NPR - The Roots of Audio Recordings Turn at 78 RPM

[novkev24]

Listen Here: http://www.npr.org/templates/story/stor ... Id=6645723

[Henry]

Ah yes, history lite. In fact, the 10" 78 can hold more than three minutes' worth of content, as proven by the Paul Whiteman "Sweet Sue" (Columbia 35667) at 4'22". And many more examples of popular song, composed before recording was even thought of, time out at 3-4 minutes (examples: most songs by Stephen Foster, George F. Root, Paul Dresser, et al.). The notion that somehow the three-minute length of popular songs was dictated by the so-called limitations of the recording process is half-baked, and stems from ignorance of music history. My own belief is that song length has more to do with the "average listener's" attention span, but I can't offer any objective evidence to support my own hypothesis.

As for the classical guys cutting their own pieces to fit, I suspect that the 20-minute version of Elgar's Violin Concerto says all it has to say musically in those 20 minutes; 50 minutes is an awfully long concerto! And in regard to the Stravinsky piece, the professor is stretching it a bit when he says "each movement lasted exactly the length of a 10-inch 78-RPM record," when we know that there was nothing exact about that length, and in any case it would have been perfectly feasible to record on 12" discs; here again I suspect that musical considerations, rather than technical limitations, guided Stravinsky's decision.

Just sign me "Grumpy Old Music Professor."

Henry

[novkev24]

Thanks for your post Henry. It's always nice to get another perspective on the topic.

-Kevin

[phonojim]

Columbia 10" #35667 is a truncated version of the original 12" #50103-D (per Rust)and can be visually determined to be a dub. It was edited down so it could be included in album C-29, a "Hot Jazz Classics" set of Bix Beiderbecke reissues. It plays for 3:35 at 78 RPM (strobed), a difference of :47. However, as Hit Of The Week proved, it was possible to record 5 minutes on a 10" 78 and do it well.
I think you are correct about the reasons for the length of popular records. Even piano rolls of popular songs seem to run around 3 min. playing time even though they could easily have been made to play much longer.

Jim

[Henry]

Jim,

"Sweet Sue" and its several versions have been discussed "at length" (ha!) on the previous board. AFAIK, the version I have on Col. 35667 (from album C-29) is complete, including the symphonic-style introduction, and does time out at 4'22", not 3'35", with my turntable running at strobed 78 rpm. I've never heard the version on #50103-D, but I'd like to; if anyone owns it and can post a dub on the board for me to compare with my 35667, I'd appreciate it, and I could post a complete formal analysis of both versions, with timings.

Interestingly, here's what George Avakian's liner notes say about the 10" #35667 version of "Sweet Sue" on album C-29: " 'Sweet Sue' contains what Bix followers consider his best solo with the Whiteman band: a beautiful half-chorus played with a derby hanging over the bell of his cornet. Originally a 12-inch record, 'Sweet Sue' had to be re-recorded on 10-inch wax for this album. This apparently impossible feat was accomplished by Columbia engineers by merely dropping the florid introduction." But on my #35667, there IS most definitely an introduction, and of some length at that. Whether this one is the same as Avakian's "florid introduction" could easily be determined by an A-B comparison.

Henry

[Swing Band Heaven]

Here is a sound file of the 12" version of "Sweet Sue" by Whiteman

Sweet Sue - Paul Whiteman

I don't have a note of the speed I transcribed this at but the length on my file is 4.18.

BTW - I wondered for years what the title of this song was after hearing it played in the film "Some like it hot". It was only after I found out years later that I got the "joke" in the film regarding the name of the band leader!

I don't have either of the other versions referred to but if anyone else has them it would be interesting to compare.

[phonojim]

Henry, I just found 50103-D on Youtube http://www.youtube.com/watch?v=VTd3jretEOE.

Jim

[Swing Band Heaven]

.....However, as Hit Of The Week proved, it was possible to record 5 minutes on a 10" 78 and do it well.....

Jim

I've just checked by Box files and see the longest recoding time I have on a single 10" side is 5.21. This is from a disk which had two almost full length songs on each side - but the grooves are very very fine and its difficult to find these type of disks in good condition as early soundboxes were very hard on them. But the sound is very good indeed and no different to a standard 10"

[Edisone]

This is from Don Rayno's book about Whiteman:

" Take 4 was subsequently released as Columbia 35667 as part of album C-29. Most issues of 35667, however, have the introduction edited out. The runoff groove of the edited issue has "2A" stamped at the top..."

The truncated version was released first, and later replaced with a finer-grooved dub of the complete piece. The origunal 12" release was Take 1

"Sweet Sue" and its several versions have been discussed "at length" (ha!) on the previous board. AFAIK, the version I have on Col. 35667 (from album C-29) is complete, including the symphonic-style introduction, and does time out at 4'22", not 3'35", with my turntable running at strobed 78 rpm. I've never heard the version on #50103-D, but I'd like to; if anyone owns it and can post a dub on the board for me to compare with my 35667, I'd appreciate it, and I could post a complete formal analysis of both versions, with timings.

Henry

[phonojim]

Edisone, Thanks for the research. I found that the copy in my album is the complete one while my loose copy is the edited one. I just compared the two and found that the truncated one is recorded at a "normal" level with the customary strong bass. The complete one is recorded at a lower level with reduced bass, all to allow the finer grooves with closer spacing. The things one can learn here...

Jim