On 8/30/2014 21:08, Tom Fine wrote:
> I've heard the Ellington "accidental stereo" recordings that are
> available for download online (lossy formats, which may effect some of
> my conclusions). They sound like unimpressive "stereo", accidental
> indeed. They do reveal some sonics of the room not available in the mono
> audio, but don't clarify or otherwise serve the music. Assuming
> Michael's story below is true, they sound exactly like that -- two
> recordings with two intents, not two purposeful recordings intended to
> provide more musical detail or clarify than either alone. I suppose
> there's a "neato" factor, but that's about it. No big deal.

Hi Tom,

There isn't any assuming, per se; the 24 CD "Duke Ellington Complete 
Victor Recordings" box that came out in 1998 contains both of them and 
lists their individual matrix numbers, the alphabetical portion of which 
reflect codes for the equipment used. They were recorded on February 3 
and February 9, 1932.

Looking at the matrix numbers now, as opposed to 1998 when several of us 
discussed this set at length on 78-L, it looks more like Victor was 
testing two different RCA systems. In the sessionography, the BVE 
(10")/CVE (12") matrix numbers are replaced in mid to late 1931 with 
BRC/CRC, which seems to indicate RCA equipment. The two "unintentional 
stereo" recordings each carry dual LBRC/LBSHQ matrix numbers, but serial 
take numbers: LBRC-71811-1/LBSHQ-41812-2 for Mood Indigo/Hot and 
Bothered/Creole Love Call, and LBRC-71836-2/LBSHQ-71837-1 for East St. 
Louis Toodle-) [sic]/Lots O' Fingers/Black and Tan Fantasy.

An additional standard take, LBSHQ-17812-3, is present for the first 
record. After the single take of the second record was recorded, two 
additional standard 78RPM records were made during that session using 
BRC matrix numbers, "Dinah" and "Bugle Call Rag".

L was long playing, B was ten inch; RC presumably stands for RCA, S 
hasn't been figured out yet to my knowledge. HQ most likely stood for 
high quality, and since I've only seen it on Victor Electrical 
Transcription discs, it's my presumption that this was an attempt at 
refining the recording chain for use with the long-playing records, to 
approximate the audio quality of the 78RPM record.

Not long after this session, the matrix numbers switch to BS/CS, which 
was the core matrix prefix for Victor and Bluebird recordings pretty 
much through 1946.

Michael Shoshani