Scott Hull on Vinyl, Part Ten

Header image for Scott Hull on Vinyl series

How many grooves are there on a typical record?

The RIAA (Recording Industry Association of America) set all the parameters for the dimensions of the modern vinyl record. The parameters needed to be made consistent so that player functions would all work. So, as a mastering engineer, I need to know that the final locked groove on an LP (33 1/3 rpm) must be at a diameter of 3.875″ (give or take 1/32″). And lots of other details. (See the disk diameter chart from the RIAA, below.) Lets look at the parts of the disk surface.

RIAA diagram illustrating the surface of an LPThe Safety Groove

The outermost grove is automatically cut a little deeper and wider than standard, and its purpose is to catch the needle if it’s manually placed on the record too near the outer edge. If you let the cartridge bounce off your turntable it will almost always cause some damage to the delicate stylus/cartridge.

The Lead-In Area

The lathe carriage — the part of the cutting lathe that moves the cutting head across the surface of the disk — moves at a fast rate in the lead in. There is some blank area there on the disk that must not have audio recorded. The reason for this is that automatic record changes would not always drop the needle precisely. The grooves in the lead-in and the safety groove did their best to keep the needle on the record.

The lathe carriage, driven by the lead screw, then slows down to standard pitch for about 3 seconds. Then and only then is audio supposed to begin. From this point on, the pitch of the grooves (how far apart they are) is controlled by the computer in the lathe. The pitch drive computer listens to a preview audio signal that comes 1.8 seconds before the audio. It’s that far ahead because that is about how long it takes the record to make one revolution at its outermost diameter. Between songs we press a “spiral” button which advances the carriage quickly for just a moment. This creates the visual band between the songs so you can see where to place the needle.

This part gets pretty technical…

Lets look at the process of cutting the groove in the first band of an album. And let’s assume for simplicity that the grooves of the left channel face towards the center of the record and the grooves of the right channel face the outer edge of the record. A modulation on the left channel moves the groove into the “virgin” area of the disk that has yet to be cut, while a modulation on the right channel moves the groove into the part of the disk that has already been cut. So to keep the grooves from colliding, the computer has to calculate how it has to turn the lead screw to avoid cutting over a previously cut groove. This happens very fast, and it’s hard to see with the naked eye, but we can monitor the progress of the groove by watching a meter on the front of the lathe. It’s calibrated in Lines Per Inch (lpi) (see photo below).

This, logically, is the number of grooves (lines) that are cut in an inch of the lateral record surface. The computer then has to store the left channel information into memory, and add that to the right channel information that is coming up on the next revolution. You see, the collision that has to be avoided is between the left channel of the first grove and the right channel of the second groove. If you make a little drawing of a disk and a squiggly groove you will see what I mean. In real time, the lead screw motor has to turn fast enough so that when the next groove comes around there is enough room to cut the groove and still leave a tiny bit of “land” between the grooves.

Photo of an LPI meterLevel and Duration

Very early lacquer lathes cut at a fixed pitch. There was no computer control. With these lathes it was virtually impossible to cut a 20 minute side of pop music with a reasonable level. It took the variable pitch lathe to cut a better sounding record — as long as you didn’t let the grooves collide.

We align our cutting system with a basic geometry assumption. We adjust the cutting parameters so that a 2 millimeter-wide groove cut with 600 lines per inch should produce no land or open space between the grooves. From that baseline, any audio that is present causes the groove to wiggle and requires that the pitch be lower than 600 lpi. Does this make sense? Ok… More music = fewer lines per inch. So the louder the music the less space to record the audio. There is a direct relationship between level and duration.


A couple other factors cause us to increase the space between the grooves. If we have audio that causes a very challenging groove to be cut, we may need to momentarily increase the land between the grooves (thus lowering the lpi) to give the grooves a little extra space. This is only for insurance, but a good practice when it’s possible. Also there is a peculiar effect when cutting into lacquer. The disk is rather soft, and it’s being cut buy a heated stylus. But what happens after the groove is cut is what is interesting. Being a “plastic” substance, the lacquer partially springs back to it original shape after being cut. Not entirely, of course, but enough to cause the neighboring grooves to be affected. This “spring back” or elasticity can cause audio to “ghost” into neighboring grooves. This is referred to as groove pre-echo and it very hard to deal with when there are soft passages followed by very loud sounds or visa versa. The loud sound can be heard one full revolution before or after the audio actually happens. Sometimes even both. Many of you have probably heard this and probably wondered why echo would have been added in the production. It wasn’t added in the production studio. This groove echo was caused by the disk cutting process itself. To avoid groove pre echo we open up the spacing of the grooves right before any sudden loud passage and right after any loud passage that stops suddenly.

Analog Tape Print Through

There is one more complication. Analog tape recording has a similar effect called print through. This isn’t due to the tapes elasticity, but it’s due to the magnetic properties of the tape. One layer of magnetic tape laying against another layer of tape tends to give off a small portion of its magnetism to its neighbor. The louder sound will “travel” up and down the packed reel of tape. This effect gets worse with age. The longer the layers are sandwiched together the more of the energy is transferred. Fortunately for records, once that master lacquer is plated in the pressing plant, no more echo can happen.

As for the question at the top of this post: as it turns out there are TWO grooves on any record. One on each side. And if you stretched one of them out it would be 1600 feet or about 1/3 of a mile long.

Fill out my online form.