I have the privilege of being the first customer to receive AR-T's Legato II USB-S/PDIF converter.
For those who has no idea what the Legato is, it is a reference grade redbook (44.1K) transport for the computer audiophiles. Legato II continues with the same tradition as a redbook only transport, but has upped the ante considerably in terms of sonic quality.
Why no high-res? Many of you may ask. The man (Pat) who created this is a strong believer in jitter. It is the gremlin why digital audio sounded "digital", instead of being "analog". When jitter is extremely low, the sound of a digital transport will become very analog, even a high-res device would have hard time matching the performance. And I was told a lot of this has to do with some physics at work. Suffice to say, it is "easier" to achieve extremely low jitter with a lower frequency clock than with a higher frequency one. And a higher frequency clock is needed for a high-res device. So it becomes very difficult to find an extremely low jitter clock for a high-res device, unless some expensive custom made SC-cut crystal or custom OCXO is involved. The Legato II is designed to be affordable (US$799), so that precludes the use of such clocks/crystals.
So what sort of jitter figure do I get for the Legato II? Audiophilleo quoted their USB-S/PDIF converter having 2.6pS RMS phase jitter from 10Hz to 100KHz. The Legato II measured the same way will have around 0.05pS (or 50fS), about 50 times lower. Not only that, Pat specifically measures phase noise down to 0.01Hz to ensure it is nice and clean at that level. But mostly, he makes sure the phase noise at 1Hz is as low as possible. This is achieved via extensive bin sorting, involving large quantity of crystals in order to find the ones that can make the grade. It is like finding diamonds in the dust. It is a very labor intensive process.
I do not know any audio manufacturer who will measure phase noise down to 0.01Hz. MSB measured their $9995 FemtoSecond Galaxy Clock module down to 0.1Hz. MSB's Femto Clock reaches -67dBc at 0.1Hz. My particular Legato II reaches -55dBC at 0.1Hz.
Besides having a superior clock, Pat also ensures minimum signal reflections along the signal paths, tight control of the output impedance, and high return loss to guard against impedance mismatch on the receiving equipment. All these painstaking details are required to achieve the superlative sound of the Legato II.
The version I received is actually using the final prototype board. The production version will have the same circuit design, but there will be a slight position change of the charge LED and the rear panel lettering will be different. It will have anodized labels instead of laser engraved labels. Mine was a one off unit that Pat was playing with the idea of laser engraving.
So what has changed from the Legato (v1) to Legato II? Legato II is now powered entirely by a SLA battery. A 6v 7.2Ah unit to be exact. With a fully charged battery, one can expect the Legato II to play music continuously for 3 days straight. The unit charges the battery when the PC is not playing music or when it goes into standby/sleep. There is an external 9v DC charger for this purpose, instead of getting the charge from the USB bus. The will ensure the unit will get charged irrespective of the status of the PC.
Besides the battery powered supply, there is now a second BNC output on the Legato II. It is labeled "Direct". The other BNC output is labeled "XFMR". XFMR means the output is digital pulse transformer coupled. This output is the same as the Legato. The pulse transformer is there to provide galvanic (ie: ground) isolation between the transport and the DAC. It is there to avoid ground noise contamination. The Direct BNC output does not have a pulse transformer. It assumes the gear downstream (a DAC in most cases) will already have a pulse transformer in it. In order to isolate the ground connection, all we need is one pulse transformer in the coax signal path, we do not really need two. An extra pulse transformer can add distortion to the signal (even when the distortion is minuscule) and will also slow down the rising edge of the pulse. So why is it so important to have a fast rising edge? Well, Pat told me that an ultra low jitter clock signal has a very fast rising edge, in the <1ns range. By having an extra pulse transformer in the S/PDIF path, the rising edge will inevitably be slowed down.
Ok, enough of the background info. So how does it sound? Well, it sounds very analog indeed. So much so that I thought I was listening to vinyl with a nice turntable setup. Everything flows with ease and yet it can be wild and dynamic if needs be.
I also happen to own the very last revision of the Legato (internally, this was labeled as rev 2). The Legato II has smoother treble (the last bit of digital glare is gone), a darker background revealing more details, longer decay/reverbs, more holographic sound stage, more dynamic sound and deeper bass. The tone density is also higher, with a more engulfing sound stage. The most striking difference to me is how much more ease in the presentation, and how much more analog it sounds.
Some of the sonic references that I am familiar with include: the Sony CDP-R10/DAS-R10 CD transport/DAC, Anedio U2 USB-S/PDIF converter, Anedio D2 DAC, Berkeley Alpha USB-S/PDIF converter, Alpha DAC series 2, Weiss DAC202, and MSB DAC IV Signature. Many of them are superb top-of-the-class devices, but none gives the same vinyl-like presentation as the Legato II does.
I have also auditioned the ClearAudio Statement turntable a couple years back. It is probably the best sounding analog source I have heard. The Legato II is no ClearAudio Statement TT, but it has many of the same sonic attributes that I find a joy to listen to.
As usual, YMMV.
Rear panel view:
Inside the Legato II:
Edited by Viper2001 - 9/20/12 at 10:12am