Searching for the Perfect Delay
by Phil Taylor
Delay is one of my favorite effects because it creates a big sound which inspires musical ideas and is an essential effect for the progressive player. For a long time I was unware of fantastic analog echo machines like the Binson Echorec and Copicat - here I describe my experiences of delay technologies in a search for my ideal tone.
Solid-State Experimentation
The first delay I ever owned was an Arion SAD-1 (a melancholy acronym) which was an expensive item for a 17 year old back in 1983, costing about £60. This was capable of 300ms delay, however the repeats were very muddy. It's likely the manufacturer was squeezing maximum delay length performance from the MN3205 BBD chip by under-clocking and therefore using excessive low-pass filtering to remove clock feed-through. BBD technology is almost universally (and incorrectly)referred to as being analog, but strictly speaking it is hybrid digital/analog since the signal is sampled in the time domain at the clock frequency. The delay length was adequate for most purposes, allowing me to replicate the intro to "Kid Gloves" and "By-Tor" by RUSH. The pedal was also capable of self-oscillation and therefore it was possible recreate classic 50s sci-fi effects as well as short delay and slapback echo.
Moving forward several years, my next delay was a quantum jump up in technology to an Alesis Quadraverb, missing the Boss DD-2 step, which was the current "hot" delay for the time. This was a small step for me (down the street to my local music store, Axe Music in Colchester), and a giant leap for my bank account to finance it! The quality of delay effect from the digital Quadraverb was a much different experience to analog BBD technology. The reverb algorithms were lush and the unit was capable of huge, spaceous, clear delays with multi-tap and ping-pong options, opening up a whole new realm of creative possibilities. This was a musically inspiring device! It could also recreate chorus, phase, flange and equalisation effects, however these were very weak sounding, lacking the presence and body of analog based pedals. At the time I put it down to the clinical nature of digital gear, however, in hindsight part of the problem may have been down to poor algorithms. I don't know the architecture of the Quadraverbs' internal DSP circuitry, so am only guessing here. Additionally the delays are overbright and brittle, resulting in "clatter-like" sounding repeats. Here are some recordings I made on a Sony Pro-Walkman back in 1993 - these sound sample excerpts from "Cold Sunlight" and "Lost Girl" exemplify what I'm describing. Some high frequency roll-off would definitely help to make the repeats sound less artificial and make for a more natural sounding delay.
Learning from the Past
My first experience of tube echo, must have been in the mid-90s when I happened to stumble upon a Copicat unit at a small-town market stall in Retford, Nottinghamshire. What a find! It was there amongst a clutter of outdated electrial junk such as toasters, electrial wire, domestic light switches and the like. Although in bad shape cosemetically and electrically, everything was there inlcuding the case and a very rusty footswtich. "Now play it cool don't look too excited, maybe this guy doesn't know what he's got", I said to myself. . .
I hurried home with this little bit of treasure, anitcipating the fun I'd have rebuilding and then playing through it. A rebuild involves replacing dried electrolytics, ceramic coupling caps with polyester types and drifting, out of tolerance noisy carbon composition resistors with metal-film equivalents. The ease of access to the internals and tag-board construction makes these units eminently servicable and consequently they are almost impossible to "write-off". Although, the original tube Copicat was a budget unit at it's time, now, almost fifty years on, came back to life to create a wonderful, rich mult-dimensional delay sound.
The primitive tape transport in these units results in considerable wow and flutter creating a warbling chorus effect. This is considered a desriable attribute by many enthusiasts, however, as an engineer I'd like to be able to exercise at least some control over this to minimize it or increase it and not merely accept whatever performance the lousy transport produces. Additionally, the playback record electronics are very lo-fi resulting in a degradation of the delay repeats. This manifests as a rolling off of high frequency content with successive repeats, making the repeats become progressively muddier sounding. Again, this can be quite useful effect, as the repeats don't conflict with dry signal and sound more distant. To the ear lack of H.F. content in a sound, as well as volume is one of the mechanisms the brain uses to judge the distance of an object - so the tape medium recreates a natural sounding delay. The repeat is not just simply low pass filtered though, it's more complex than that. There are probably other effects that add to degradation of the repeats signal such as tape saturation and the response of the tape heads themselves. To reiterate, these shortcomings of a mechanical tape delay are not neccessarily a bad thing and are considered by many enthusiasts as the magic of these old tape echo units, however it's the beautiful, warm and involving sound that I find so special.
There is a trend to develop stompbox delays that emulate the sound of tape delay by simply low-pass filtering the repeats. Some pedals also feature wow and flutter controls to vary the repeat pitch as a tape unit would do if the tape transport was inconsistent, for example the pinch roller was distorted or out of true. For me this is missing the point. These are the shortcomings of tape delay, not the desirable attributes. Tube based tape delay has a lush quality that is seems somehow intrisic to the technology being used. Although I have a nostalgic love for the wow and flutter and "happy accidents" it can create, focusing on that so much in most tape emulation delays is truly missing the point. The warmth and clarity of the pure tube signal path is what really makes a delay outstanding. Everyone seems to think that repeats need to get dirtier as they decay. I don't understood this - I prefer the sound of echos merging and melting together in a lush swell with higher feedback levels. The repeat degradation, wow and flutter has it's place, but there's much more to the tape delay that merits consideration. For reason, I'm looking into improving the quality of magnetic delay technology and hope to make some geniune advances and contributions. I'm using the Binson as a platform on which to do this.
The Binson Project
Tube tape delay still offers warmth, clarity and involvement cannot be satisfied by modelling, digital or analog bucket brigade technologies. For this reason I think this technology merits some serious further investigation! Fortunately I'm privileged to have owned and played through vintage tube tape echo units from the fabulous Watkins Copicats to the more the highly sought after and prized Binson Echorec drum delays. I'm considering an approach of using a Binson as the building block for a high performance delay unit project. The approach is to take the best magnetic transport available, the Binson and uprate the electronics. Binson UK have already made improvements to the transport by adding varispeed control based on a D.C, motor manufactured by Trident Engineering Ltd 2-40/1212. Further to this, I'm looking at designing a new printed circuit board with high quality polyester coupling caps, metal-film resistors and D.C. heaters for the tubes. I'm also considering the option of photo-optical companding noise reduction circuitry to improve signal-to-noise of the record/playback process to produce clearer repeats. It should also be feasible to reduce the physical size of the unit substantially. I'd like to add a send and return so that you can introduce effects into the repeats. So, for example you could put Vibra Chorus in there to add some swirl on the repeats or an overdrive to add controlled desgradation of the repeats. There are many possibilities. With all these improvements we should have a very nice delay unit!
At it's time the Binson units were considered the top of the range echo. Binson, Milan became involved in a research effort to develop a storage medium that was an improvement over tape to create the delay. The result was a unique device, utilizing a specially designed steel/alloy disc or drum, which carried a durable flat metal 'tape’. The drum was driven by an A.C. motor via a rubber idler wheel (similar to a Garrard 401 turntable). Record and playback heads were arranged around the edge of the drum. The stability of this transport was a significant improvement over tape with reduced wow and flutter. The Binsons still suffer from excessive noise though and too much degradation on the repeats, This is my only concern, that despite being more stable than tape, the drum may have inherently poor fidelity in comparison to tape - this is something I need to quantify.
I'll be adding some more informatiion here shortly - so do stay tuned. This is a long term dream for me and will almost certainly take a few years to refine. The Binsons are a starting point and a pretty good one at that because they use a drum which is more stable and lasts indefinitely if properly maintained.
