The Talking Machine Forum — For All Antique Phonographs & Recordings

A while ago I bought some RCA home recorded discs from eBay. They were supposed to have radio recordings on them and it turns out only one was used but that is what it says on the label. When I listened to them (on a 70s record player), besides being very noisy they sounded too fast and they didn't sound like the versions of the songs written on the label that are on the internet. I know that the record player is at the right speed. I would expect the noise because these need a 4-6 mil stylus to be played correctly but the speed and difference in the songs is weird. I don't know how they are at a odd speed because the phonographs these could have been recorded on only had 33 ⅓ (from program transcription records) and 78 rpm. Is their a way to find out what speed these were recorded at? I know with other formats like tape professionals might use a bias frequency or a hum to correct speed but I doubt that could work in this situation. And is their any way to further improve the sound quality? Even with noise reduction in audacity there is some parts with some much noise the music is inaudible. They are interesting because they are such early recordings of radio (December 16 1933). They are RCA Victor pre recorded discs.

Recording machine could have just been malfunctioning in some way. 1933 is early, these should be worth trying to wrangle some sound out of. Audacity is a handy tool, and it's high and low pass filters are fine, but it's other noise reduction capabilities are limited. A surfeit of click and pop noise is better handled in a program like Click Repair.

I don't know much about these blanks but there is radio forum on line that mentions them.
In that there is a comment that the recordings seem to run fast ( about 4% )
The system seems to have been a pre pressed groove in medium soft plastic (soft at the time, as it probably hardens in storage) and reverse driving the radiogram pickup as a cutter or embosser of the groove.
The resulting modulation is probably in the top part of the groove so a special larger than average stylus would be needed to get the best out of the playback.

The speed alteration is almost certainly due to drag on the turntable while recording.
The suggested 4 % would be more than noticeable in the pieces of classical music that the label suggests.
obviously you need to compare the pitch of the much rather than the tempo which can vary according to the orchestra conductors interpretation

The recorder could well have been running off speed, either because of mechanical issues or because of drag from the cutting stylus. The Moszkowski may be a bit harder to identify, but Bizet's L'Arlesienne music is a hardy perennial. Find a recording of it on YouTube (here's one, for example https://www.youtube.com/watch?v=hBlNa9_RCNw ) and compare pitch. Adjust your turntable as necessary to match. I would guess the Moszkowski number would run at the same speed, assuming, as seems overwhelmingly likely, it was cut on the same machine. [edit] In case you aren't a classical buff, note that the L'Arlesienne music is a suite of pieces; your record may not start with the same one as the YouTube video does. You may need to skim through the latter to find what you have on the former. Apologies if I'm being Captain Obvious here!

One issue that you should hope you don't have: early electric English Columbia records often have pitch creep, getting faster as the record progresses, because the lathe motors weren't strong enough to hold a constant speed against the changing drag of the cutting stylus. If you have that problem, you'll need to adjust the speed in fine increments across the entire surface of the record to start and end at the right pitch; you'll probably be only approximately right at most points in between.

If your turntable doesn't have pitch adjustment, or enough pitch adjustment, you'll need to record to your computer at the wrong speed and do the job in software. Can't help you there; I do have a pitch adjustment program, but I've never applied it for that purpose. Note that up until 1930 or so, the one speed at which "78s" never seem to turn is 78 RPM.

Are you playing the records in mono? Stereo playback will dramatically increase noise. Ideally, you should have a collection of styli in various sizes to find a good match for the groove ("good match" being defined as a size that rides above or below the level where playback with the punishing gear of your record's youth has damaged it). I reviewed one set of such here: https://www.tnt-audio.com/vintage/rek_o_kut_e.html That would be overkill if you have only these few records to play, but it would be a wise investment if you play lots of 78s on a modern turntable.

The conductor of your record would almost certainly have been Frederick Stock. Much under-rated today. Good score there!

Audacity speed correction tool works well. You want to use the speed correction in Audacity, not the pitch correction.