Graham Slee Accession

A Second Listen

Author: David Hoehl - TNT USA
Published: May, 2018

[Accession and Jazz Club]

This article supplements Graeme Budd's evaluation last June of the Graham Slee Accession phono preamp (top component, photo right).[1] Taking it from the perspective of one devoted to modern LPs, Graeme liked what he heard. I wasn't surprised; I've been happily playing 78s and LPs alike through a Graham Slee Jazz Club (bottom component, photo right; TNT review of earlier version here) since 2010, from day one considering it very possibly the best audio purchase I've ever made. Graeme's review, however, did contain a few remarks that caught my attention like flares from a Very pistol (emphasis mine):

...While we're on the front panel it also sports a mono switch, 3 EQ curve options and the option to remove EQ completely for recording so you can add EQ in the digital domain if you wish.

This would be a good moment to mention the EQ on the Accession as it's rather unique in that first it EQs the cartridge and then applies the selected record EQ curve. I'm not aware of any other devices that do this....

The alternative EQ curves can be a real bonus for classical collectors who have records that may have been cut with other EQ curves than the standard RIAA most of us are used to. The ability to remove the EQ completely and add it later in the digital domain should put the Accession high on the shopping list of anyone who takes their archiving seriously...

After briefly consulting with TNT powers-that-be, I contacted the company and very promptly (embarrassingly so, considering how long it's taken me, for a variety of "real life intrudes" reasons, to get words onto virtual paper) received a spanking new silver finish Accession for my own assessment as a collector of vintage records.

I'll not beat around the bush. For those of us who collect records made with the horn-and-diaphragm mechanical system prevalent before the introduction of microphones and electric cutting heads in the mid-1920s, the Accession is revolutionary: it demands the user reassess not his gear but his ear, throwing away ideas of how things should sound based on what has gone before and starting afresh.

The reason has much to do with the physics of magnetic cartridges. Good news for you: rather than suffer through any attempted explanation by this non-engineer, you can get the inside scoop straight from Grahm Slee himself in a series of blog posts available here.

Making a ski slope out of a pancake

Therefore, I'll--mostly!--leave explaining first principles to the expert and pass on to their consequences, at least as I understand them. Here's a quick summary of what's absolutely necessary to the discussion: magnetic cartridges inherently have a +6 dB per octave rising response with frequency. The boost starts at the lowest frequency of which the cartridge is capable, and it continues at that constant rate all the way to the highest, meaning it affects tonal balance all across the audible response range. In other words, the rising magnetic cartridge response differs from treble boost in RIAA or such equalization, as it will affect the lowest note of a big pipe organ, or at least its fundamental, not just the fundamentals of those sounding above some middle frequency.

Modern phono preamps routinely compensate for this rise in combination with their setting for RIAA or other curves, but in every preamp known to me until now, playing a record without curve compensation removes the cartridge compensation as well. Collectors of acoustically recorded 78s often seek "flat" output by switching out recording curve compensation or playing the records through a preamp without it, such as a microphone preamp. What they get, however, is far brighter than a true "flat" response.

Enter the Accession

The Accession, as far as I know, is the first phono preamp to address this dilemma head on by dividing cartridge equalization from recording curve compensation. Set one switch on the front panel to "CA/flat" ("CA" for "constant amplitude"), and the preamp will adjust for the magnetic cartridge rising output but apply no other equalization. At last, the collector of acoustic recordings can play them back on a high-quality modern turntable through modern electronics and hear them as they were recorded, without electronic artifacts. The system finally is in harmony with its input.

That was easy. Now comes the hard part.

Friends, Romans, Countrymen, Lend Me Your Ears

If, like me, you've spent years or decades playing pre-electric recordings on modern gear set to bypass playback equalization, your ears are trained to hear a certain type of not quite realistic sound from them, one liberally admixed with fusilades of surface noise, that your "mind's ear" translates into terms of what you know a real orchestra, a real singer, a real violinist, or whatever actually sounds like. Now, abruptly, that sound is replaced by something else, at once more natural and a bit alien, and on the basis of my auditions so far something that can vary pretty wildly from one record to the next.

[Beethoven 9th Sym, Vocalion set]

Shortly before I plugged the Accession into my system, I lucked into a copy of Beethoven's Ninth Symphony in its first complete recording, a rare acoustic set from 1923 in which Bruno Seidler-Winkler led a brave little ensemble billed as the New Symphony Orchestra of Berlin, soloists, and tiny chorus; my copy is a licensed issue on US Vocalion, the record label of the Aeolian piano company, in that company's characteristic red clay colored shellac. This most recent purchase seemed a natural choice as a first test of the Accession.

I began by playing a couple of sides through my Jazz Club as I usually set it for acoustic records, with no treble compensation and as little bass boost as possible (the Jazz Club does not provide for switching bass boost out entirely). Then I substituted the Accession, expecting something--well, I guess, something more like an early electric recording, maybe. Definitely something "better."

What I heard was utterly disorienting.

At first I thought, can this be right? It sounds like one of those awful overfiltered LP reissues from the pre-digital era. And where's the surface noise?

What I was hearing, of course, was the sound of a truly flat playback for the first time. Yes, until then I'd known intellectually that I was listening to a treble-boosted signal when setting my equipment for what I called "flat," but, like Michael Flanders when confronted with the airport sign reading "Beware of Low Flying Planes," I'd concluded "Not much you can do about that." Now the Accession was demanding that I face the consequences of +6 dB per octave without any lazy mental fudging, any thinking "oh, well, it's probably a little bright, but that just offsets the weak treble response of acoustic recording." The Accession was presenting what was on the record, no more and no less, and to one coming from a world of boosted response, suddenly everything seemed so very dim and dull--which is to say, true to the limits of a purely mechanical recording system.

And so my ears began a voyage of rediscovery. Determined to give them every opportunity to explore this new aural territory, I moved on to other records, for the most part chosen more or less at random, on a variety of labels, and gradually, reluctantly, they came to understand this strange new reality. Early on, I found that as a choice for first exposure, the Beethoven Ninth was more "jump in the deep end" than "dip a toe in the water." In advertising the set, Aeolian was not bashful, describing it as "The Greatest Recording Achievement of Modern Times!" The company had good reason to be proud: capturing a work of that magnitude by the acoustic recording process, with its narrow technical limitations and hellish, high-stress environment in the studio, took tremendous engineering skill, pushing the boundaries of the possible, and almost superhuman endurance from the musicians. Nonetheless, even the severely reduced performing forces that could be crammed around a recording horn would perforce have been set some distance away, and on the strength of hearing other records it seems distance from the horn dramatically affected the amount of treble a recording diaphragm would register.

The most obvious way in which the Accession changed the acoustic recording experience was surface noise, or more properly the relative lack of it. I'd been accustomed to tremendously increased noise whenever switching out compensation for electronic curves, and I'd always assumed the increase resulted from removal of what in effect is a high filter, the RIAA (or other) treble cut calibrated to bring an equalized curve back to flat response. The Accession revealed that the increase came not from absence of a cut but from presence of a boost, that +6 dB per octave rising response of my magnetic cartridges. Compensate to bring the response to truly flat, and suddenly conventional shellac 78 surfaces, at least those not damaged by mistreatment, could be remarkably quiet. Not LP quiet, necessarily, but certainly unobtrusive in a way they never were before.

[Edison Diamond Disc]

The next, and perhaps most important, step was a dawning recognition, as the new type of sound became more familiar, that within the new limits of their presentation, instruments and voices alike seemed more natural, that the old progressively boosted way of hearing them made them--well, artificially bright and often robbed them of body. Take, for instance, a Victor record of Schubert's beloved song "Die Forelle" ("The Trout") sung by Ernestine Schumann-Heink in 1911. Now, Schumann-Heink was a true contralto, perhaps the recorded era's most celebrated example of a voice type not often encountered today, lower than the mezzo sopranos who are universal in low female roles on our current stages and concert platforms. Playing the record for comparison first uncompensated and then through the Accession, my initial impression was that the latter way sounded dull, but as the song progressed I started to notice how the voice had a richness and depth not present in the Jazz Club rendition, which upon close listening had a more "soprano-ish" character, a high edge not characteristic of a real voice in her class. The piano accompaniment, although definitely suffering from remote placement relative to the recording horn, also seemed more "piano-like" when heard through the Accession.

A Case Study: Edison Diamond Discs

Collectors of acoustic recordings widely, albeit not unanimously, consider Edison Diamond Discs, or "Re-Creations," to embody the pinnacle of recorded sound quality achieved during the acoustic era. For that and other reasons, I thought they would be fertile ground for testing the Accession.

[Edison sleeve label]

Some background: Edison was a latecomer to the disk format; for a dozen years or so after flat disks appeared on the market, his company resolutely adhered to producing only cylinders. When he finally did enter the disk market, however, Edison made the acoustic era's closest approach to what we consider "high fidelity" today: sound heard emerging from an Edison machine playing an Edison record was to be identical to that which had entered the recording horn.

Edison's thus was the grandfather of all "audiophile" record and equipment labels. The phonographs were expensive, with powerful, heavily built, complex mechanisms machined to a high standard. The records were not like those of other makers: they actually ran at their stated speed of 80 RPM; were precisely grooved at 150 threads per inch, maybe half again the size of an LP groove but far finer than anything else of their day; and played with precision-ground permanent diamond styli rather than the single-play steel needles common to most of the competition.[2] Alas, like its latter day audiophile progeny, the company, under its founder's influence, also tended to make less than ideal A&R choices.

[Edisonic Schubert]

The combination of Edison discs on Edison machines does produce sound remarkable for its day, but I've never found my attempts at electronic reproduction of Edison discs satisfactory, and that includes with the Jazz Club. For a start, on average they have notoriously high surface noise. Moreover, although many record buyers at first considered them too loud when played on original gear, for whatever reason they've tended to yield a faint, thin signal in electronic playback.

Upon playing a number of diamond discs through it, I think the Accession, set for flat reproduction, hugely improves this situation, to the point that electronic playback is not only feasible but actually enjoyable. Flattening the treble boost of the magnetic cartridge helps tame those heavy surfaces, and if the signal remains relatively faint, it no longer sounds thin. The music emerging from those idiosyncratic surfaces actually sounds distinctly accurate, if nowhere near as wide-range as a modern recording.

[Edison Diamond Disc]

Rather than imposing my own judgment on you, I'm offering two examples, one instrumental and one vocal, to let you form your own opinion. In each case, I've included three copies. Two are taken from a modern turntable: one uncompensated, played through the Jazz Club; one flat, played through the Accession. The third, demonstrating how these records sounded on the machines for which they were intended, is captured by microphones from a late 1920s "Edisonic Schubert" model Edison Diamond Disc phonograph.[3] Edison promoted his system by staging concerts, known as "tone tests," in which the audience would be challenged to distinguish between Edison artists performing in person and a phonograph playing their records. Difficult as it is to believe today, supposedly many listeners at these events couldn't tell the difference (although I've read that at least one of the singers involved later claimed they were coached to modify their vocal delivery to match what would be heard from the phonographs).

The selections are as noted below. "Stop" will halt play of any in progress.

  • Russell: "Young Tom o' Devon," sung by Arthur Middleton with orchestral accompaniment, released in 1917. Edison promoted Middleton as a member of the Metropolitan Opera; in fact, he sang only a handful of performances, all in concert appearances or minor roles, during a brief tenure with that company. He was the Edison house bass and in that capacity, under his own name and pseudonyms like "Eduard Mittelstadt," recorded a far more substantial repertory; of his records I've heard, I consider this his best.
  • Bazzini: Le Ronde des Lutins, a showpiece for solo violin with piano accompaniment performed by then-celebrated Czech violinist Vasa Prihoda accompanied by his countrywoman Asta Doubravska. Prihoda was age 20 when he made this record in 1921.
  • A Digression: Pre-RIAA LPs

    Although the Accession caught my attention as a potential tool for the collector of acoustic recordings, it also makes provision for records cut between the end of the 78 era and the eventual ascendancy of the RIAA curve. Besides "flat," it offers a choice of three curves: RIAA, of course, but also two labeled "British" and "American." The former is for FFRR recordings; the latter for curves adopted by US labels, particularly the NAB curve.

    To test one of these settings, I chose an early 1950s record of Friedrich Wuhrer playing Schubert's D. 960 piano sonata in B-Flat, for which American Vox specified NAB equalization. Playback through the Accession set to "American" and through the Jazz Club with settings adjusted for NAB EQ was a near-perfect match. If I could point to any distinction, transients may have been ever so slightly more incisive through the Accession, but the difference, if in fact it was there, was perceptable only upon repeated close comparison of certain passages. In short, compared to "dialing in" NAB equalization given variable controls, the Accession's fixed setting gets it right.

    Limitations and Extensions

    The Accession excels at addressing one issue facing the collector of acoustically recorded disks: it removes the rising response of magnetic cartridges. It does not address shortcomings inherent in a given recording or in the acoustic process generally: it can do nothing about horn resonances, poorly responding recording diaphragms, disadvantageous placement of instruments or singers relative to the recording horn, or the like. For most of my time with the Accession and in the recordings I've included above in this review, I made no adjustments to what the preamp yielded by itself except to normalize the recorded files; I have neither adjusted any aspect of tone nor applied noise reduction, even manual declicking. I can see, however, how attainment of truly flat response might arouse interest in judicious use of graphic or parametric equalizers to address deficiencies beyond the rising response of magnetic cartridges. I would expect the Accession to provide a better starting point than older equipment when applying an equalizer, as it eliminates the need to deal with the sum of issues from the recording and issues from the cartridge.

    To test that hypothesis, I dug out the first equalizer I could retrieve from my closet of audio skeletons. I hooked it up to the Accession's variable level output jacks, feeding the signal from the Accession straight into the equalizer's line input and then the equalizer's line output on to the rest of my audio chain. In the old days, the conventional location for the equalizer would have been connected to a tape loop built into the amplifier, but my Rogue Sphynx, like most modern amps, lacks that feature. Note that the Accession's provision of variable output is likely to prove invaluable in these circumstances, as it ensures a particularly "hot" recording won't overload the equalizer's inputs.

    Having so inserted the equalizer, I proceeded to give the upper midrange or lower treble a very slight boost and made similar cuts in the central midrange down into the upper bass, reflecting where the typical acoustic recording head would roll off and where its output would be disproportionately strong. The adjustments spanned fairly narrow ranges and at maximum amounted to maybe about 2 dB; with equalizers usually "less is more." The result does have a bit more air on top and a bit less system resonance boom, and I don't think it unduly brightens Mme. Schumann-Henik's vocal production as the uncompensated magnetic cartridge did. You can judge for yourself below:

    Conclusion

    With the Accession, Graham Slee has played a part like Toto in The Wizard of Oz: he's pulled back an obscuring curtain to reveal how hidden machinery has been projecting a false reality. What to do with that knowledge, however, is for the listener to decide. In the seemingly eternal battle between the respective champions of tubes and solid state, the argument often arises that tubes may not be "pure," but their deviations from purity are euphonious and actually enhance reproduced music. Some might make a similar statement about the magnetic cartridge's treble boost, at least when playing acoustic records with less than average high frequency content: the extra treble may be "wrong," but it stands in for sound that purely mechanical recorders were incapable of capturing. Others might conclude, as I have, that acoustic records played truly flat may lack "brilliance," but they do give us a better idea of the actual sounds preserved in their grooves. There's no "right" answer here; it's a matter of the listener's ears and willingness to accept one type of limitation in preference to another in an inherently imperfect sound world. I hope the samples above will help you form some initial impressions, but they can't substitute for a longer exposure; love it or not, if you collect acoustic recordings, you owe it to yourself to spend some time with the "truly flat" sound of the Graham Slee Accession.

    The Accession, then, promises to be an essential tool for the collector of early 78s. That said, does it render obsolete earlier preamps designed for the vintage record market, like my Jazz Club? The short answer is, "no." While the Accession, in my view, is a "must hear" component for acoustic 78s, it is not a complete solution for 78s generally. Disks cut between the advent of electric recording and the end of the 78 era, roughly 1925 through 1950, and then on into the first several years of the mono LP era, incorporate various electronic equalizations, and therefore collectors who have 78s from after the acoustic era or non-NAB mono LPs on some labels like Bartok, Concert Hall, or Urania will still want the kind of selectable curves a component like the Jazz Club offers, not just the pre-set RIAA, NAB, and FFRR choices of the Accession. Granted, the Accession's literature suggests copying records to a computer "flat" and then equalizing in software, but that approach doesn't work well when the owner, rather than engaging in a restoration project, wants simply to play a correctly equalized record for pleasure. Moreover, I wouldn't be surprised if artifact-free software equalization requires relatively expensive, powerful software, whereas setting physical switches on the preamp definitely will introduce no conversion issues, and the high level of quality built into Graham Slee products ensures it won't introduce mechanically induced issues, either. Dare we hope for something like a "Jazz-cession" some day, a preamp incorporating both the Accession's "true flat" for acoustic disks and the Jazz Club's flexibility for electrical ones? That would be a dream solution for records of every era.

    What the Accession does render, if not obsolete, then only occasionally needful is application of noise reduction software like "Click Repair" to acoustic recordings. Before making the Accession's acquaintance, I was accustomed to running a light application of Click Repair on nearly every dub of an acoustic 78 (or electrical one, for that matter) to remove the worst noise spikes and allow normalization to yield a better, higher playback level; usually, when subjected to +6 dB per octave, the pops and clicks of even a clean 78 become so pronounced they dwarf the level of the music, frustrating any efforts at boosting its level. During my experimentation with the Accession, by now I've copied a wide assortment of 78s in varying condition, by no means all close to pristine; in only two or three instances have I found noise reduction software needful. Otherwise, a file with a bit of manual click reduction or even the unmodified recorded sound has been perfectly good enough for normalization. The surfaces of acoustic 78s are revealed to be much quieter than I had ever suspected. Now, I regard Click Repair as a remarkable bargain; inexpensive as it is, it yields excellent results in the home, far better than what we heard on professionally mastered 78 reissues during the heyday of the LP. Nonetheless, inevitably such filters affect the recorded sound, if only slightly; the ability to put them aside except in special circumstances can only be a boon. I expect to continue applying Click Repair to copies of my electric 78s, but I couldn't be happier that during my sojourn with the Accession I've largely been freed of it when dubbing acoustics.

    To sum up, then, I can do no better than echo Graeme's earlier review: "I think the best conclusion I can give is that I'm certainly not going to be phoning GSP up any time soon to remind them to pick this one up!"

    [1] - Since Graeme wrote his review, and while mine was pending, Graham Slee has released a version of the Accession for moving coil cartridges. Like Graeme's, my review is of the initial version for moving magnet cartridges.

    [2] - Pathe discs and a few imitators also played with a jeweled stylus, the famous "sapphire ball," but in every other respect Pathe's standards were much more relaxed. That said, Pathe records often yield sound far better than they have any right to, and played on the paper-coned "diffusor" reproducing mechanism some develop positively ear-shattering volume, probably giving a better idea of operatic voices' sheer size than any other records of their day.

    [3] - The Edisonic is actually somewhat later than ideal for this experiment. A last-gasp product issued when the company, again belatedly, adopted electric recording techniques, it features a redesigned reproducer and enlarged horn calculated as an answer to Victor's Orthophonic line. For these recordings I substituted a reproducer contemporary with the records. So equipped, the Edisonic yielded sound reasonably similar to that of my earlier C-250, which developed mechanical issues when I attempted to record it.

    Equipment for this review:

    Turntable: Strathclyde STD 305D

    Arm: SME 3009/S2 Improved

    Cartridges: Several Shure V15Vx-MR, one fitted with the corresponding Shure VN5xMR stylus for LPs (original SME headshell), the others with Shure N78S cantilever assemblies retipped by Expert Stylus with truncated elliptical styli of various sizes for lateral records and one for Edison discs (all in Ortofon 2M headshells). Also a Shure M44-7 fitted wtih Shure cantilever assembly retipped, again by Expert, with a Pathe stylus (Ortofon 2M headshell).

    Phono preamplifiers: Graham Slee Jazz Club, Graham Slee Accession

    Amplification: Rogue Sphynx

    Speakers: Pinnacle BD 650 Mk II

    Digital recording: Edirol UA5 USB interface; Dell Latitude E6410 running Windows 7 Professional, Ocenaudio, Cool Edit 96 and Kastor Free Mp3 M4a Wma Converter software; Tascam DR-100 MK III digital recorder

    Equalizer: Kenwood 1070KE

    Vintage: Edison "Edisonic Schubert" diamond disc phonograph fitted with standard Edison disc reproducer.

    Records for this review:

    Victor 87104, Schubert: "Die Forelle." Ernestine Schumann-Heink, contralto; Katharina Hoffman, piano. Rec. August 12, 1911, Camden, New Jersey.

    Edison 83067-L, Russell: "Young Tom o' Devon." Arthur Middleton, bass; orchestra. Rec. June, 1917

    Edison 82227-L, Bazzini: Le Ronde des Lutins, op. 25. Vasa Prihoda, violin; Asta Doubravska, piano. Rec. July, 1921

    Manufacturer comment

    Dear David,
    Thank you for your review. It came as quite a pleasant surprise.
    It makes me very happy indeed to see the CA/flat function being used, that you understand the device's workings, and that you have explained it so well. I first started questioning the frequency boost of records nearly 40 years ago. It was then explained that magnetic cartridges have a flat output, and the record has a rising response, and that a ceramic or crystal cartridge had built-in RIAA equalisation.

    The explanation seemed to be a little too convenient. I had been working in disco in the mid 70's and had seen Sonotone ceramic cartridges which had the metal cover removed. They looked a little too simple to have RIAA equalisation built-in, to my eyes.

    It was not until I read the chapter on record EQ in National Semiconductor's Audio/Radio handbook (1980) that the penny started to drop. It was so difficult to understand because of my previous indoctrination with the above explanation.

    Slowly my brain started to unravel what it meant, by which time many years had elapsed, what with the pressures of earning a living doing other things in electronics.

    Finally, a few years ago, I decided to have a go at equalising the (magnetic) cartridge, which is very similar to doing an RIAA stage, but without the “kinks” in the response needed for its time constants. Then, by applying record EQ – the “kinks” in such as the RIAA EQ – I got the same results as a normal RIAA stage. This however did not prove the rising cartridge response – the old explanation could still ‘hold water'.

    There was also the fact that numerous magnetic cartridges come with a frequency response plot, which shows a flat output. How could this be? Eventually I landed on Roger Russell's page about Sonotone cartridges, and saw the frequency response plots, the test circuits, and Mr Russell's explanation of how these cartridges work.

    I then had to assume that the flat frequency response plots of magnetic cartridges were somehow manipulated. I created a simulation model in “Easy Spice” and it showed the rising response which was further proof.

    I decided to make the Accession anyway as in my opinion evidence from the National book, my own simulation model, and Mr Russell's explanations were proof enough. Then, about 2 years ago I happily discovered this reprint of an article written in 1996. It is a lengthy read, and still quite difficult to grasp, but it is further evidence that magnetics have a rising +6dB/octave response. And in the reading it can be deduced that a “special record” is used to give the flat frequency response plot which comes with so many magnetic cartridges.

    But you have provided concrete proof in your article that this is the case, and now my mind is set at rest. Thank you.
    Graham

    Dear Graham,
    In turn, what a pleasant surprise to hear from you in response to what I wrote, and many thanks for recounting some of how the Accession came to be. Having lived with it now during an extended period while publication of the article was pending, I can only say again: the Accession, with its truly flat output, opens up an entirely new world for the playback of acoustic records. (I won't pretend that I haven't been putting it through its paces in the interim, to the extent my wife has become quite cross at all the time I'm spending listening to those old records!) If anything, the review unit, which was brand-new-in-the-box when I received it, has become even better with break-in. I can't imagine trying to play acoustic records any other way--and recognizing I'd need to send it on after the review hit "print," for the past month I don't suppose I've played more than one or two records made since the early 1920s. Beyond what I might arrogate to myself, I don't have any authority to speak for the community of acoustic record collectors, but I think nonetheless I can safely thank you on all our behalf for creating this wonderful new tool with which to enjoy our favorite hobby.

    As long as I'm being bold, might I offer a suggestion? The stereo-mono switch on the back panel of my Jazz Club has a third, apparently unused position. I know adding on new circuitry to an existing component is even more perilous than settling new wine in old bottles, but how lovely it would be if there were an option to return the Jazz Club to the factory for addition of a mono "CA/flat" option accessed by that third position. For those who fancy playing stereo records "flat, of course, it wouldn't be a substitute for the Accession's stereo option, but between what it did offer and the switchable EQ curve settings, it would certainly meet the needs of the 78 collector head on.
    David

    © Copyright 2018 David Hoehl - drh@tnt-audio.com - www.tnt-audio.com