Written by our Staff Editor


Interest in 78rpm recordings has reached a new level of appreciation over the past decade or so. Fuelled by the internet, hundreds of sites have appeared relating to 78s – including our own – where 78s are sold for download and shared or discussed in detail on social network sites like Twitter and FaceBook. In the process, new audiences are rediscovering long lost or forgotten recordings that would have otherwise remained locked away in vaults and not reissued.

Thankfully, depending on where you live, most 78rpms recordings – particularly those made before 1950 are now out of copyright for one reason or another and happily (for us), a new breed of restorers has appeared, showing off their results as well as newly acquired skills at digitising and restoring 78s for others to enjoy.

As one website states in their notes,

“…from Caruso to Callas, there has been an explosion in historical vocal releases in the past decade… many have appeared on enterprising small labels like Romophone, Pearl and VAI, which rely on the generous resources of enlightened collectors. In both instances, a handful of dedicated American sound engineers have produced some of the most significant results. “None of us are in this for the money,” observes Ward Marston, a pioneer in transferring old 78s. “We’re all in it for the love of singing, for the love of old records. Most of us are collectors.” [1]

And at the heart of this new wave of interest in old 78rpm recordings is the love of collecting and restoring long forgotten gems – no matter the genre.

For a new generation of collectors and listeners, these sites are a hidden treasure trove of forgotten sounds and history. Shame then, that for many who make money from reissues – the major companies (originally responsible for recording most of these forgotten sounds in the first place) are far less bothered in reissuing treasures from their vaults – and are either hampered by archaic copyright laws or internet file-sharing illegalities.

Despite this, digitalising 78s by hobbyists and semi-professionals continues unabated.

With each new restored share – the fascinating process of creatively transforming old recordings grows and improves -spurred on by discussion and debate.

This, rather lengthy, two-part article delves into the fascinating world of the intricate and time-consuming processes of transferring, digitalising and restoring 78rpms and what makes the ‘restorer’, tick.

So how do restorers choose what to restore?

For many non-commercial/hobby restorers, it comes down to personal choice and preferences. While one restorer may prefer jazz and swing – others may prefer classical or operatic. In most cases, the restorers we interviewed for this article are experienced collectors of 78s, and as a result have preferences for certain genres of music. Interestingly, choice can also be driven by collecting habits, a personal liking for certain performers as well as scarcity of a particular type of record. As Bill Anderson; an active member of the 78rpm Collectors Community, noted in our interview with him,

 “the lack of commercial reissues, or poorly engineered reissues, are also contributing factors. I sometimes transfer discs that have unique historical importance, culturally or politically (e.g. the Talich/ CzPO ‘Sokol March’ recorded at the time of the Munich Agreement in 1938)”[Ed. You can hear Bill’s excellent transfer of this 1938 recording below:]

Sokol March – Talich/ Conducting the Czech Philharmonic Orchestra (HMV 1938)


Of fundamental importance is the need to obtain the very best source material – be it “mint” copies of 78s, “used” shellac copies found in second-hand record stores, online or at record fairs, or even the original metal masters of a particular recording. By far the best source material is the metal master.

“The metal positive, compared with a shellac pressing, is astonishingly quiet and relatively click free.” [2]

Original metal positives or ‘mothers’ from which stampers pressed 78rpm shellac records is a generation much closer to the original wax disc. Often then not, they do not display anything like the imperfections that a 78rpm record does. However for many collectors, access to these masters or metal stampers is often impractical or downright impossible. In the majority of cases, “clean” 78rpm shellac pressings are used, but with the correct playback stylus and turntable and the use of modern computer software, these ‘secondhand’, ‘clean’ shellac pressings can still produce wondrous results.

The Restorer as Editor and Censor:

Whether aware of the fact or not, many restorers of 78rpms are editors and to an extent, censors. Often not, 78rpm restorers select material to restore based on individual likes, preferences for music genres and to some extent, public interest. Through individual, religious, political, cultural or personal choices, they select, dismiss or promote certain types of music over others.

Thankfully, choices are rarely criticised. In many cases, many restorers have extensive blogs which they use to explore and discuss recordings they restore – and have large lists of followers who promote a healthy, if not often, heated debate on selections. Many of these blogs feature lengthy comments and opinions from listeners – our own site, 78rpmcommunity.com – not withstanding.

Some interesting blogs worth exploring include:


Because there are many elements to transferring and engineering a recording to digital form, a variety of software is often needed. One, well-known restorer, who we interviewed for this article, had this to say.

“For the initial transfer to the digital domain, I use DC8. I use this for some basic de-rumbling, before then de-clicking, equalising and de-hissing in ClickRepair, Equalizer, DeNoiseLF and DeNoise, all created by Brian Davies and available at clickrepair.net. After all of this I merge the recording down to mono. For speed adjustments and for stitching multi-side recordings together, I go back to DC8. I use HarBal to carry out fine-tuning of the equalisation, and possibly then DC8 or DeNoise to carry out further de-hissing after this, and other adjustments – reducing any residual thumps using spectral editing, and so on.”

“With acoustic recordings, I will tend to perform major adjustments to equalisation, in order to achieve a more natural instrumental sound, and in this case, noise reduction is better carried out afterwards.”

This contrasts with another restorer – Bill Anderson, who notes,

“…using the Algorithmix tool … is a good starting point for my transfers. If the records are fairly clean, then I decrease the settings, the opposite for worn or problematic sides. Also, de-clicking takes more than one pass, sometimes as many as four passes are needed to get the desired results. The key is to listen carefully and when the high end musical signal starts to dull, I back off the filter settings and/or bump up slightly the high frequencies with an equaliser.”

What is obvious, however, is that no matter the approach taken to restoring 78s, a variety of software is essential to achieve optimal results. In most cases, the inherent “scratch” found on even the cleanest of 78rpm pressings can distract and hide upper and lower frequencies embedded on shellac discs. In many cases, these “lost” frequencies are only recoverable when filtered by software.

As Bill Anderson (and many other restorers) know from experience, software is only part of the process in restoring 78s to listenable tolerances. In many cases, multiple layers of software intervention is required to bring a 78rpm to listenable levels. Take Bill’s examples below which received three passes to achieve an acceptable level of tolerance.

In the example below, Bill transfered the 78rpm as raw files using the following tolerances: 300Hz turnover, -5 db rolloff at 10kHz stereo.

During the first filtering, he employed DC/ART: Average Filter setting 2 ( a ‘ripple’ notch filter with a moderate drop after 10Khz) and used Algorithmix: 78 declick/decrackle settings.

During the second filtering, he converted the files (both sides of the recording) to mono and then applied DC/ART: Average Filter 3 (steep notch around 12kHz)and used Algorithmix: moderate 78 settings ( the one he uses for most 78 work).

In the final filtering process, he used Algorithmix again using the “moderate vinyl” settings and applied DC/ART: Continuous Noise Filter (CNF) – 1st setting moderate. To finish off, he added, Paragraphic equalizer to bump up higher frequency to compensate for the CNF effect – (Continuous Noise Filter, 2nd setting light).

The results below illustrated the differences in results and quality:

Kikimora Op. 63 (Liadov) – RAW transfer

Kikimora Op. 63 (Liadov) – Final


One shortcoming however of shellac records that cannot be denied is the material – originally “shellac”, made from the secretions of an Indian scale insect – often has a larger grain than vinyl. This surface compound/formula limited the upper frequencies obtainable in playback, and the actual structure of the record material is audible, creating surface noise.

In “modern” remastering to digital media, sound engineers do everything they can to eliminate this noise, but at the same time can also mistakenly eliminate essential components of the original sound. That’s the main reason why many restored recordings sound dull at best, and usually catastrophically bad.

Most modern playback equipment now-a-days accentuates the scratch found on 78s – so some filtering of the original sound needs to be controlled, reduced or eliminated with software or playback hardware. Thankfully some software is freely available on the internet, while other software can cost many hundreds of dollars.

Listen to the examples below of Percy Grainger playing a Chopin sonata to see how ClickRepair effectively removes unwanted ‘noise’ from 78s. Recorded 1925/Shellac pressing/Electrical recording:

No Processing:
Featuring Dehissing and Declicking:


Hardware is just as important in transferring a 78rpm as software. The ‘correct’ type of stylus, turntable and record cleaner can make the difference between a good transfer and a poor transfer.

Finding the most suitable stylus for transferring a 78rpm recording is important and choice can significantly alter the character of the recorded sound. As one of our interviewees replied,

“I like to choose a stylus not necessarily by the age of the recording but rather by the resultant sound effect it will achieve. I often change a stylus during the transfer of a side, particularly if excessive wear or ‘chatter’ is evident. Despite the available science, cyclic swish is still a problem, although varying the stylus can often minimise this effect for the listener.”

Article continued in part 2

[1] – Marston Records
[2] –http://voxnovamedia.com/lehmann/llf/intro/llf_introduction18.html

CONTRIBUTORSMany thanks to the following contributors to this article:

  • Damian Rogan
  • Bill Anderson

SOFTWARE: (Software mentioned in this article)

OTHER SOFTWARE: (The list is endless…)

EQUIPMENT: (The list is endless…)

1 Comment

  1. Romophone was a fine label, but they’ve been out of business for about 15 years. The quote is quite old. Ward Marston used to do a lot of transfer work for Romophone and now he’s had his own label for some years and still chugging along. Mark Obert-Thorn was another well regarded engineer who did work for Romo and other labels. Haven’t heard of anything he’s been up to in awhile. Which isn’t to say he isn’t up to some projects somewhere.

Leave a Reply