The sample rate converter (SRC) market is definitely heating up; this is fifth SRC circuit I’ve written about in the past year. With the rapidly expanding installed base of equipment handling 88.2 and 96 kHz sample data, the ability to convert up to, and down from, these higher sampling frequencies becomes increasingly important.
Product PointsApplications: Recording and mastering studios, post production and broadcast
Key Features: Asynchronous sample rate conversion from 30 kHz to 100 kHz including varispeed and common pull-up and pull-down rates; 16-, 20- and 24-bit lengths; AES11 and word clock external sync capabilities and single and dual-wire connections; permits conversion between double-wide and high-speed AES/EBU formats
Contact: Lucid at 425-742-1518 Web Site
+ Dual operation permits two separate input data streams to be sample rate converted simultaneously
– Asynchronous conversion – misses that last small bit of ultimate conversion quality (for situations that do not require pull up/downs)
The Score: Unless your budget permits purchase of the significantly higher-priced alternatives, this is obviously the box to get.
Lucid’s new SRC9624 ($1,995) is the latest entry into the fray, and it comes equipped with more features than just about any of its competition.
A single-rackspace unit with Lucid’s new platinum, 3D-look, the SRC9624 possesses multiple I/O and sync possibilities, as well as several other useful features. The rear panel contains two sets of digital inputs and digital outputs in three formats: AES/EBU on XLR connectors, and S/PDIF on coaxial and TOSlink plastic fiber-optic jacks.
Additionally, connections are provided for word clock I/O on two BNC jacks, as well as an AES11 input (on female XLR connector), which can accept either AES digital black signals (clock but no data) or a standard AES/EBU stream (from which it will remove the data and extract only the word clock.) There’s also the ubiquitous IEC AC power jack.
Five three-position toggle switches on the front panel are used to move the selection of parameters up and down. The parameters are: routing, input-A source, input-B source, output sample rate and output dither.
Routing choices are: independent (input A drives output A; input B drives output B); distribution (input A drives output A and output B simultaneously); 96 kHz Dual AES source (two wire in/high speed single wire out from output A), and 96 kHz Dual AES receiver (high-speed single wire in, and double-wide out); and, finally, output dither on or off. The power switch is also on the front panel.
The input source choices are XLR, RCA or TOSlink. All may be connected concurrently at the rear panel and selected via the two input source switches. The digital output choices are the same three, and all connectors are fed simultaneously.
The digital output sample-rate switch determines the sample rate at both outputs A and B; this rate can be chosen from internal clock frequencies of 32, 44.1, 48, 88.2 and 96 kHz, as well as taken from the word clock input, AES11 input or input A. One cannot have different output sample rates on the two sides of the Lucid unit. Thus, one cannot upsample on one side, send the data through an external double-sampling effects processor, and then downsample back to the original rate on the other side, as can be done on the $3,500 Weiss SFC2.
All parameter choices are clearly confirmed by illuminating appropriate LEDs adjacent to the particular labels. Additional LEDs show digital input status – pro, consumer, audio and nonaudio (AC3).
The SRC9624 has an interesting twist on nonvolatile memory storage. Approximately half a second after the last switch change, the LEDs will blink, indicating that the current configuration has been written to nonvolatile RAM. The next time the unit is powered on, it will remember its last state and initialize in that configuration.
The SRC9624 enables pull-downs and pull-ups between 44.056 and 44.1 kHz by making use of an external clock, unlike the Weiss unit, which being a synchronous converter, cannot do. To convert video frame rate signals to standard sample rates, one simply drives the SRC9624’s digital input with the video frame rate digital-audio signal, and selects the appropriate output sample rate via the front panel switches.
To do the opposite – convert a standard sample rate to the video frame rate – one must drive the word clock input with the video frame rate sync signal and select the external clock; the output signal will conform to the sample rate of the external clock input.
To carefully judge the sound quality of the SRC9624 under two scenarios (converting 88.2 and 96 kHz stereo data streams to 44.1 kHz, and upsampling CD playback from 44.1 kHz to 96 kHz), I set up a nifty little input and output switcher using two M Audio C03 boxes – one for inputs and one for outputs.
My input sources were various master tapes I’ve recorded during the past year at 88.2 and 96 kHz, and a representative collection of CDs – including classical ones I produced myself and pop and jazz mixes made by other engineers and producers.
With the flick of various M Audio toggle switches, I could instantly compare high sample-rate playback from my TASCAM DA-78/dual Apogee PSX-100 combination four-track 96/24 recorder, (or standard CDs played back on my trusty Audio Alchemy CDP CD transport) and the outputs of three SRC units: the Lucid SRC9624, the Weiss SFC2 (PAR, 6/00, p. 58) and the SRC section of my dbx Quantum processor (PAR, 1/00, p. 78).
The outputs of the three variously priced SRC units and the two original digital data streams were fed to my Lucid DA9624 digital-to-analog converter and monitored through a pair of Manley Tannoy loudspeakers.
When compared with my original tapes, the three SRC units all sounded slightly different from the sources and from each other. I set them all to output 44.1 kHz 16-bit data and carefully ensured that all were configured for “ordinary” (usually TPDF) dither with minimal, low-order noise shaping.
The Weiss SFC2 sounded the closest to my source and gave a relaxed, liquid impression with the most faithful low-frequency response of the three. The Lucid SRC9624 came in a close second, with highs almost as smooth, and lows almost as deep sounding.
I consider this a remarkable achievement, because the Weiss costs nearly twice as much as the Lucid unit, and is a true synchronous converter, while the Lucid is definitely an asynchronous box. With the dbx Quantum – another unit using asynchronous conversion – set to do SRC, the results were not quite as good as either of the dedicated boxes. The lows, although sounding quite strong, seemed in a slightly different time frame as the other frequencies (almost as if they had been slightly phase-shifted), while the mid-bass was a bit muddy.
The highs were also not as bright and crisp as the Lucid and Weiss units. But to keep things in perspective, although the dbx Quantum box costs the same as the Lucid unit, sample-rate conversion is only one of its multitudinous features.
On all the types of music I played through it, the Lucid SRC9624’s sound quality was close to the Weiss SFC2. I must point out, however, that the data output through its TOSlink S/PDIF connectors – even when using only six feet of good quality plastic fiber optic cable – was definitely inferior to that output from its coaxial S/PDIF or the AES/EBU ports. The sound was rougher, thinner and murkier; this finding corroborates my customary poor impression of digital data transmission in the professional arena via plastic fiber: it is best to use other methods if at all possible.
As I have said before, upsampling playback from CDs to 96 kHz or 192 kHz – and then playing that data back through appropriate high-resolution DACs – provides a level of sound quality significantly improved over straight 44.1 kHz playback, even through high-class audiophile DACs. Both the Lucid and Weiss units are capable of upsampling to 96 kHz, while the dbx Quantum is limited to only 48 kHz without going into analog first; thus I concentrated on comparing the former two boxes. The Lucid DA9624 – as I pointed out in my previous review (PAR, 12/99, p. 32) – is a very good sounding DAC and, thus, would easily reveal differences between the two SRCs.
Both units presented clearer, crisper and cleaner highs (when resolved through the Lucid DA9624) than did original CD playback on my Wadia audiophile DAC unit. The lows, again, were slightly different between the two SRC boxes, with the Weiss sounding a bit larger and rounder and the Lucid appearing a bit tighter and deeper. The important middle frequencies seemed equivalent between the two units, and at least as focused as the source 44.1 kHz playback through the Wadia digital-to-analog converter.
I was pleasantly surprised by how good the Lucid SRC9624 sounded when compared to the source and with its higher-priced competition. It seems I’ll have to relax my prejudice against asynchronous conversion! One must also bear in mind that synchronous converters – for all their theoretically reduced coefficient jitter – cannot accomplish the pull-ups and pull-downs and varispeed tasks, which the Lucid unit can do easily. When a sample frequency converter sounds as accuratel and smooth as the Lucid box does, one needs not feel shortchanged in the sound quality department in any significant way. It really sounds good.