The Berresford Blog | Mark Berresford RARE RECORDS | Mark Berresford

The Berresford Blog

More on Styli for playing 78s

I get so many emails from collectors about what stylus to use for playing 78s that I thought I’d address it on my blog page, even though I have written one with a link to this information before. I suppose having the information on one page makes more sense, so here it is...


By Ron Geesin and Mark Berresford

westrex cutting electrical

One of the most commonly-asked questions we encounter from new collectors (and some not so new!) is "What size stylus do I need to play my lovely jazz/blues 78s?" Answer: There is no such thing as a standard ‘play all’ 78 stylus! In the almost 120 - year existence of the flat record each and every company varied in the size of their grooves, not just periodically but on a day to day basis – the recording engineer was a cross between an alchemist, scientist and hands-on engineer and each guarded his own secrets and innovations jealously, not only from other companies but from in-house colleagues as well! Paul Whiteman in his autobiography recounts how the Sooy brothers, Charles and Raymond, Victor’s top recording engineers, guarded their personal cutters - to which they were forever making their own improvements - from one another, to the extent of taking them home with them in the evening!

 We want to share with you our observations collected from many years of gentle but relentless groove-digging for the very best-sounding music. Even the most proficient computer-aided retrieval and restoration techniques need the best input from the groove - separate signals from each side of the groove preferred so the final optimum monophonic signal can be projected. A groove can be fitted, or misfitted, with a stylus. If you are using audio criteria as part of your grading assessment, a groove misfit could lead to a disparity of some two notches of grade either way. But our main reason is to dispel the mystique and to encourage you to enjoy the very best sound. The subject really needs a whole chapter in a book. Here we will keep it tight and packed full of information.

 There are basically two types of monophonic groove configuration: ‘lateral’ and ‘vertical’ (hill and dale). Since ‘vertical’ uses a constant width, variable depth method, once the Edison, Pathé or ‘other’ is matched with its particular ball-type stylus (again, there are varying factors to consider – see below), combined with an electrical means of getting only the vertical signal, you have more or less covered the stylus problem, so we will concentrate on ‘lateral’ cut recordings.

 In the beginning, there was the cutter. This was a tiny bar of sapphire sharpened to a V with an angle varying between 60º and 90º, Europe tending towards the former and America the latter. The depth of cut, maintained by counter-weights and sometimes supplemented by a ball travelling on the wax just in front of the cutter, was supposed to be about .0025" (2.5 thou.) but often varied due to local conditions on the day and engineer’s judgement. In our early jazz field, we have to concernourselves also with the smaller, or specialised, or less technologically-aware companies where blunt or misshapen cutters and odd groove-depths occur. Every day may have produced a different setting!

Drawings of the cross-sections of 78rpm record grooves that you may have seen in technical books are misleading. The theoretical V-shape with rounded bottom mostly turns out in practice to be the bottom curve of a U with the legs cut off. This contour is due to a combination of the plastic nature of the materials and electroplating processes. You will see in the following illustration from the rare but informative book "The Gramophone Record" by H. Courtney Bryson, 1935, that there is actually not much room for error…

groove sections (1)

However hard a ‘shellac’ 78rpm record may look or sound by tapping, it is still a crumbly plastic full of clay, carbon black, aluminium flake, gypsum, powdered limestone, fibres, Fuller’s Earth and slate dust – yes, Bryson says, "records have been manufactured containing as much as 70 per cent. of this material." – all bonded with shellac and other resins. When one scrutinises a field of grooves through the microscope, one can observe just how easy it is to push through a groove wall with a specially-mounted stylus – like pushing a finger through some classic English Cheddar cheese. 

 In the days when our favourite records were being manufactured, and there are those of us still doing this (!), the steel needle aided by a heavy pickup actually wore to fit the groove: that’s one excuse for the abrasives listed above. This wear could be tested/seen by rotating a used-once steel needle tip under a single-point light source and observing the glinting: the ‘flats’ worn by the playing process. The ‘missing’ metal meanwhile had been deposited into the groove walls and can frequently be seen as metallic ‘glints’ on heavily worn records. The average consumer (an apt word in this case) would then rotate the needle in its holder to get another play out of it, chiselling its way along the groove – consumed indeed. Hence so many old 78s with worn starts. The other favourite of the past, especially among collectors of classical or operatic records was the fibre needle. This did not wear the groove, only filled it up with powdered fibre, blocking good clean sound (and is nigh-impossible to remove) and providing the ideal breeding ground for mould, which loves shellac too. 

 More recently, we have learned to caress our precious groove with a polished sapphire or diamond stylus, mounted in a very sensitive light-to-medium-weight cartridge and pickup. Sapphire-tipped styli are much cheaper than diamond-tipped, but they really are a false economy; their wear life is a fraction of that of a diamond and the initial savings will be offset by the frequency of replacement, so we do not recommend their use. You can see by now that fitting the groove with something that is not going to wear to fit it is quite a science, if not an art.

 Although there are only a few specialists in the world supplying dimensionally tailored styli, there are enough for our needs. The terms you might encounter are, principally:

Conical: cone shaped, round section (viewed from above)

Elliptical: cone-shaped (viewed from along the groove), elliptical section (viewed from above)

Full Radius: the tip is exactly spherical

Truncated: the tip is slightly flattened, better fitting the cut-down U-shape groove

Tip Radius: for 78s, anything from just under .002" (2 thou.) to .006" (6 thou.)

 For shellac 78s, Conical is probably to be preferred to Elliptical. The round section rides the grit, whereas elliptical will engage the grit and other imperfections because of its slightly reduced surface contact area. In certain critical circumstances, particularly when playing very clean copies, an elliptical stylus will give a little better definition of the higher frequencies, but they can be quite unforgiving on scratched or worn records. Elliptical is sometimes preferred for best vinyl and microgroove reproduction.

 Again, for our purposes, Truncated is usually preferred to Full Radius, for the reason given before.

 While the professional sound restorer may have a full range of stylus assemblies, maybe up to 20 and all to fit in one cartridge, the average record collector might do quite well with three or four. If you are happy with approximation and/or on a tight budget, then a .0028" (2.8 thou.) Conical Truncated would be the one to get, but bear in mind the reference table below. The post-1945 groove tended to settle at 2.5 thou. radius, though again collectors of post war blues recorded for shoestring record companies will realise that there is no such thing as a ‘standard.’ So, a small kit, capable of dealing with the majority of records one is likely to find in a collection with a 1910-1950 timeline, might include 2.5, 2.8, 3.5 and possibly 4 thou. radii.

 Now we move to the main business of fitting the groove. ‘Bottoming’ occurs when the stylus is too small for the groove and literally contacts the bottom of the groove (remember the signal on a laterally-cut recording is on the groove walls) and is easily heard as an irregular swishing: the effect being rather like a toboggan down the Olympic track. It is probably best to start with a 2.8 thou. and work up or down in stylus size from that as necessary. The optimum point in the groove is just below the shoulder, because there you get maximum deflection, side to side movement of the stylus tip, giving you the best signal-to-noise ratio. But, that is also along the line that a previous steel needle may have chiselled out, so you have to come down in size a bit to track the groove area that has not been damaged. Surface scratches may also force you that way. Occasionally, running an undersized stylus and a tracking brush (Dust Bug or similar) can help keep the groove as clean as possible. This of course brings in the thorny subject of to wash or not to wash records and how to do this without causing irreversible damage to the record. This subject is worth an article on its own (let us know if you want one - MB) and is not our main concern here, but bear in mind that a clean and detritus-free groove is not only going to sound better but will also reduce the wear and damage on those expensive industrial diamonds!

 In the end, the art of compromise has to be practised. Sometimes, the groove width and/or depth may change from start to finish of one recording, causing the restorer to change radii and make an edit later for the best result. Tracking weight also has to be considered. Too much will force the stylus arm up into the cartridge body, or, more accurately, force the cartridge body onto the record surface: too little will result in inaccurate tracking, possible skipping and swishing.

 All this talk of racks of multiple sized diamond styli may seem rather daunting and an expensive luxury when there are records to be bought, but the financial outlay will be far outweighed by the pleasure derived from hearing the enormous improvement in the sound quality of your records. Play away!

 The following reference guide gives you an even better indication of where to start fitting the groove – as you have by now gathered there are no hard and fast rules but, based on years of playing a variety of makes and ages of 78s and covering a timeline of 1895-1950s, this table will give the reader a fairly good idea of where one should base their starting point. Anomalies abound, as one might expect, and some notable ones are mentioned where appropriate. We’d love to hear from you if you know of similar exceptions to the rule! 

 Whilst we do not claim that this list is exhaustive or 100% reliable in each and every case (see comments above about riding above or below groove wear/damage) it does give a good starting point for individual experimentation. Ultimately, your ears will be the best judge and it is always worth noting a disc’s preferred stylus size on the sleeve for future reference. As can be clearly seen, the vast majority of records fall into just three stylus sizes – 0.0024/5", 0.0028" and 0.0035". If one’s budget only runs to three styli, these are the ones to go for.

Further References

Bryson, H. Courtney, THE GRAMOPHONE RECORD, Ernest Benn Ltd, London, 1935.

Read, Oliver and Welch, Walter L., FROM TINFOIL TO STEREO, Howard W. Sams & Co. Inc., and The Bobbs- Merrill Company, Inc., Indianapolis & New York, 1959.

Thanks also to Russ Shor and the late Malcolm Shaw for their assistance.




Useful for some acoustic Richmond Gennett recordings: Some early aluminium and acetate transcription discs.


Extremely useful for the majority of acoustic Richmond Gennetts. Strangely, the Wolverine Orchestra’s Fidgety Feet/Jazz Me Blues like a whopping 0.0035"! Also a number of acoustic OKeh ‘field’ recordings (e.g. many of the 1923 King Oliver OKehs) like this size.


Very useful for acoustic, ‘Truetone’ and some electric OKehs, pre-1920 Victors, most post-war records. Many Paramount 12700 series. English Columbias 1925-26 mainly prefer this size. This is the most common ‘off the shelf’ stylus size you will encounter.


Post 1922 acoustic and electric Columbias, most electric OKehs, acoustic and early electric Paramounts, Gennett (acoustic and electric New York recordings and electric Richmond recordings), Brunswick/ARC (acoustic and electrical), US Decca, Victor 1921-25, most Plaza group acoustics, pre-1931 Parlophone, Black Swan, Cameo, Pathé/Perfect lateral recordings. A good starting size to work up or down from.


Useful for some Brunswick/Vocalion electrics, post-EMI merger English Columbia/HMV/Parlophones, Columbia acoustics to 1922, some early US Columbia electrics, also some later Columbia electrics where the master has been heavily polished or high numbered stampers have been used. Autograph, some Paramount 12800s (though these vary enormously – up to 0.004"!


Most Victor Electrics post 1926, some very early Victors (1901-06), Acoustic Vocalion, HMV/Zonophone 1925-31, most QRS, Plaza group electrics, Edison Diamond Discs, most Electrobeam Gennett post-6400 issues.


Early electric Victors (1925/6). One notable anomaly is the October 1927 Brunswick session by Johnny Dodds’ Black Bottom Stompers, where the cutter was particularly blunt and all the sides from the session need a 0.004" stylus.


eBay has dozens of sellers offering ’78 styli’ and, whilst most are perfectly respectable, it is a very specialised field, with only two companies (in fact only one who offers a comprehensive, customised, service) in the world who can be truly called ‘specialist’.


Expert Stylus Company, PO Box 3, Ashtead, Surrey KT21 2QD, UK. Tel: +44 1372 276604; Fax: +44 1372 276147. Expensive (and a near – monopoly), but they are the best in the business and the first choice for professionals and collectors the world over. Their years of experience and friendly (and free) advice will benefit both neophytes and old hands alike.]


KAB Electro - Acoustics, P.O. Box 2922, Plainfield, NJ 07062 Tel: (908) 754-1479 /Email: [email protected] Web site: A very helpful and knowledgeable specialist company, usually with a limited availability for given stylus sizes.



The First ‘Cajun’ Recording

Last December I was very fortunate enough to acquire what is probably the finest collection assembled of Jazz and Blues 78s recorded in New Orleans in the 'Golden Age' - the 1920s.

As well as immaculate copies of all the classic sides by Sam Morgan, Celestin’s Tuxedo Orchestra, Johnny Bayersdorffer etc., there was something I’d never seen, or even heard of before. It looks like a normal c. 1925 OKeh 78, but with ‘Roach Record No. 2-A’ and ‘Roach Record No. 2-B’ in place of the normal OKeh catalogue number. The A side has one James F. Roach, accompanied at the piano by Agnes Farrell Roach singing the Cajun song ‘Gué Gué Solingaie’, whilst on the reverse Agnes Farrell Roach plays the piano solo ‘Reflets Ians L’Eau’ [sic], which is best described as ‘Debussy-esque’. 

This recording of ‘Gué Gué Solingaie’ is thus the oldest known recording of a Cajun song by some three years. As for ‘Roach Record No. 2’ James F. Roach sings ‘Gué Gué Solingaie’, a known Cajun song (also called ‘Song of the Crocodile’) in a classically-trained manner - after all he was no Cajun, but a respected physician, originally from St. Louis. However, its importance is that is is the first!

An excellent, in-depth blog article on this song and more about James F. Roach can be found at

The matrices are 8889-a and 8911-a. 8911-a is the first traced matrix of the January 1925 OKeh field trip to New Orleans which were recorded in the four-storey  Junius Hart Piano House at 123 Carondelet Street. Matrices 8886-8888 have never been traced and at least two of them may be what went to make up ‘Roach Record 1’ - if such a disc was ever issued.

Until my copy surfaced, only one copy has ever been found of this disc, and no label photos have been made available, nor the music made available to collectors. However I will be selling this disc shortly (not before quality transfers are made), on eBay, so if you are interested, make sure you get onto my Mailing List!

IMG 0914

Getting the best sound out of your 78s

The question I’m probably most asked about playing 78s is “What size stylus do you use?” Well, the answer is actually not one, but several, of different radii and profiles. However the bulk of pre- and post-war 78s can be played with just two different-sized styli. A 0.0028” full conical stylus will play most acoustic and electric 78s, and for pre-1935 Victors and HMVs, a 0.0035” full conical will get the best out of those deep, wide boomy-bass grooves. Dependent on the breadth and age of your 78s, you may need more - a lot of early acoustic 78s like a small radius stylus - 0.0024” or even 0.002” on some records made on the legendary Gennett label at their Richomond, Indiana studios - think King Oliver’s Creole Jazz Band, the Wolverine Orchestra, Jelly Roll Morton’s piano solos, Whistler’s Jug Band and many more. A few years back I co-authored an article with Ron Geesin that appeared in Vintage Jazz Mart magazine, and its contents are still accurate and pertinent. The world’s leading suppliers are still Expert Stylus Company in Surrey, England - if you do contact them please mention my name!  Click HERE to see the full article.

'Sicily Jass - The World’s First Man In Jazz’

Phew! Two days of hard work with the Italian film crew making a new film about the Original Dixieland Jazz Band and their connection with Sicily through trumpeter Nick La Rocca and drummer Tony Sbarbaro. Despite the dozens of takes of me switching on tape recorders, boiling kettles and putting ODJB 78s on the turntable, not to mention the relentless questions, it was all jolly good fun! 

Latest Update! The finished film 'Sicily Jass - The World’s First Man In Jazz’ is now out and currently doing a tour of European film festivals prior to general release. You can see a trailer for the film on Youtube at  It looks and sounds great and I’m very pleased to have been associated with it!