Reviewing: Manley ELOP® Elctro Optical Limiter
by Dr. Frederick J. Bashour @ Pro Audio Review
As a classical music recording engineer and producer, compression has not usually been a part of my arsenal of sonic management tools. As a keyboard synth musician, however, I purchased an original Aphex Compellor back in 1987, and it has served me and my music well over the years.
Recently, with the "tubification" of my classical recording chain now complete, I began to wonder how some of the newer vacuum tube compressors would help my synth studio's sound, especially since my current project will employ several vocalists. I was able to get a hold of three of the reputable new tube compressors available: the Manley Electro-Optical ($2,500), the Inward Connections Vac Rac TLM-l ($3,424), and the Eclair Engineering LA-LA ($3,600). All are completely tubed, all are designed to emulate the classic UREI LA-2A design and sound, and each is reasonably expensive.
But the bottom line is that each unit significantly "improved" the sound of just about any program material I fed into it, and in a much more "euphonic" way than my old trusty Compellor.
Theory of operation
As I see it, there are two classic types of compressor/limiters: the Fairchild 660/670 series, and the Teletronix/UREl LA-2A "leveling amplifier." Over the years, both limiters spawned many emulations - and for good reason. These two designs are so different in intent and execution, and so good at what they do, that one hardly needs look any further if one desires the same vintage sound as is present on just about every record made in the sixties. The Fairchild is best known as a program compressor - that is, an overall compressor for the entire mix, while the LA is thought of as a "tracking compressor." The Fairchild's design is based around a variable gain double triode tube, while the LA has a photocell driven by a side chain. (Thanks to "Hutch" Hutchison of Manley Labs for much of the following theory).
The leveling amplifier's traits can be divided into two aspects: electronic and operation. The electronic concept is clean and simple: use the audio to light up LEDs or lamps, which shine onto photo-resistors. The photo-resistor, in combination with a fixed resistor, simply acts as a voltage divider to attenuate the signal. The (vacuum tube) line amplifier functions only to provide extra gain to make up for attenuation losses, and to act as a line driver. "Simple, elegant, and minimal," as Hutch writes in the extensive manual for the Manley EL-OP Limiter. The operational aspects are equally simple, elegant and minimal. There are usually only threshold and gain controls. Most LA-type limiters have no user adjustments for attack, release, ratio or functions for de-essing or the use of external side-chains. The user is thus "stuck" with fixed time constants and a feature list that seems anemic compared to dynamic processors costing far less.
So why are the LA-style opto-electronic limiters so popular - then and now? In my experience, they work correctly on vocals! And, in my tests of these three units, they also sound great on acoustic guitar, bass, and piano. And what of the restrictive, fixed parameters? The attack, release, knee, and ratio are, by design, simply a function of the light-dependent element used. Manley and Vac Rac use a Vactrol Cell (incorporating a LED), while Eclair uses the identical sealed electro-luminescent panel and light-dependent resistor module found in the original Teletronix LA-2A.
The time and slope characteristics of opto-electronic elements are not easy to describe and even more difficult to simulate. The attack is only moderately fast, but still fast enough to "catch" consonants. The limiting action is also level-dependent: at lower reduction levels arid lower peaks the element is slower. It becomes faster with sharp peaks and heavier levels of reduction. Release time is similar to the attack characteristic, but between ten and twenty times slower. Quick peaks are handled with quick release; as the gain reduction nears zero, the opto-electronic element gets slower, "gently braking to a stop," as Hutch puts it.
The slope or compression ratio is also difficult to simulate. The initial ratio becomes higher with more gain reduction until the red LEDs in the Vactrol or the blue source in the electroluminescent panel light up fully and further reduction is not possible. The upper limit of reduction is in the area of 20 dB, but this is a rather severe setting. If one pushes it farther than that, the sound becomes distorted.
So where does one use opto-electronic limiters? As stated earlier, for vocals, bass, guitars, and pianos, but also for brass, saxes, synths, and similar sounds. They seem less effective with percussion or for mixes where the percussion is already well mixed, since the element reacts fast enough to peaks to reduce the level of drums within a mix. Used alone on drum tracks, I have found that individual drums tend to have little of their initial transient let through, but the main drum tone is diminished. But then again, one does not usually use an LA type limiter on a mix or a percussion. That is where the Fairchild comes in.
The three units
The Manley Electro-Optical Limiter is a two-rack space processor with the typical controls: gain reduction, output gain, bypass, meter function (gain reduction or output) and stereo link switches, two large rectangular "modern" VU meters and, uniquely, a stereo balance control. The rear of the chassis contains XLR and 1/4" jacks for its balanced and unbalanced inputs and outputs, an IEC connector, and a fuse.
The Vac Rac TLM- 1 The Limiter is quite different in execution. The Vac Rac 4000 "system" consists of a three rack space high unit containing a vacuum tube power supply (the power transformer is in an external chassis) with vertical slots for four devices, each module measuring 3" wide, 5.25" high, and 12.5" deep.
Inward Connections, the manufacturer of the Vac Rac equipment, currently sells four such modules: the TMP- 1 Tube Mic preamp, the TEQ1 Tube Step Equalizer, the TII-l Tube Instrument Interface, and the TLM- 1 limiter, so one can mix and match. The units are beautifully built, both inside and outside.
The TLM-l limiter's front panel features a very retro-looking hemispherical VU meter, switches for bypass, stereo link, and meter function, a recessed trim pot for meter calibration, and the customary two rotary controls for reduction and gain. In this case, the gain control is an input attenuator. This is quite a different situation from the Eclair and Manley units, whose gain control is on the output side. In effect, it is much more difficult to overload the input of the Vac Rac; on the other hand, this feature makes it more difficult to over-drive it to get a distorted and compressed electric guitar sound.
The "reduction" control is more customary in its threshold-determining effect. However, both the gain and the reduction controls are detented pots, not linear pots, as in the other two units. When queried about the decision to use detented pots rather than stepped attenuators, the manufacturer told me that there was simply not enough room in the small chassis for the attenuators and their resistors. At any rate, the presence of detents makes settings easier to log, since there are only 12 choices. On the other hand, however, the limited choices make it more difficult to dial in precise amounts of gain or gain reduction. In fact, during my tests, I found it necessary to use the Vac Rac unit as the device to which I calibrated the other two, to ensure similar level and threshold settings.
The Vac Rac's rear panel features balanced XLR or unbalanced 1/4" input jacks and unbalanced XLR and 1/4" output jacks. There is also a 1/4" "link in" jack that allows two TLM- 1 modules to be linked for stereo operation.
Eclair Engineering is a small custom design and production company located in western Massachusetts. So far it sells two products: the Evil Twin vacuum tube line- level output direct box and the Model 62 LA-LA limiter, under consideration here. The limiter is built and finished to a very high standard, employs premium parts throughout, and occupies three rack spaces. It features the identical gain control element and control circuitry as used in the Teletronix LA-2A, two large round (unfortunately unilluminated) Weston VU meters, and Jensen transformers for inputs and outputs. Controls on the front panel are the usual input and output pots, named "threshold" and "output level." and four heftv toggle switches - two for meter function and single ones for stereo link and power. Two imposing rack handles flank the ends of the front panel. The overall design is uniformly retro and, in my opinion, quite well done.
Inputs and outputs on the rear panel are through typical XLRs and 1/4" jacks, balanced and unbalanced, respectively. Only the XLRs are driven by the Jensen transformers. Furthermore, the two sets of inputs and outputs can be used simultaneously. One can actually feed each input with a different source (although, of course, any relative level adjustment must be done upstream.) There is also a ground lift switch on the rear panel.
The Ii5tCfling 1e515
The studio's tube monitoring console section was set up so that the same source was fed simultaneously to all three tube limiters. Their outputs were patched directly to three toggle switches so that, by flipping them on and off, one after another, the units could be A-B-C compared in any sequence. Levels were calibrated so that, with no limiting taking place, the apparent volume levels from the three units were identical. The monitor amplifiers were McIntosh MI-200s and VTL MB-300s; the speakers were Manley Tannoys coupled with IMF transmission line subwoofers. Music sources were a combination of live microphone feeds, digital recordings played back from Studer Dyaxis II hard disks, and 30 ips analog master tapes.
For a couple of weeks before testing the units as limiters, I listened to them as line amplifiers to see if there were noticeable differences in sound quality between them. It did not take long to decide that there were. If one thinks of a "sound color continuum" stretching from "very tubey" in one direction, all the way to "very solidstatey" in the other, the three devices under test here arranged themselves in different places on this continuum. My subjective conclusions were repeated over and over again on different types of music. Within a day or so of beginning to switch among the three units, I was able to unquestionably pick which was which.
The Vac Rac is the farthest in the solid state direction, with an extremely clean, slightly dry sound. It is not at all harsh or bright, or thin, just much less "tubey" than the other two. In the opposite direction is the Eclair LA-LA box. It is definitely euphonically colored, in a very similar fashion to my D.W. Fearn VT-2 microphone preamp. That is to say, it makes everything sent through it sound wonderfully smooth. Its low end, for instance, is the most "palpable" of the three units, but it definitely does not sound boomy; its highs are silky smooth, but not at all rolled off. It sounds "just right" to me on most material and I found myself returning to it again and again.
The Manley has a very characteristic sound for me, since I own so many other Manley units. The company makes basically two - very different - line amp topologies. The EL-OP limiter uses its singled-ended "totem pole" version of the classic "White cathode follower" circuit, which tends to sound quite neutral but possesses just the right amount of tube warmth. It also sounds a bit "hi-fl" which, to me, means strong on both the top and bottom. The mids are very euphonic and full of dimension. All in all, the Manley El-Op has a very satisfying sound.
Now that I knew how they sounded as line amps, it was time to test the units as limiters. I set them up for moderate processing no more than ~8 dB of gain reduction. The first source was a variety of female vocal tracks, recorded using the top section of my C 24 last summer, played back from the Dyaxis.
For this source, I preferred the Manley, which enhanced my recording of Judy Malafronte's voice in a good way, making it both larger and clearer. Since my recording of her voice was already smooth and warm, the Eclair seemed to soften her smooth voice up a bit too much. The Vac Rac made it sound less warm than the original, and a tiny bit thinner.
I should stop right now and remind the reader that these "differences" in sound were usually very subtle. In my tests, when I say a device was "warmer" or "drier" than another, I mean, "There's a very small just noticeable difference between them; one sounds a tiny bit warmer, while the other seems just a bit drier."
Acoustic guitar was the next test - a 30 ips analog master tape of classical guitarist Elliot Fisk playing Latin American music. Although the tape was recorded 18 years ago, the Agfa 468 tape seems to have retained all the sparkle and excitement I remember from the sessions. In this test, I preferred the Eclair LA-LA. No matter how hard I pushed it, the limiter just took the music in stride. Even with more than 10 dB of gain reduction on the peaks (when Elliot strummed hard at the end of a passage of quick notes), the limiting action was inaudible (except, of course, to make the guitar sound huge and wonderful). The Manley seemed to add a little brittleness to the top, while the Vac Rac, although it sounded uncolored and was the "fastest" limiter of the three, did not add any "excitement" to the sound, as did the other two.
I guess I should add another note here, for I do not mean to damn the Vac Rac with faint praise. Every designer who works with sound, whether he or she be an electronic engineer or an organ pipe builder, has a certain "sound ideal" in mind. In my experience, all designers try to get their work to sound like their ideal. While I know through personal experience the "sound ideals" of David Manley and Bruce Seifried (designer and builder of the Eclair equipment), it is only through a mutual friend that I know of, Vac Rac designer Steve Barker's, sound philosophy. Basically, he has tried to design vacuum tube equipment that sounds like the best solid state gear, "only better." And, with that aim, he has succeeded. Never before have I heard tube gear which, in a blind test with high-quality solid state equipment, could be thought of "being in the same family." The Vac Rac gear I have heard (my 4000 "Rac" was supplied with two microphone preamps as well as the two limiter modules) is extremely clean, but not clinical, and entirely free of the kind of tube sound that, in the old days, gave tube equipment a bad name. To my ears, it is tube gear for people who like solid-state equipment.
Next was piano. This was a digital recording of mine, done on a powerful Hamburg Steinway at the Konzerthaus in Vienna a couple of years ago. Here, I preferred the Eclair and Vac Rac limiters to the Manley. The Eclair, especially, was particularly unobtrusive in its limiting action -there was none of that "squashed" quality one hears so often on acoustic piano recordings, just the big, bold sound originally captured on that stage, but tailored to sound better on smaller stereos. The Vac Rac also limited pretty unobtrusively but, in keeping with its designer's sound philosophy, sounded more neutral to the source (albeit slightly brighter) than did the Eclair, whose slightly smooth euphonic coloration made my recording sound "better" to me. The Manley, on the other hand, really made the upper register notes of the piano extremely bright and brittle. This surprised me, for it is the first time I have heard something of mine through a Manley product produce a sound which I did not like.
Subsequent research revealed t~ me that the sonic differences I heard have more to do with the choice of gain reduction element (Vactrol or electroluminescent panel) than with any other circuit feature of these three units. The Vactrol elements sound increasingly brighter as they are pushed harder, while the electroluminescent panels just get more "pillowy." Had my piano recording been on the dull side, the two Vactrol-based limiters might have made it sound "better" than the Eclair, the added brightness would have been a "distortion" of the original sound.
A quick test with some highly dynamic digitally-recorded choral music showed an advantage of the Vac Rac's design choice in having the gain pot on the input rather than the output side of the limiter. Since much of my listening tests were with material at full program line level - rather than microphone preamp output level (which is more probably how the units would be eventually used) - I found that the input level settings that sounded fine for the other tests quickly overloaded the Manley and Eclair units, necessitating my turning down the levels at the source device. With the Vac Rac, simply adjusting the little detented (input) level pot calmed things down quickly, eliminating the clipping in a jiffy.
I can make certain informed guesses about the (fixed) parameters of the three units after a couple of weeks of auditioning. I concluded that the "sound" of each of the three processors depends on both its overall quality as a line amp as well as the particular fixed attack and release parameters of its specific gain control element. I suspect that the Manley and the Vac Rac, both of which use modern Vactrols, have more in common in this department than does the Eclair, which uses the vintage electroluminescent panel and light dependent resistor module that the original LA-2 uses. The Vac Rac seems to be the fastest in action, followed closely by the Manley.
It is not that the Eclair is slower - its success at piano and acoustic guitar gave evidence to the contrary. I hypothesize that, rather, the gain control action is more linear with the Vactrols than with "puffy" or "pillowy" sound of the old electroluminescent panels. I got so that I could generally predict how the Vac Rac and Manley would sound, but the Eclair kept surprising me with its "magical" effect on certain program material. Perhaps it is this very "non-linearity" or unpredictability of the action of that old module that makes vintage Teletronix LA-2s so desirable. Whatever source material these devices work well on, they really work well on it!
If you have been following the preceding subjective test results carefully, trying to find the winner, I will bet you have been frustrated. There is no clear "winner" of this shoot-out. All three limiters are great pieces of gear that would make wonderful additions to just about any studio of which I can imagine. My main conclusion is that, if you are in the market for a limiter in this class, check each one out with the type of program material you expect to use with it, for each one has certain things it sounded better on than the other two in my tests.
Furthermore, depending on your own particular sound ideal, one or the other of these devices might fit your application better. Solid state lovers would probably prefer the Vac Rac, while those desiring a bit more euphonic tube coloration would more likely prefer the Eclair. The Manley is smack dab in the middle, sound-wise and, for some reason, costs about $1,000 less than each of the other two - so it is the obvious bargain in this group.
For my music, I lean toward the Eclair. Coupled with my U 47 and C 24 microphones and my D.W. Fearn microphone preamp, I cannot imagine a more flattering front end for the vocalists singing on my ambient electronic medieval music project!
Dr. Fred Bashour holds a Yale Ph.D. in Music Theory, and currently performs as a jazz pianist and church organist, in addition to working as a classical music producer/engineer. He is also a contributor to Pro Audio Review.
The following review is reprinted with permission from the July 1997 issue of Pro Audio Review magazine. Pro Audio Review can be found on the web at http://www.proaudioreview.com.