Many audio recording systems are based on a concept known as “pre-emphasis” and “de-emphasis”. This is a process where a signal is distorted (here, I use the word “distorted” to mean “changed”, not “clipped”) at the recording or encoding process to counter-act the effects of something that will happen at playback. One example of this is a RIAA equalisation that applies an overall bass-heavy tilt to the frequency response at playback, and therefore the signal is given the opposite tilt when it’s cut onto the vinyl master. Dolby noise reduction for analogue magnetic tape follows a similar philosophy.
Another type of intentional distortion applied to an audio signal is based on assumptions of what happens at playback. Mixing engineers for television often emphasise lower frequency bands, assuming that everyone’s television loudspeakers needed some help. Pop and rock recording engineers check the mix on a low-quality mono loudspeaker and may make adjustments to the mix – to make sure it survived a clock radio or a portable Bluetooth loudspeaker (depending on which decade we’re talking about). Stereo vinyl records can’t have big low-frequency differences in the two audio channels otherwise the needle will hop out of the groove, so they’re mixed and mastered accordingly.
I’ve been reading “The RCA Victor Dynagroove System”, by Harry F. Olsen, published in the April 1964 issue of the Journal of the Audio Engineering Society. In it, he describes the entire recoding chain, including something that piqued my interest called a “Dynamic Styli Correlator” which is a distortion that is applied to the audio at almost the last stage of the signal path before it reaches the cutter head of the lathe that creates the lacquer master. You can see it here in Figure 1 from the article (I drew the red box around it).
Cool name; almost worthy of Dr. Heinz Doofenshmirtz (although it’s missing the “-inator”). But what is it?
One of the problems with playing back a vinyl record is that the shape of the needle on your turntable is not the same shape as the cutting stylus on the lathe. Consequently, the path that the needle tracks is not exactly the same as the path of the stylus. The result of this mis-match is that the electrical input signal that is used to make the master (the original recording) is not the same as the electrical output signal that comes out of your turntable (what you hear).
The idea behind the Dynamic Styli Correlator was that the actual path of the playback needle could be predicted, and the groove cut by the stylus could be modified to ensure that the output was correct. In other words, the distortion caused by the playback needle was estimated, and a distorted groove was cut to make the needle behave. This is shown graphically in Figure 29 of the article:
This is a great idea if the system works and if the prediction of the playback needle’s path is correctly predicted. However, neither of these two assumptions is guaranteed; so a number of things can go wrong here, and if anything can go wrong, it probably will.
However, it does mean at least as a start, that if you play an old RCA Victor Dynagroove record with a stylus shape that wasn’t invented yet in 1964 (say, a contact line stylus made for CD-4 Quadraphonic records, for example). Then you might wind up doing a much better job of reproducing the distortion that RCA created in the first place, instead of what they thought you were supposed to hear.