Lavry Engineering LavryBlue Converter System

Lavry Engineering, the company formerly known as dB Technologies, made their reputation providing "gold" converters for best-of-class studios and scoring stages. Not every facility can afford the gold series at $4,000 per channel, but Lavry AD122 and DA924 converters are frequently found in rooms that can.
Publish date:
Social count:

Lavry Engineering, the company formerly known as dB Technologies, made their reputation providing "gold" converters for best-of-class studios and scoring stages. Not every facility can afford the gold series at $4,000 per channel, but Lavry AD122 and DA924 converters are frequently found in rooms that can. With his new "blue" series, designer Dan Lavry has successfully ported key high-end features into an economically priced product with outstanding results. Novel and rigorous design principles and modern parts combine to make the LavryBlue well worth the attention of studio engineers. Lavry Engineering LavryBlue Converter System
Product PointsApplications: Studio

Key Features: Modular system; 24-bit, 96 kHz A/D-D/A conversion; CrystalLock; varispeed

Price: Two channels start at $1,680

Contact: Lavry Engineering at 206-381-5891, Web Site.

The LavryBlue system (AKA LE4496) begins with a 1RU modular chassis that accommodates up to four stereo converter cards plus a master sync card. Modules currently available are the M·AD-824 A/D, M·DA-824 D/A, M·BY2 synchronous sample rate converter and the M-SYNC clock. Modules can be installed in any combination and the master sync card is only required if there is at least one ADC module installed. The supported sample rates are 44.1 kHz, 48 kHz, 88.2 kHz and 96 kHz, plus a varispeed mode that extends the usable range to 38 kHz - 102 kHz. Additional modules may be offered in the future, such as a possible stereo mic preamp card that the company currently has under consideration. My review model came with an M-SYNC, an M·AD-824 and an M·DA-824. Construction is solid and I experienced no technical problems at all in working with the system, including the installation of cards, which was simple and straightforward.

The DAC module offers a number of design concepts that up until now have only been found in the uppermost class of high-end converters. The most significant concept, in my opinion, is the CrystalLock buffer. Clock jitter has probably created more confusion and myth regarding digital audio than any other single factor. Jitter is inherent in the medium of transmission, not the message, and so identical digital files can sound maddeningly different depending on the condition of the clocks, cables and transports used during playback.

All of these factors are sources of jitter, even though they are all external to the data being transmitted. Retransmit the data under different circumstances and the audible signature of the jitter can be different even though the numbers are not. CrystalLock completely removes transmission jitter from the equation.

The M·DA-824 reads the incoming data into a small temporary buffer while proprietary DSP analyzes the incoming data rate and sets a custom "pullable" oscillator to read the data back out to the converter stage at a matching speed but with the pico-second precision and stability of its own timing circuitry. The DAC is therefore completely isolated from any jitter in the source. Lock time is also very fast and free of digital noise.

If all DACs functioned like this there would be a lot less controversy about various workstations, CD-Rs and pressed CDs sounding different while playing the same files. The effects of passing a jittery clock along to a DAC are most noticeably a lack of solidity and definition in the bass and reduced clarity and impact over all. These problems are not exhibited by the LE4496.

Another high-end concept employed is the use of two converter circuits and two anti-imaging filter chips per channel. The net effect is to increase dynamic range by 3 dB while simultaneously reducing artifacts. Upsampling of single rate sources to double rate before conversion is also implemented, and optimized output buffers provide a final touch of finesse.

The DAC board also allows two additional clocking modes - "wide" and "narrow." Wide is intended for nonstandard and varispeed digital inputs, but its lack of jitter reduction stands in strong contrast to CrystalLock and provides a quick study in how critical a stable clock is to conversion. The narrow mode allows for incoming speed variations just beyond the lock range of CrystalLock, yet still sounds very respectable. These extra modes make the DAC module a problem solver for difficult sources. The DAC module also includes a set of internal jumpers for synchronizing multiple cards to prevent multichannel phase shift. An additional jumper set configures the output stage for unbalanced operation. All in all, the DAC provides not only exemplary sound quality, but also Swiss Army knife utility. One tiny nit is that the digital pre-emphasis bit is not recognized. This obsolete issue never comes up in modern work, but as a mastering engineer, I have encountered it on the extremely rare occasion in very early digital sources.

The ADC module handles 16, 20, and 24-bit conversion at sample rates from 44.1 kHz to 96 kHz. Flat dither and two types of noise-shaped dither are provided. Clock speed is derived from the required M-SYNC module, which in turn can be set for internal or external. The M-SYNC also features a rear panel 2X switch that allows double-speed clocking to a single-speed word clock input, a very helpful feature to have when multiple sample rates are needed.

The M·AD-824 also offers two types of soft limiting, which Lavry describes as analog soft saturation and digital soft saturation. Digital soft saturation adds 6 dB of level to the converted signal in the digital domain followed by a nonlinear transfer function to prevent values over full scale. The intent of digital saturation is to model tape compression and to add apparent loudness. If not pushed too far it does its job well. I tried digital saturation on rock, jazz, and even classical mixes with clean results. Although the unit does not give any indication of how much gain reduction is actually taking place, I gauged about 3 - 5 dB before noticeable distortion occurred.

In contrast, the analog saturation mode is designed purely to reduce the occurrence of unintended overload distortion. An analog nonlinear transfer function kicks in at -3 dB but can also be internally jumpered for -6 dB. The analog protection circuit is clean and provides a safety margin of about 2 - 3 dB before audible artifacts set in.

A single toggle switch is used to both select and set the ADC's various functions. This arrangement is a little fiddly at first, but once you get the feel of it, the desired settings are quick and easy to achieve. Metering is accomplished with a double row of LEDs giving reference level, peak level, and over level indication. I was able to calibrate the converter's input level to 1/10th dB accuracy by lowering the 20-turn front panel pots until the LEDs for the target level just turned off.

In Use

The thing that struck me within the first few seconds of hearing the LE4496 DAC was a surprisingly analog-like sense of ease. The sound was open, meticulously detailed, and effortless. Bass and kick were rock solid and I could literally feel the airspace forming between the cymbals of the high-hat as the drummer closed the pedal. Vocals sounded live in-the-room without a trace of harshness. This was absolutely first-rate conversion. Likewise, I found the sound of the ADC to be completely comparable - remarkably transparent and well balanced.

Of course, the inevitable question for any converter set is how accurate? As an acid test I played some digital sources through my workstation's AES output into the LE4496 DAC, looping the DAC's analog out back into its own ADC, and captured the digital results for an easy A/B against the original files.

Beginning with a 44.1 kHz source and returning through the D/A/D loop at 24-bit 44.1 kHz, the end product of this double conversion was practically indistinguishable from the source. I repeated the test starting with a 96 kHz file (in this case, a new high-resolution recording of the Gorecki Third Symphony with orchestra, mezzo-soprano, and chorus) and redigitized to a 24-bit 96 kHz end product. In a blind test to distinguish the 96 kHz looped file from the source, let's just say that I scored poorly.


The LE4496 series seems to be as close to the ideal balance of cost versus function that I have seen in a piece of pro audio equipment. Lavry Engineering has created a converter that performs far beyond its price. Dan Lavry has done a service to engineers by creating an affordable converter set with all the hallmarks of the high-end.