[78-L] Late 1930s ARC/Brunswick sound quality, was Acoustic/electrical recordings....

[Royal Pemberton]

On Fri, Jan 8, 2010 at 9:57 AM, Julian Vein <julianvein at blueyonder.co.uk>wrote:

> Dan Van Landingham wrote:
>   <snip>Brunswicks.I felt that their recordings in
> > the mid to late thirties were horrible-I mentioned a 1937 Brunswick I
> have of Gus Arnheim's
> > orchestra doing Shubert's "Serenade" and it was truly awful.
> ====================
>   <snip>But, of course, there are the late 30s'
> Brunswicks with their mushy sound. It made Eddie De Lange and Fletcher
> Henderson sound the same!
>
> It puzzles me how a record company could adopt such poor recording
> equipment. Did reviewers of the day comment on the poor sound? If a
> recording company is considering updating their equipment, either by
> their own development or buying in, wouldn't they make some test
> recordings first?
>
>      Julian Vein
>
> Somewhere I read that ARC began recording onto what amounted to lacquer
discs, that they called 'instant-o-tiles', around the end of 1936.  As this
was a new and as yet unproven technology then, I expect they would have
continued recording onto waxes simultaneously.  This perhaps explains the
existence of the two copies of Artie Shaw's 'Streamline' (recorded 23
December 1936) David Lennick has or had, which had the same matrix and take
number, but appeared to had been cut on two different lathes, AND sounded as
though they were taken from different mike feeds (so one of them was not
pressed from a matrix made by dubbing the matrix of the other copy).

My copy of 'Streamline' (Parlophone R 2984) looks like it was done on the
older equipment, but it has some aspects of the later styles of cutting.  It
has a rather shaky lead-in spiral as one with the grooves of the recording;
the lead-out spiral is the same coarse-pitch deep groove like most Columbias
have from c.1927 onward, with the eccentric ending introduced around early
1934.  However, there are two eccentric grooves close to one another, the
second groove just inside the diameter of the first one (the first the same
as that of the earlier single-groove eccentric ending).  The pitch of the
grooves in the recording are somewhat closer together than what I typically
see on Columbia (disregarding 5 minute, 'longer playing' 78s) and most any
contemporary ARC product I've seen.

The other side of R 2984 is Shaw's 'Sweet Lorraine', mx B.20453-2, from the
same session (and the next matrix in the ARC series), and it also has the
same general appearance of 'Streamline'.  Sound quality for both much the
same as for early to mid-1930s Columbia fare, betraying somewhat distant
miking in a smallish, deadened room.

It seems to me like all the poor-sounding ARC sides of the late 1930s (with
that ringy, wheezy, screechy top end) are those made on the lathe that gave
records that sharp, fast lead-in spiral, and spaced the first and last few
grooves on each side almost twice as far apart as the rest of the grooves,
with the added lead-out spiral having both the closer pitch and the
'machine-y' eccentric stopping grooves as also seen on red label era
Columbias through c. 1951 or 1952.  (Was there possibly some bug they hadn't
quite worked out with cutting to lacquers as opposed to cutting to waxes?)

But in 1939, after CBS completed their extensive equipment upgrades, the
sound greatly improved, and the groove pitches settled down....