Studies in variations of electric recording quality.
Posted: Mon Apr 06, 2009 3:52 am
The recent thread concerning a 1926 National Music Lovers release of some Banner sides led me to post some further examples of early electric recordings not made with the Western Electric technology. As WE would not sell or lease their equipment to any labels other than Victor and Columbia, the other record companies had to get by with whatever they could find or devise. Apparently that came from RCA, at least in Brunswick's case, and possibly in that of Plaza Music, aka Banner Records.
From what I understand, the first electric Banner sides were recorded in the autumn of 1925. The quality wasn't good at first, often thin and/or highly distorted. I wonder if January into February 1927 wasn't when they'd managed to 'de-bug' their equipment, based on Banner 1944's sides.
'Somebody else', by the 'Imperial Dance Orchestra' (here a pseudonym for Adrian Shubert and his orchestra) was assigned matrix number 7071-3 when recorded on 25 January 1927. It has the thin, rather distorted quality described above. Recording turntable wow was also rather pronounced as well. Vocalist is Irving Kaufman.
'A lane in Spain' by Fred Rich's Dance Orchestra was assigned matrix number 7095-2 when recorded on 4 February 1927, and tonally is much better balanced than the other side. (Vocal is by The Joy Boys.) There is still some turntable wow, but less than that of the other side.
I transferred both of these sides at 76.5 rpm, with the same EQ and filtering, so you can hear the great differences in quality between them.
Brunswick's problematic and much-ballyhooed 'Light Ray' recording tended to miss more than it hit. Some days they got rather good recordings with their equipment, only to get quite poor ones the next.
Here's an example of each, recorded on consecutive days, by the Ray Miller orchestra, released on Brunswick 3046.
'Dreaming of a castle in the air' is an easygoing fox trot of the mid-1920s, recorded here on 18 January 1926 (matrix E.17515). The orchestra is well-balanced, although the peaks are gently constricted, as if a smooth limiter were in use. The frequency range is nicely wide for the time as well (note the brightness of the cymbal crashes).
'Sweet nothings' was recorded the next day, 19 January, and is quite inferior to 'Castle'. (Matrix number is E.17550.) The orchestra was set up the same way in the studio, as evidenced by the balances of the sections, amount of 'room tone' having to do with distance from the mike, etc., but strangely, the overall sound is much more closed-in, and quite distorted at least 80% of the time. I wonder what was adjusted, and how, to have caused this result on the recording. It's a shame, as otherwise the record is quite good.
Both of those sides were transferred at 78.26 rpm, and with the same level, EQ and filter settings, one to the other.
From what I understand, the first electric Banner sides were recorded in the autumn of 1925. The quality wasn't good at first, often thin and/or highly distorted. I wonder if January into February 1927 wasn't when they'd managed to 'de-bug' their equipment, based on Banner 1944's sides.
'Somebody else', by the 'Imperial Dance Orchestra' (here a pseudonym for Adrian Shubert and his orchestra) was assigned matrix number 7071-3 when recorded on 25 January 1927. It has the thin, rather distorted quality described above. Recording turntable wow was also rather pronounced as well. Vocalist is Irving Kaufman.
'A lane in Spain' by Fred Rich's Dance Orchestra was assigned matrix number 7095-2 when recorded on 4 February 1927, and tonally is much better balanced than the other side. (Vocal is by The Joy Boys.) There is still some turntable wow, but less than that of the other side.
I transferred both of these sides at 76.5 rpm, with the same EQ and filtering, so you can hear the great differences in quality between them.
Brunswick's problematic and much-ballyhooed 'Light Ray' recording tended to miss more than it hit. Some days they got rather good recordings with their equipment, only to get quite poor ones the next.
Here's an example of each, recorded on consecutive days, by the Ray Miller orchestra, released on Brunswick 3046.
'Dreaming of a castle in the air' is an easygoing fox trot of the mid-1920s, recorded here on 18 January 1926 (matrix E.17515). The orchestra is well-balanced, although the peaks are gently constricted, as if a smooth limiter were in use. The frequency range is nicely wide for the time as well (note the brightness of the cymbal crashes).
'Sweet nothings' was recorded the next day, 19 January, and is quite inferior to 'Castle'. (Matrix number is E.17550.) The orchestra was set up the same way in the studio, as evidenced by the balances of the sections, amount of 'room tone' having to do with distance from the mike, etc., but strangely, the overall sound is much more closed-in, and quite distorted at least 80% of the time. I wonder what was adjusted, and how, to have caused this result on the recording. It's a shame, as otherwise the record is quite good.
Both of those sides were transferred at 78.26 rpm, and with the same level, EQ and filter settings, one to the other.
