Elliptical stylus for 1916-1921 Columbia?

[dlovrien]

I have a collection of Columbia saxophone records from 1916-1921. Some are on the rare side and my only copy is pretty worn, but I can still hear a decent frequency range inside the surface noise. With a lot of computer processing I've gotten a few nice transfers, but would like to get a better initial play from my Dual 1229. You can hear my current results here:

http://www.dws.org/saxtette

I'm currently using a Grado 78C and am wondering if an elliptical stylus would make big difference. On one site I heard some sound samples with 3.0, 3.5 and 4mil ellipticals used on a worn disc and the surface noise was increasingly reduced by the larger stylii. Feedback appreciated.

[bart1927]

I have a collection of Columbia saxophone records from 1916-1921. Some are on the rare side and my only copy is pretty worn, but I can still hear a decent frequency range inside the surface noise. With a lot of computer processing I've gotten a few nice transfers, but would like to get a better initial play from my Dual 1229. You can hear my current results here:

http://www.dws.org/saxtette

I'm currently using a Grado 78C and am wondering if an elliptical stylus would make big difference. On one site I heard some sound samples with 3.0, 3.5 and 4mil ellipticals used on a worn disc and the surface noise was increasingly reduced by the larger stylii. Feedback appreciated.

I assume your grado stylus is 2,7 or perhaps 3 mil. A bigger stylus would certainly help to reduce surface noise, those standard 78 styli are more suited for playing 78's from the 1950's.

A general rule is that for playing worn 78's a conical stylus is better than an elliptical one. For your Columbia's I think a good stylus would be the 3.5 tc (truncated/conical). Those styli can be ordered from Kurt Nauck (http://www.78rpm.com) but they are not cheap, about $ 150 !

[syncopeter]

Saxophone records are quite hard to restore properly. Try to invert the file, i.e. putting it back to front, when you apply declick and denoise filters. And afterwards put it back again. That may make a bigger difference than investing in an expensive stylus. Programs often have problems separating impulse noise from music. Inverting a file helps I found. It will cost you no more than a little time. I agree with Bart1927: a worn record often sounds better with a fatter conical stylus, but there is no fixed rule in this. Sometimes increasing playing weight can give a huge reduction in surface noise too, but sometimes it doesn't.

[Wolfe]

Since your working on computer, doing a half speed transfer (or any speed like 40 rpm or so) and then restoring the original speed with an audio editor can help matters. A lot of the high frequency noise, upon restoring the speed, will go right up outside the frequency range cutoff at 22 kHz, since it's been doubled as well. Of course, one will have to fix the EQ a bit, but that's not really a problem with limited frequency range acoustical records. Of course, one can always just filter the noise, but half speed seems to work even better, IMO, if one wants to go to the trouble, especially with things like those noisy laminated Columbias.

[gregbogantz]

The purpose of the elliptical stylus shape is to have a smaller scanning radius while still having a large enough bearing radius so that the stylus does not sit too far down in the bottom of the groove. These styli carry a double size specification. For example, they may be rated 0.7 mil by 3 mil - the smaller number is the scanning radius and the larger is the bearing radius. Shure called these styli "bi-radial" which is more appropriate to their shape and use. The smaller scanning radius allows more accurate reproduction of the smaller recorded wavelengths associated with the higher audio frequencies. This results in a brighter sound with less distortion. This true for 78rpm as well as for microgroove 33s and 45s. The bearing radius that works best for a worn record will depend on where the wear pattern is on that particular record. Adjusting the bearing radius allows the stylus to hit the groove wall either above or below the wear pattern, thereby scanning the unworn parts of the groove walls. Thus, you need to have a variety of stylus sizes - some will work better than others with certain worn records. So, simply using an elliptical stylus will not, per se, result in less record noise. You still need a selection of elliptical styli with different bearing radii, just as you need a selection of conical styli of various radii. That said, elliptical styli WILL result in cleaner high frequency audio when you play good records or the unworn areas of worn records. This is particularly noticeable at the smallest diameters of the record where the recorded wavelengths are shortest. If you like to listen to a lot of Harry James or other loud brass instrument electrical recordings, an elliptical stylus will sound cleaner than a conical one.

[VintageTechnologies]

On the general subject of audio processing, I would caution anyone not to over-do it, because sometimes the resulting distortion and low-frequency sound artifacts are worse than the record hiss. As an example, I download a lot of acoustic recordings from www.archive.org. Someone named "grim ripper" has gone to a lot of trouble uploading tons of music (which I certainly appreciate), but much of his music has been overprocessed to the point of being unbearable to hear. Most people (myself included) don't know how to process sound like the professionals do and it becomes very evident. In such cases, I would rather hear the hiss than the distortion.

[Wolfe]

On the general subject of audio processing, I would caution anyone not to over-do it, because sometimes the resulting distortion and low-frequency sound artifacts are worse than the record hiss. As an example, I download a lot of acoustic recordings from http://www.archive.org. Someone named "grim ripper" has gone to a lot of trouble uploading tons of music (which I certainly appreciate), but much of his music has been overprocessed to the point of being unbearable to hear. Most people (myself included) don't know how to process sound like the professionals do and it becomes very evident. In such cases, I would rather hear the hiss than the distortion.

That might the one I downloaded a track from and the level had bee set so high, that on a visual waveform all the louder peaks were completely smashed / flattened out (sometimes called brickwalling, which is also done with modern CD masterings.) Inept, or just a totally inappropriate technique to use with 78's, it sounds terrible.