation or even give them out as promotional items.
How they get stereo on these things
Back in the days of mono, the groove cut on the disc was a simple spiral, and it was shifted back and forth in the plane of the record with the signal (this is called lateral cutting). The engineer could adjust the groove pitch (that is, the spacing between each successive rotation) so that for loud passages the grooves were widely spaced and for soft passages they were tightly packed. This maximized the amount of time on a side, while still allowing a lot of dynamic range on peaks.
When stereo came along, there were a number of odd schemes used first off, but the industry quickly settled on what was called the 45-45 system (see Figure 1). With this method, the lateral movement of the groove carries the mono signal (that is, the L+R sum of the right and left channels), while the depth of the groove is varied to reflect the differences between the channels (that is, the L-R signal). This meant that the new stereo records could be played on older mono equipment accurately.
The reason it is called 45-45 is that the popular way of cutting this involved right and left channel coils in the cutting head, each of which were arranged 45 degrees from the plane on which the stylus was mounted, so that the sum and difference signals were generated mechanically.
Another way of looking at this is to think about the V-shaped groove of the record: the inner groove wall carries the left channel, modulated on a 45-degree tilt to the right, while the right channel is cut on the outer groove wall.
As a result, if there is a lot of information common to both channels, the groove is made deeper and shallower, while if there is a lot of information not common to both channels, it slides back and forth laterally. The net effect is that if you sum the left and right channels you get mono, but with the appropriate pickup you can get stereo.
However, there are some cutting heads with individual L-R and L+R coils, one mounted parallel to the record surface and the other mounted perpendicular to it. And there are a lot of mastering houses that still generate L+R and L-R signals for separate processing and then recombine them to L and R before sending them to the cutter, despite using the more common cutting heads that take L and R signals as input.
The disadvantage of the whole stereo LP scheme is that although the mastering engineer can adjust the groove pitch for loud or soft passages, the maximum and minimum depth of the groove are constant. Cut too much low frequency information with a wide stereo spread, and you get a lot of deep peaks and valleys in the groove and styli tend to pop out of the groove. Turn that down, and your stereo image collapses.
So the amount of stereo information has a lot to do with the level that can be cut to disc. No matter what you do beforehand, out-of-phase low frequency content will lift the stylus from the groove or drive it into the substrate. On the other hand, in-phase low frequency information causes lateral excursions wide enough to cut into the previously cut groove and into the area where a groove would be cut in the next revolution.
The mastering process
Mastering today has become a catch-all term for any kind of post-mixing audio processing, but LP mastering is the process of making an acetate out of the original tape. The processing is secondary. However, the processing is almost essential to get the most out of the limited channel.
There is a lot of poking and prodding that is often done to get the stereo signal to fit into place, because the LP has less information on it than the original master tape does. Often, you’ll see mastering engineers roll off a lot of the very low bass and add a false bass peak around 200 Hz or so, just to compensate for the mechanical limitations of the equipment. The other alternative is to reduce the running time per side radically.
The one thing that saves us from bass being a big problem is the RIAA pre-emphasis curve. Most of the noise in the recording process is at higher frequencies. So on record, we pre-emphasize the signal by pumping up the highs, and then on playback the phono amplifier has a roll-off curve that is the exact inverse of the curve in the record chain, which rolls them off. This means that the music has the same frequency response, but the noise is reduced, primarily on the high end.
Even with this, though, the mastering engineer is constantly juggling signal processing versus recording time versus groove pitch. Most systems today automatically control the groove pitch, although an expert engineer can override them to some extent and make constant tweaks to get that last bit of performance out.
Traditionally, the way this is done is with a “margin control” system. The tape is played back on a machine with a special “pre-hear” or “pre-listen” head that picks up the signal about half a second before the playback head (i.e., for 30 ips tape, the head is about 15 inches away from the playback head), and feeds that signal into some control electronics.
This means that that control system has information about what the signal level is going to be like on the next rotation of the disk. It can constantly ride the groove pitch up and back so that the grooves don’t get too close that the walls between them get deformed. But they are still as close as possible so that the maximum running time on the disk occurs.
This, incidentally, is why 12″ singles are invariably cut much, much hotter than conventional LPs. There is plenty of space for very wide groove spacing, so they are cut as hot as possible and therefore play back much louder. Some of the 12″ singles are even cut constant pitch, without any margin control, because there is just too much safety margin available for the mastering engineer.
When working with a digital signal source, a digital delay line today usually replaces the deck with the pre-hear head, but still the overall principle is the same.
Another issue here is that the frequency response of the disc is different in the outer grooves and in the inner grooves, because the stylus is moving much faster across the outer grooves (it goes through a greater distance per revolution). This means that the mastering engineer may have to tweak the high end response up progressively during the cutting process.
It is a good idea to leave a large blank area around the label, because the very inner grooves have serious sonic problems. Not only do they have high frequency loss, but the tracking distortion is a lot higher, especially as the records wear. The less time you try and get on a side, the more space the engineer is able to leave there to keep things clean.
The number one thing to remember when you are mixing down for vinyl is that the L-R and L+R signals are being separated off, and you want to keep as little unneeded information on the L-R channel as possible. This means you want to try and keep everything identical between channels.
The easiest way to do this is to monitor in mono as well as in stereo; check the mono compatibility. Make sure your monitoring system has a mono button so you can listen and see what is happening. If instruments drop out or the tonality of anything changes, you’ve got a problem.
I once had a fellow who had a graphic equalizer he routinely mixed down through, directly before his DAT deck. He never noticed any problems when getting stuff issued on cassette. But when he brought me the master for cutting an LP, I found a phenomenal amount of differences between channels because it was impossible for him to set both channels absolutely identically to one another. Because of this, I had to cut the record about 6 dB quieter than I otherwise would, as there was just so much stuff that wasn’t common to both channels.
All the stuff that is mixed to the center will appear in the difference signal if you have even minor differences between channels. And remember that the L-R signal has less dynamic range than the L+R signal in most cases, so it’s usually the limiting factor for the mastering engineer.
Therefore, the number one rule when mixing for LP is to get all of that processing gear out from between your console and your recorder. Let the mastering lab do all that stuff.
If you need to add effects and processing, try and do it in the individual channel and try to avoid stereo effects. If you have a phase meter on your console, learn to use it. If you want overall equalization or compression, write a short letter to the mastering house and tell them. If you don’t want overall equalization or compression, tell them that also. They have special mastering equalizers and compressors that have click-detents and that are calibrated precisely to be identical between channels when set identically.
Some mastering places split the signal up into L+R and L-R components and process them separately, which gives them additional flexibility. Use your equalizer and compressor for monitoring, so you can tell what you want it to sound like, and let them duplicate (and possibly improve on) the effects in the mastering room.
The number two rule is always try to mix sources with heavy bass content into the middle. If there is a difference between channels on a signal with a lot of bass, you get bass in the L-R channel and the groove on the record becomes very shallow at points. It will have a tendency to pop out of the groove, unless the mastering engineer either limits the bass or reduces the level. If you want the disc loud, keep all the bass in the center.
Remember that if there is a loud bass sound in the L+R signal, the mastering engineer can just reduce the pitch of the record at that point so there is more room between grooves and more space for that sound. But he cannot do anything about large bass signals in the L-R signal other than just turning everything down.
For minimalist recordings, you want to try and minimize large phase differences between channels for precisely the same reason. This means that spaced omnis are really not such a good idea if you can avoid them.
If you can’t avoid them, try and put loud bass sources in the center of the soundstage, as close to the center mic as possible. Even if you are using coincident miking, this is a good idea.
I do strongly recommend near-coincident techniques like ORTF or baffled omnis, because they will give you good imaging while still making the mastering engineer’s job easy. (I’ll recommend them for plenty of other reasons, too, but that’s the subject of a different article altogether—see elsewhere in the library.)
The number three rule is to try and keep each side short. The longer each side is, the closer together the grooves have to be and the lower peak levels the mastering engineer has to work with. The hotter he can cut the record, the more real dynamic range you have available to you.
Also realize that as discs get banged around and poorly treated, the noise floor increases. Unlike CDs, you have to expect that someday they aren’t going to sound as good as they do coming out of the plant, and they’ll be covered with some DJ’s fingerprints and mustard from his lunch. So give yourself as much margin as you can.
There are a lot of judgment calls in the LP mastering and pressing process because there are a lot of variables that have audible effects. This is a real shock to someone who is used to cutting CDs, where one can send off a tape with directions to cut it flat and expect something back from the duplication house that sounds identical.
This, in short, is what makes audio production an art, and what makes it so much fun.
LPs are not perfect sound forever. So what did you want?