This high-end stereo compressor boasts three different flavours of optical compression.
There are two fundamental signal-processing tools that are almost impossible to do without in any form of audio recording or post-production: equalisation and compression. Both are available in a wide variety of different formats, flavours, designs, and degrees of quirkiness. The device reviewed here, the Electro-Harmonix NY2A, will score fairly highly on most people's 'quirkiness' scale: there aren't that many stereo compressors that require a massive 3U-high and 14-inch deep rackmounting cabinet, for a start!
Mike Matthews, the man behind Electro-Harmonix, declares in his introduction to the handbook for the unit that the NY2A is 'the best compressor ever made'. I dare say there would be more than a few people wanting to challenge that claim, but how exactly do you define 'the best'? If we are talking technical specifications, the NY2A doesn't look so hot. The distortion figures suggest 0.37 percent is as good as it gets, and the signal-to-noise ratio is a very modest 65dB with 0dBu signals in and out. Of course, if you drive it harder, the noise performance improves in proportion, and a little 'nice' distortion is a thing of sonic beauty for most people. So maybe technical specifications aren't really what this unit is all about.
The NY2A is an unusual stereo compressor in almost every sense. I have already mentioned its gargantuan size (and it weighs a mighty 26lbs as well!), but the controls are just as unusual. Generic compressor controls like Ratio, Threshold, Attack, and Release are eschewed in favour of Pre-gain, Compress, Post-gain, and Squash! But perhaps I'm getting a little ahead of myself here... Let's start at the beginning.
The New Sensor Corporation — the parent company for Electro-Harmonix — also own and import the Electro-Harmonix, Sovtek, and Svetlana brands of Russian-made valves. Consequently, the NY2A is not only an optical compressor, but one which uses exclusively valve-based amplification. However, it doesn't have just one optical compression element per channel — that would be far too conventional — the NY2A has a trio of different opto couplers, each using a different light source to provide three distinct dynamic responses which the user can select to suit the source and musical style, as well as the desired effect.
The first opto device uses an LED as the light source to provide a fast attack time (roughly 10ms — which is slow in comparison to some solid-state compressors), and a uniform response across the entire audio frequency range. The second uses an incandescent lamp, which naturally has a much slower attack time of about 50ms and a rather less even response to all signal frequencies. The thermal inertia of the lamp filament means that it tends not to respond to high-frequency level changes, with the result that it tends to sound brighter than the uniform LED arrangement.
The third option uses an electro-luminescent panel as the light source, and this has a very nonlinear behaviour. A particular characteristic of this kind of device is that it changes colour with input frequency: it starts off a yellowish colour at 30Hz, and progresses through green, blue, and on to purple by 20kHz. The light output also varies considerably with frequency, giving much less light with low frequencies than with high. The result is a compressor that compresses high frequencies a lot more than low frequencies, and is thus a kind of natural de-esser.
The valve amplifiers in the NY2A are anything but 'standard' either. The circuitry does not employ negative feedback to linearise the signal path as most amplifier circuits do. The result is a higher distortion figure on the test bench, but a much better transient response and, many would claim, a better sound character.
Both channels in the NY2A are identical, and there is a linking switch to tie the side-chains together for stereo operation when required. The balanced input is interfaced with a Lundahl transformer, providing galvanic isolation from the outside world, and providing an unbalanced signal for the internal circuitry.
The first amplifier stage employs an EH1 dual-triode valve and provides a fixed 20dB of gain. The signal level feeding the rest of the signal path is controlled at this point with the Pre-gain control, and there is sufficient level available to force the unit into limiting or peak distortion. In effect, this control sets the compression threshold as well as the input signal level, but the threshold is also affected by the Compress control. This determines the proportion of input signal (taken after the pre-gain control) that feeds through to the side-chain to drive the illumination source, and thus sets the amount of compression.
Following the opto sensor, the signal is buffered by an EH2 triode valve stage before reaching the Post-gain control which sets the output level from the device. This is followed by another fixed-gain valve amplifier, using an EH3 dual triode to provide 32dB of gain, and the Lundahl output transformer.
The front-panel controls described above are accompanied by a few additional switches and features. The main controls are arranged in a row along the bottom half of the front panel with four black retro-style detented knobs. The first three adjust the Pre-gain, Compress and Post-gain parameters, while the fourth selects one of the three opto compressor light sources. Each channel also has three silver toggle switches with associated LEDs. The first operates a relay bypass to connect the input directly to the output, the mode being indicated with a green LED when the toggle is flipped down. To the left of the light source switch is another toggle switch marked Squash, and to the right a third labelled Attack — both have red LEDs to indicate when they are active. The former activates a gentle high-frequency shelf when using the electro-luminescent light source to further emphasise the de-essing characteristic of this mode, while the latter doubles the attack time for the incandescent lamp mode (raising it from 50ms to 100ms). These two switches only work when the corresponding light source has been selected, but rather perversely they operate when flipped up — the opposite direction to the Bypass switch. In fact, the panel labelling suggests this is the case (the 'Bypass' legend is written below the switch, whereas 'Squash' and 'Attack' are written above the switches), but it is confusing ergonomics nonetheless.
Above these operational controls are the metering sections for each channel, delineated by a darker panel background. To the left of a large white VU meter is another silver toggle switch which selects either output-level or gain-reduction metering, and there is also a recessed screwdriver trimmer to calibrate the meter to read zero when there is no gain reduction being applied. In addition to the VU meter, there is also a 'magic eye' display valve (EH80) for each channel which always shows the output level. This responds much faster than the VU can, and so is a better guide to signal peaks and overloads. The handbook claims that the 'eye' is calibrated to 'close' at 100dBu, but this appears to be one of a rather large number of unfortunate typographical errors found throughout the handbook. In fact, the NY2A is set up so that the 'eye' closes at 0dBu, and 0VU on the meter also appears to equate to 0dBu at the output.
Between the two channel metering sections lies the toggle switch for stereo linking (up for on, with a green LED), and to the right of channel two's metering section is a power switch toggle (also up for on, but with a blue LED). Moving briefly to the rear panel, the audio inputs and outputs are all via XLR, and the two inputs each have an associated toggle switch to operate at a nominal 0dBu or +20dBu. There is also a ground-lift switch, a fuse holder, and the standard IEC mains inlet. It's worth mentioning that the unit's front and top plates become almost too hot to touch after a couple of hours of use, so the provision of adequate ventilation will be vital when rackmounting.
Whichever way you look at it, this is a very quirky product, which means you'll either love it or hate it — there isn't much room for a middle-ground opinion! Initially, I was probably leaning towards the 'hate it' camp. The sensible engineer's side of my brain had me screaming in frustration at some of the quirks. The main problem was that the thresholds for the three opto circuits were radically different, which made it completely impossible to compare sonic characteristics simply by switching light source. Electro-Harmonix argue that the threshold variations have been chosen to optimise and enhance the characteristics of each opto circuit, but I remain unconvinced.
For example, if I adjusted the Pre-gain and Compress controls to give a measured 8dB of gain reduction on a static tone signal with the electro-luminescent setting, (although I noted the VU meter showed only 6dB), switching to the LED mode produced around 10dB of gain reduction (and the VU meter showed 20dB+). Switching to the incandescent bulb, the gain reduction fell to just 3dB (and the VU meter agreed with that one!). So in practice you have to juggle the Pre-gain, Compress, and Light Source controls in order to find a setting that suits the material being compressed — and that becomes a real handful if you are working with stereo material!
Another issue which I struggled with was the operating levels. While the VU is calibrated to suggest a nominal working level of 0dBu, the device really needs to be driven a lot harder than that to maximise the signal-to-noise ratio. It's not such a problem if you are working with digital recorders and converters expecting peaks of +20dBu, but most budget analogue mixers will be struggling long before such levels are reached. The last of my major complaints was the acoustic hum coming from the mains transformer. Measured with a noise meter about three feet away, the acoustic hum was dominated by a 200Hz component which was over 20dB higher than the background noise when the unit was switched off! I wouldn't want this box humming away in my rack, that's for sure.
However, if you can accept all these quirky features, I have to admit that the NY2A can sound really pretty good. It's not the kind of ultra-transparent and subtle compressor you would want to use for mastering — it's more of a 'down and dirty' effects compressor, designed to help bring out the attitude of a sound source, or to create a dynamic edge and tension. The very lazy attack of the incandescent setting suited percussion well, adding serious character to a dull drum kit, especially the kick drum. It made a nice job of vocals too. The electro-luminescent mode seemed better suited to more complex material, especially electric guitars and heavy synth parts — it made a very good job of controlling some exaggerated synthesised strings, for example. The LED mode was almost on a par with a solid-state compressor — fast, aggressive, and tightly controlling — much more like a peak limiter than a linear compressor, but still very useful in the appropriate circumstances. By juggling the gain controls it is easy to dial in a degree of gentle valve saturation which tends to thicken the sound, the harmonic distortion filling out the mid-range in particular.
At the UK list price, the NY2A has a lot of very strong competition. Within £500 either way, there are the Amek 9098CL, Avalon AD2044, Chandler TG1, Crane Song STC8, Dbx 160SL, Focusrite Red 3, Manley Variable Mu, Neve 33609J, Prism Maselec MLA2, Summit Audio DCL200, and Thermionic Culture Phoenix. None offer quite the same feature set as the NY2A, but they are all extremely fine and well-respected dynamics units, some with their own distinct quirkiness too!
The NY2A is a unique and colourful product that is capable of some interesting and characterful results. For me, the frustrations probably outweigh the benefits, but I'm sure many will find this a very powerful and flexible tool with which to shape their sound. It is well built, with very high-grade components, is easy to set up and adjust, and certainly looks very impressive.
- Three distinctive sound characters.
- Imposing retro styling.
- High-quality components throughout.
- Significant acoustic mains transformer hum.
- Inconsistent thresholds between modes.
- Requires unusually high signal levels.
A rather quirky dual-channel valve compressor employing three different opto elements to provide distinctly different dynamic characteristics. Operation is simple if unconventional, and the unit is built to very high standards.