The Talking Machine Forum — For All Antique Phonographs & Recordings

I have a bunch of 12 inch 78s recorded between 1902 and 1920, how long can these play? They say 4 minutes though I have seen some playing as long as 6 minutes. how long can they really play?

Why not play the records and find out for yourself?

It was not a standard... I believe they adapted the groove separation to the kind of music they were recording, and softer music was recorded with more grooves to the inch, and so the recording might be longer. There is great variability among different recordings, and of course between different brands. I've not chronometred it, but some British 12" acoustic Vocalions seem to be quite long... I remember especially violin records by Albert Sammons, and those of Felix Salmond, Warwick-Evans and others of soft violin or cello with piano accompaniment. Some German Polydor/Grammophon electrical recordings of late 20s have also long 12" sides...

Just working on this subject for various reasons.
Currently measured some Mid 1910s records and it looks as if Columbia were using a fixed pitch of 100 TPI which with the full size label gives a full 5 min HMV of the same period is courser at around 95 TPI which reduces this to 4min 45 seconds.
the Gramophone co and victor lathes of the time did have choices of feed pitch but for anything other than non- dramatic speech around 100 TPI would be optimum a setting on the Victor/ Gramophone Co machine of 120 TPI was available but the records would have to be of lower sound volume and possibly more difficult to process into good pressings it doesn't mean they don't exist and could run to 6 minutes.
In the case of courser pitches these were often used to fill out a short piece to make it look good value.

If for example all Columbia records of short duration on 12 inch format ( say from the 1910s ) have a lot of blank space it might suggest they did not have a selection of lead screw settings.
if any one has a 12" record that shows as a full record but runs for say 3 minutes then the feed must have been courser.

On the development of electrical recording the process of compressing the signal did allow for very slightly finer grooves without the groove crashing into each other as seen in the 4 in 1 of the late 20s with 2 tunes on each side getting about 4 min plus on a 10 inch side with a normal label size.

recordmaker wrote: Fri Jan 19, 2024 1:36 pm Just working on this subject for various reasons.
Currently measured some Mid 1910s records and it looks as if Columbia were using a fixed pitch of 100 TPI which with the full size label gives a full 5 min HMV of the same period is courser at around 95 TPI which reduces this to 4min 45 seconds.
the Gramophone co and victor lathes of the time did have choices of feed pitch but for anything other than non- dramatic speech around 100 TPI would be optimum a setting on the Victor/ Gramophone Co machine of 120 TPI was available but the records would have to be of lower sound volume and possibly more difficult to process into good pressings it doesn't mean they don't exist and could run to 6 minutes.
In the case of courser pitches these were often used to fill out a short piece to make it look good value.

If for example all Columbia records of short duration on 12 inch format ( say from the 1910s ) have a lot of blank space it might suggest they did not have a selection of lead screw settings.
if any one has a 12" record that shows as a full record but runs for say 3 minutes then the feed must have been courser.

On the development of electrical recording the process of compressing the signal did allow for very slightly finer grooves without the groove crashing into each other as seen in the 4 in 1 of the late 20s with 2 tunes on each side getting about 4 min plus on a 10 inch side with a normal label size.

The longest 12 inch Columbia I have ever found was only about three minutes and thirty seconds on a disc from 1915. The longest 12 inch Victor I have was exactly four minutes long. Next time I go looking for 78s, I will try to see if I can find some 6 minute Victors.

recordmaker wrote: Fri Jan 19, 2024 1:36 pm On the development of electrical recording the process of compressing the signal did allow for very slightly finer grooves without the groove crashing into each other as seen in the 4 in 1 of the late 20s with 2 tunes on each side getting about 4 min plus on a 10 inch side with a normal label size.

I will disagree with this assertion. Being able to ride the gain was available to control loudness hence groove modulation level but very detrimental to the dynamics of the music. Electrical recordings tend to be louder hence there will be greater lateral modulation requiring a coarser groove.

But the largest problem was capturing and reproducing the bass in the new process electrical recordings. In the days of acoustical playback (Orthophonic Victrolas) the playback characteristic was "constant velocity" which meant that the rate of change or slope of the lateral groove swing had to be constant from low to high frequency. Hence the amplitude of the bass lateral modulation had to be increased whilst the treble was reduced. This meant electrically recorded groove pitch, especially for the first few years after 1925 had to be coarser to accommodate the wider bass modulation and and at the same time to keep lower level treble amplitude above surface noise. It was a troublesome compromise.

The constant velocity characteristic continued well into the 1930s and it was only from around 1940 with the preponderance of electronic playback (which facilitated passive electrical component equalization filtering) the characteristic began to approach a more practical constant amplitude characteristic.

.... how long can they really play?

Well, it all depends upon the listener and what the person is listening to.

For example, personally, my Madam Buterfly records seem to last for only a couple of minutes - as opposed to some Caruso records which go on and on and on for-ever ! *


*There is some attempt at dry humour in there some-where

They play - as long as they play... What difference does it make, unless you are manufacturing new records to fit a given size blank?

"On the development of electrical recording the process of compressing the signal did allow for very slightly finer grooves without the groove crashing into each other as seen in the 4 in 1 of the late 20s with 2 tunes on each side getting about 4 min plus on a 10 inch side with a normal label size." Recordmaker

"I will disagree with this assertion. Being able to ride the gain was available to control loudness hence groove modulation level but very detrimental to the dynamics of the music. Electrical recordings tend to be louder hence there will be greater lateral modulation requiring a coarser groove." Gov. Flyball

This is not an assertion it is reference to a particular brand of record (The 4 in 1 Record ) that attempted to fit more on one side, the results, as you correctly suggest , have limited dynamic range , poor bass and are fast to wear out. I also assume poor sales, I only have 1 example in my collection. I made no comment about the quality just the increased duration and use of compression.
This was not common or desirable practice for many of the reasons you state in your additional comments that cover the recording characteristic compromises in that post acoustic era nicely.

It is of course possible to time a record and measure the track length to calculate the groove pitch and support your suggestion that on the introduction of electrical recording groove pitch got courser for some if not all music.

100TPI uses 0.78 inch per minute at 78RPM for example

With regard to constant velocity and constant amplitude for electric recordings.

The 1925 report by Maxfield and Harrison of Bell Labs established the idea of turnoff, which is the frequency were constant velocity changes to constant amplitude as the frequency decreases.
Most record Companys used a turnover of 250 Hz, then as time went on, during the late 20's and into 40's, the values would change until the LP RIAA curve established fixed values.

Marc