Library #7: Drums And Acoustic Recording Technology – An Experiment

Today’s Library article represents a slight departure into new and unfamiliar territory for DITT, and some of the information I’m going to present below is some way outside my traditional comfort zone. Luckily, I’ve got a very knowledgeable person with a wealth of practical experience in this field to help me out, since this is another first for these pages – something approaching a collaboration…

Nearly all of our Heroes‘ early recording careers were made in studios using acoustic technology rather than electric. The difference between one and the other might not, at first, seem to matter too much – but as we’ll see, the practicalities of the acoustic recording process had a significant impact not only on the way drummers performed in the studio, but also how their contributions were captured, and subsequently heard by us today.

Acoustic recording is a completely mechanical procedure, axiomatically involving no electric power at all in the actual capturing of the sound. Simply put, in an acoustic recording studio the recording artist makes sound into a large brass horn, rather than a microphone. The sound waves travel down the horn and hit a glass or mica diaphragm, which vibrates in turn and transfers these vibrations to a stylus, which cuts detailed grooves into a pliant material – often soft wax. In these grooves is contained all the information ‘heard’ by the horn; a similar stylus/diaphragm/horn apparatus (i.e. a gramophone or phonograph) can then be used to reverse the process, retracing the sound grooves and replaying something approximating the original sound. The original ‘master’ material that the grooves were first cut in is usually replicated in a harder material for durability’s sake – such as hard wax cylinders, and later, 78rpm shellac discs. For some idea how all this looked from the musicians’ point of view, have another glance at the photo above; this shows four members of a put-together group organised by Bix Beiderbecke (far R) taken at the Gennett studio in Richmond, Indiana in 1925. The two huge cones either side of drummer Tommy Gargano’s ‘traps setup are the recording horns. Note the close proximity of him and Paul Mertz’s piano to the horns, and also how close all the musicians are situated to each other. As there was no way of mixing after recording, any balancing of the sound was done purely by the engineers altering the positioning of the artists around the horns, and/or asking them to alter the volume of their playing.

It’s probably at this point that I should introduce the real facilitator and factotum behind this article – my friend and frequent colleague Colin Hancock. Colin, as well as being a fanatical record collector and superb hot jazz multi-instrumentalist, has for many years been running his own ‘Semper Records’ label at his home in Buda, Texas. Colin uses a vintage 1947 RCA radio transcription lathe which he laboriously rebuilt to cut discs acoustically using a period soundbox and sapphire stylus – essentially reversing the process, much in the same manner as the Nordskog label on America’s West Coast did in order to record the first sides by Abe Lyman, Herb Wiedoeft, and Kid Ory. After a few years of fine-tuning the process, he has been able to produce a catalogue of beautifully-crafted and utterly authentic recordings. As a fun mini-project in lockdown during the early months of 2021, he invited me and a small group of colleagues to record some socially-distanced tracks in our respective homes, for him to then assemble and play at a realistic volume into his machinery. Hearing our playing back on a genuine acoustic 78 was a genuinely thrilling and surprising experience; not only were certain sounds captured better or worse than we expected, but the slightly eerie sensation many of us felt was summed up perfectly by one collaborator: ‘I sound like somebody else!’

Playing aside, hearing my own familiar drum sounds back on 78 was a revelation in itself – the frequencies that ‘cut through’, the ones that were lost, and the way the kit blended with the other instruments were all different to what I might have expected, and made me think again about the acoustic recordings and styles of our Heroes. Could an understanding of what the acoustic process did to my sound in 2021 help me to hypothesise how they might have sounded in real life, in 1921? It occurred to me that this experience might make an interesting DITT article – but when I asked Colin if I might be able to use his recordings for it, he very generously suggested that instead I make a special acoustic disc of my own, focused on drums in particular. How could I refuse?

Now in the interests of clarity, it’s important to state that the recordings below mustn’t be taken as completely identical facsimiles of genuine 1920s acoustic recordings, since there were a number of important factors at play in the process that simply can’t be replicated a century later. Firstly, of course, none of the musicians involved were present in person to perform into the horn – this project happened during the Covid-19 lockdown and all instruments were recorded in isolation from each other. The sounds of all the instruments entering the horn thus were hi-fi digital recordings mixed together and played at a realistic volume, rather than the real thing. And of course, sound waves from a speaker do behave differently to sound waves from an instrument. Another important difference to real 1920s record production was that neither wax recording blanks nor shellac 78rpm discs are manufactured any more, meaning Colin had to use his stylus to cut soft vinyl blanks instead. These do work, but are slightly less responsive than wax, resulting in a lower level of signal compared to surface noise. These caveats aside, it’s still a fascinating process to experience, and something that has to my knowledge never been done before with specific reference to drums – so let’s get into the experiment.

First of all, here’s a photo of the drum setup I used for this recording – comprising 28″ bass drum, ‘Snapper’-style snare drum, Chinese tom-tom, woodblock, cowbell, one heavy Turkish-style and one Chinese-style cymbal. A very standard early-20s configuration, as photos of many of our earlier Heroes in action will confirm.

Now here’s a copy of the part I wrote for myself, taking a popular standard of the period (Fred Fisher’s ‘Chicago’ from 1922) and arranging a typical trap drum part for it. It’s based somewhat on the contemporaneous styles of drummers like Chauncey Morehouse, Ben Pollack and Vic Moore, and I tried to showcase all the most important parts of the kit for at least a few seconds at a time. This in itself is a slight anachronism of course, since I’m thus squeezing the amount of textural variation that you might expect over the course of a 3-minute record into a single chorus – but at least this way we get to hear everything at least once.

The same single, identical recording of me performing this part will be used throughout this process as the foundation of all the different versions heard below – this is our baseline. First, here it is recorded to the best of my ability at home, using a modern digital condenser microphone.

Next, let’s hear what the acoustic recording process does to that same recording!

One issue that affects how well we hear a drummer, is, of course, when they aren’t playing alone. Let’s add Andrew Oliver on piano, playing the melody of ‘Chicago’ and improvising an extrovert accompaniment for himself somewhat in the style of a novelty pianist like Frank Signorelli or Zez Confrey might do. Piano-and-drums duos like this were extremely common in cafés and restaurants as well as smaller silent picture houses throughout the late 1910s and early 1920s.

Now let’s see how this piano/drums duet sounds when recorded acoustically.

To complete the experiment we’ll add more instruments to make it a quartet, with cornet (played by Colin Hancock) and clarinet (Michael McQuaid) making up a frontline, and Andrew’s piano now dropping to more of an accompanying role. As always, the very same drum recording used from the start is underpinning the whole thing.

And finally, the same ‘Hot chorus with Band’, as recorded acoustically.

So what (once again, all caveats aside) are the most important lessons we can learn from all this? I have several immediate observations:

Firstly, let’s think about the reality of the process itself. Due to its shape and size, the brass recording horn used to capture the sound, like any confined space, has its own inbuilt sonic bias which naturally amplifies certain frequencies more than others. This, of course, can alter the perceived pitch, tone and characteristics of sound- Colin calls this ‘hearing the horn’. Listen to the sound of the snare drum each time, for example – on the modern digital recording many more of its overtones are equally audible, meaning its pitch is indistinct and fairly close to ‘white noise’. On the acoustic recordings, however, as certain overtones are amplified by the horn and others omitted, the sound emerges seemingly with a stronger fundamental pitch, closer to a musical note played on a piano.

A quick note about bass drum: I know it was an anachronism to use any bass drum at all on an early/mid 20s style acoustic record, where historically, they never appeared – but the whole reason for this experiment was to try to find out why that was! Thus, I stuck to a stylistic 2-in-a-bar bass drum pattern (such as we know, from looking at period sheet music, would have been played on gigs) throughout the track. Needless to say, the relatively narrow sonic spectrum (i.e. not much high or low sound) captured by the horn hasn’t done my 28” drum any favours, and whilst it’s perfectly audible on the solo 78 version (a bit less so on the piano duet), it’s more or less blasted out by the rest of the band on the quartet rendition. When it is audible, its warmth, sustain and bottom end have mostly gone, and all we hear is a dull, percussive thud.

Now we’ll consider cymbals. The limited sonic range of the acoustic records means that much of their ‘shimmer’ and top end is also lost, as is the ‘tail’ of their decay – meaning that even a beautiful vintage Zildjian instrument sounds much quieter and duller than in real life, with a shorter sustain that dies away very quickly. The takeaway for me, therefore, is perhaps that the appallingly leaden-sounding cymbals on records like Baby Dodds‘s with King Oliver’s Creole Band probably are, as has long been suspected, extremely misrepresentative of what those instruments’ true characters might have been when heard in real life.

Going back to the snare drum, one frequently-suggested hypothesis for the unofficial but seemingly-uniform ban on drums in recording studios in the early-mid 1920s has been that the frequencies of snare drums didn’t register at all well on acoustic technology. This theory posits that, rather than drummers deafening their colleagues and ruining band balance and cohesion with a barrage of snare drum just in order to have it appear on the record at all, engineers unilaterally encouraged drummers to stick to those instruments that did cut through a band and record well – the customary ‘recording traps’: woodblocks, cowbells and cymbals. Yet on our recording of ‘Chicago’, the one part of the kit that is consistently the most audible is the snare drum. Compare the two band versions for example. The blocks and cymbals, whilst perfectly audible on the modern recording, almost disappear behind the frontline on the acoustic version, whilst the snare drum remains perfectly clear (if rather cardboardy-sounding). The idea that ‘recording traps’ only came about because the acoustic process somehow gobbled up snare drum sound is evidently erroneous.

The conclusions I’d tentatively draw from all this (again, remembering that this was far from a completely fair test) include the following points:

1) If it was drummers’ bass drums that were really responsible for upsetting the recording apparatus as we’re told, it must have been the physical shaking of the floor thanks to someone playing a foot pedal (which of course I wasn’t physically there to do in this case) rather than their sound alone. The actual sound of the bass drum – whilst not captured very faithfully by the technology – did register, didn’t cause any gear meltdowns, and sounded more or less OK.

2) Leading on from this, it seems woodblocks actually don’t ‘cut’ as much as you’d expect – certainly not as much as they do when heard in real life. To get the really penetrating woodblock sound you hear on many late-10s and early-20s records, the drummers must have been right up close to the recording horn. A full drum kit is a large object, and its presence would have meant that fewer musicians could gather closely around the horn – perhaps this is a viable alternative theory for the prevalence of the compact ‘recording traps’ as seen in the photo at the top of the page?

3) Imagine all cymbals to have been slightly nicer-sounding in real life than what you hear on acoustic records, in general.

4) Snare drums record really well through a horn. The only thing is that they emerge sounding a bit different (and less nice) compared to how you tuned them.

There may well be other thoughts that strike you on hearing these records – do let me know if you feel you have any insights. Either way, it’s been a really enjoyable and enlightening experience having a taster of the real acoustic recording experience, and hopefully one day when things like international travel are possible again I might get the chance to visit Semper Records in Texas and record again with Colin, this time in person.

I’m very grateful to Colin for his faultless enthusiasm and expertise, his excellent cornet playing and his kind suggestion to make the disc in the first place – not to mention his invaluable assistance getting the technical details of this article right. I’d also like to thank Michael McQuaid for vital audio mixing and editing (plus unsurprisingly-exceptional clarinet duties), and to Andrew Oliver for the two equally perfect versions of the piano track.

If you’re interested in hearing more of Colin’s work I’d highly recommend the Semper Phonograph Co.’s channel, which overflows with fascinating band, solo and overdubbed recreations, experiments, explanations and much more.