Re: [ARSCLIST] Mass Digitization/ 78's

[George Brock-Nannestad <pattac@xxxxxxxx>]

From: Patent Tactics, George Brock-Nannestad

Don Andes took the challenge, and Mike Richter, David Lennick, and Steven 
Barr commented:

> As some of you may have already guessed, I'm open to crazy ideas, so I'll
> pursue the logic of creating 78's as an archival medium, since they have
> been hanging around in vaults for a good long time.
> 
> But I have the following questions in regards to 78's:
> 
> What is the maximum uninterrupted record time?
> What is the maximum fidelity expectation?
> Could the format possibly handle multi-channel recordings?
> If there any way to embed a video track (if you will) to address audio that
> is required to sync with a video source?
> Could you possible embed a data track to handle metadata since Physical
> labels have limited space, can fall off, or become obscured?
> In today's world of chemical regulations, is it still legal to make a 78
> record, or does include some gaseous by-product, or radioactive waste?
> 
> Without viable real world answers to ALL the questions above, I believe I
> for one would take the idea of using 78's as non-viable.
> 
> Any takers?

----- I am a taker:

A direct answer to your questions does not provide suitable information for a 
decision maker, because they would address limitations that must be 
considered artificial today.

First of all I would think that the expression "78" is just a symbolic 
expression for "mechanical modulation". The fixed rpm value 78 merely 
established itself as a relevant value in the period ca. 1894-1962 (in the 
Western world). This was in the period in which the material contained 80% 
slate dust, held together with a suitable binder; from a manufacturing point 
of view it was shellac for a very long time.

Joseph Maxfield established the relationship between uninterrupted record 
time, maximum amplitude and rpm in 1925, but that was when only a fixed 
groove pitch was available. With variable pitch (that was commercialized in 
the 1950s), you could get e.g. 15' on a 78 rpm side, dependent on recording 
level.

The larger the volume taken up by modulation is, the safer the signal. 78s 
have coarse grooves, i.e. a large volume, and two huge flanks, each 100 
microns (4 thous) in "height". This was required when the tracing stylus was 
a steel needle. Once the microgroove came along ca. 1950 (known as 33 and 45 
respectively), all development of coarse groove pickups stopped, and the 
efforts were thrown at the new format. Many modern pickups are quite able to 
cope with the required recorded velocities for 33 and 45, but they are no 
longer able to provide a linear velocity signal when the recorded amplitudes 
approach those of coarse groove records. As an aside, this is one of the 
reasons why some click and crackle removers have problems: when the noise 
impulses are traced near the max excursion, the recorded velocity (that is 
the signal we want) is very low, almost zero, but the representation of the 
noise waveform has suffered distortion and is more difficult to handle. A 
disco pickup is much more capable, as it has to deal with a 45 rpm disco 
single - which has coarse grooves in stereo!

So, if we now move from the high and safe rpm 78 and to the standards between 
ca. 1950 and - today, i.e. 33 1/3 rpm and 45 rpm, there are many things that 
can be done. If we accept sophisticated data and signal processing when we 
transfer to digital, why should we not also accept this for mechanical analog 
recording? If we open that field, we can obtain all kinds of multichannel and 
synchronization, some of it using proven recording technology (SQ and other 
four-channel recording systems did work, but the mechanical reproduction was 
difficult), and using state-of-the art (vintage 1987) optical replay (such as 
the ELP Laser Turntable). Variable speed to provide linear surface speed 
under the writing and later the reading head could be synchronized by a 
wobble-like signal (like in the pre-groove of a CD-R). 

AS you will note, the field is rich for standardization; on the other hand, 
if this is professional only, why bother. The digital solutions made 
available today are definitely not adhering to international standards; 
something that will be sorely felt if re-formatting (migration) has been 
pushed too late for the endlessly-happening overlap window.

The record material is one interesting area. The idea of using slate dust (as 
the "loading") with shellac binder (and some lubricants and the inevitable 
carbon black) was to make an aggregate that worked in principle like 
concrete, only on a smaller scale. Good aggregate requires both large 
particles and smaller and very fine particles, in order that it can pack and 
carry a load. For steel needles providing a side pressure in the groove this 
was an exellent idea. When all modulating forces were vertical, as in the 
Edison Diamond Discs, it was possible to use not only very fine grooves in an 
unloaded phenolic material but also huge pressures requiring a polished 
diamond - but the wear was still tolerable. I feel very confident that we can 
find an environmentally acceptable material as a binder and obtain a fine 
surface that will not be worn, because of optical replay. So, our only worry 
will be the environmental maintenance of the surface. Still, if we are only 
considering a professional format, what is the matter with storing nickel 
mothers? A negative is much too sensitive, because the information is lifted 
above the plane, but in a positive all the information is below the surface 
called the land. Copper mothers have also been demonstrated to have a very, 
very long life, provided they are given a suitable coat that may be removed 
by steam or the similar. This was done in matrix stores in India in the 
1910s, to protect the investment from corrosion.

The metadata is another area: a data track is only relevant if we are now 
discussing digital signals stored without need for migration because it is on 
a very durable carrier. In a pressing of an analog recording an embossed 
label would be quite sufficient, based on engraving the original metalwork. 
It will not fall off. On a nickel mother, there could be direct engraving or 
laser ablation in the central surface of the record that is not used for 
recording signal.

All the above addresses the question of creating an analog system for long 
term preservation and retrieval. Those recordings that are already in the 
suggested format (nickel mothers, protected copper, shellac/slate dust) 
"only" need a reasonable environment for their survival.

I feel that it has not been in the manufacturing (including the record 
industry) industry's interest to provide backwards compatibility when they 
are creating systems to enhance the value to customers (and themselves). We 
have been dependent on the manufacturing industry for low-cost solutions, the 
larger the scale of manufacturing the lower the cost. This development now 
comes back to bite the record industry, and we cannot even say "serves them 
right", because they are the holders of a large part of the commercial 
recorded history. And if they give up it may be lost. Preventing others from 
giving access by means of draconic copyright measures adds to the insult.

If you have read this far, you may still dismiss my technical arguments as 
"speculative". Let me say that I have cut, and still do, lacquers, waxes, 
Direct Metal Masters, and cylinders, both by the electrical method and 
acoustically, using recording horns and soundboxes from the period in 
question. I have etched grooves, Berliner style (using copper and not zinc, 
because of crystal boundaries). I have full control over the parameters, 
which is more than can be said for recording a CD-R. And I should mention 
that Cornell University in their wild-life sound work regarded wax mastering 
as much better than lacquer as late as 1960 (personal information from Dave 
Wickstrom). The frequency range requirement for bird song is high. The only 
reason that mechanical recording went out of fashion is that the novelty had 
worn off after 90 years.

I hope that there will somewhere be a forum where realistic approaches to 
analog long-term storage without the need for migration may be discussed. 
Cost calculations are possible on every level, so we can effectively move 
away from daydreams. The amount that may need storage is daunting but that is 
no different from all other long-term documentation of human endeavours.

Kind regards,


George