1
by Frank Wells with Jim Kaiser T he Benchmark ADC1 A/D converter has literally been years in the making. The companion to the hugely successful DAC1 D/A converter, the ADC1 is a reference- quality, 24-bit, 192 kHz-capable, 2-channel converter in a half-rack width, 1U package. The DAC1 set a high standard for rea- sonably priced, high-performance D/A conversion. Well worth the wait, the ADC1 is poised to do the same for A/D conversion. The package is straightforward yet full-featured, and flexible with balanced and unbalanced AES out- puts, a second unbalanced output that can be switched to a dithered 16-bit mode and a TOSLink optical port that can employed in AES or ADAT mode. Balanced analog input is via XLR connectors. There’s also word clock input (WC, Super Clock, AES) and word clock out on BNC. Sample rates are selected on the front panel, where there are also two channels of 9-segment LED metering (switchable to a 6 dB/LED or 1 dB/LED resolu- tion). Input levels are controlled by a 3-position range switch and either the detented rotary con- trols or recessed precision trimmers. It’s always been a treat to put Benchmark gear on the test bench. The ADC1 did not disap- point; in fact, it dazzled me. Pulling up old test files of other manufacturer’s gear as a starting point, the performance of the last device meas- ured is visible on my test gear’s computer display. Running the same test on the ADC1 consistently produced results that were much better than those of a broad array of previously tested gear, so much so that I often had to adjust the display parameters to even see an entire sweep. Frequency response is razor flat (+/- .01 dB across the audio band), with a sharp cutoff just below the Nyquist point. Dynamic range is excel- lent, and the ADC1 measures perfectly linear to well below -90 dB below full output, and within a half dB to -130 dBFS. Crosstalk measurements are excellent, and the noise floor is clear of any discernable artifacts. Harmonic distortion plus noise (THD+N) numbers are simply remarkable (in the 0.000x percent range) under almost all in- put conditions below clipping. That’s close to un- precedented in my experience, certainly so at the ADC1’s price point. Intermodulation distortion test results are also very impressive. The bench results predict an excellent sonic performance, transparent to the source. PSN turned to mastering engineers, Hank Williams and Ken Love, and director of technology, Jim Kaiser of MasterMix mastering studios in Nashville for critical listening tests. The trio com- pared the ADC1 to their high-dollar, “house standard” converter package. They set up in the “neutral” listening environment of their sur- round mastering suite (listening only in 2-chan- nel, of course), using an EMM Switchman II ana- log controller which drove the Nova Applause S/Threshold main monitors and Bryston-driven Bag End dual 10-inch subwoofers. The listening source was a 1-inch, 2-track, 15 ips analog master from a new Sony Nashville artist, played on a Mike Spitz ATR-102 using the ARIA electronics. The song selection was a medium-tempo, solo vocal w/ an acoustic-oriented, light rhythm section. All con- verters were calibrated to a reference level of +4 dBv = -14 dBFS, and no program “peaks” exceeded -6 dBFS. This listening was not per- formed “double-blind”; however, the master- ing engineers did not know which converter was being selected to relate to the source. Williams was first up. At 44.1 kHz/24-bit, he found it “easy” to reliably differentiate between various converters and the source, and to “iden- tify” each converter’s character. Hank described the Benchmark as “a bit laid back in the midrange” compared to the source and refer- ence converter. Williams found that switching to 88.2 kHz/24-bit “closes the gap” between the source and converters. At that point, he re- marked, “The Benchmark is very true to the source and my reference.” The setup was changed to have the choice of the ADC1 or MasterMix’s reference A/D driving their reference D/A, and Love joined the party. Listening first at 88.2 kHz/24-bit, Williams felt that “the gap slightly widened” between source and in this configuration. Love commented that the ADC1 was “slightly more constricted” than the reference, with “slightly less definition.” However, both agreed that the differences were minute. At 44.1 kHz/24-bit, Kaiser reported the differences to be more ob- vious, “as one would expect.” In conclusion, both Williams and Love de- scribed the differences that they heard as “very subtle” and “difficult to describe, as they are so slight.” Williams said he “really enjoyed the Benchmark ADC1/DAC1 combi- nation at 88.2/24-bit, and would have no problem using this equipment on a similar project.” He actually used the word “shocked” to describe how close the pack- ages sounded at higher sampling rates, espe- cially considering the price differences. I predict the Benchmark ADC1 will match the success of the DAC1. It’s a great package at any price, and a spectacular package at its mod- erate actual price. [ 54 ] Worth the Wait Benchmark ADC1 A/D Converter Product information ADC1 2-Channel A/D Converter: $1,775 Benchmark Media Systems www.benchmarkmedia.com The Drawing Board J ohn Siau, director of engineering for Benchmark Media Systems, says the clean, con- sistent electrical performance of the ADC1 comes in part from meticulous optimization of each gain stage: “There are a lot of calculations that have gone into each stage, as far as currents, as far as the types of components that are used—and all those things contribute.” An AKM package was eventually chosen for the converter stage. “We did a lot of evaluation on all the available parts,” explains Siau. The AKM part was selected in part for its low distortion, and in part for its filter characteristics. “We’re not happy with some of the things a couple of other manufac- turers have done in terms of digital filtering. Particularly at 192 kHz, they’ve taken a lot of short- cuts in the DSP processing.” Like the DAC1, the ADC1 employs Benchmark’s UltraLock anti-jitter circuitry, based around a novel use of an Analog Devices sample-rate converter package, for a claimed 100 percent rejection of jitter from the clocking circuits. “One of the things that that does,” says Siau, “is it gives us consistent per- formance. No matter how good or how poor your master clock system is, or whether or not you’re running from an internal clock source, or running locked to an external studio reference, the perform- ance is the same.” Siau says he’s particularly pleased with the “very wide sweet spot” that the ADC1 exhibits across its range of gain settings. “You’ll achieve the signal-to-noise numbers, you’ll achieve the distor- tions numbers, at a very wide range of gain set- tings,” he elaborates. The linear power supply in the ADC1 is also built to perform under even extreme conditions. “We have a lot of immunity to what’s going on in the power supply rails. I can do some pretty terrible things to those power supply rails...and the ADC1 is still clean.” Siau also points to the circuit board layout as critical to the ADC1s performance, “a 6-layer, inside- out construction.” Benchmark’s isolation techniques were developed to prevent crosstalk between chan- nels, between channels and the power supplies, and between the analog and digital circuitry. “That’s a big part of keeping clock noise out of the audio, and keeping all the digital trash out of the audio.” tech @psn in the field pro sound news.com October 2005

54 Worth the Wait inthefield - Music Direct · D/A conversion. Well worth the wait, the ADC1 is poised to do the same for A/D conversion. The package is straightforward yet full-featured,

  • Upload
    others

  • View
    5

  • Download
    0

Embed Size (px)

Citation preview

Page 1: 54 Worth the Wait inthefield - Music Direct · D/A conversion. Well worth the wait, the ADC1 is poised to do the same for A/D conversion. The package is straightforward yet full-featured,

by Frank Wells with Jim Kaiser

The Benchmark ADC1 A/D converter hasliterally been years in the making. Thecompanion to the hugely successful

DAC1 D/A converter, the ADC1 is a reference-quality, 24-bit, 192 kHz-capable, 2-channelconverter in a half-rack width, 1U package. TheDAC1 set a high standard for rea-sonably priced, high-performanceD/A conversion. Well worth thewait, the ADC1 is poised to do thesame for A/D conversion.

The package is straightforwardyet full-featured, and flexible withbalanced and unbalanced AES out-puts, a second unbalanced outputthat can be switched to a dithered16-bit mode and a TOSLink opticalport that can employed in AES or ADAT mode.Balanced analog input is via XLR connectors.There’s also word clock input (WC, Super Clock,AES) and word clock out on BNC. Sample ratesare selected on the front panel, where there arealso two channels of 9-segment LED metering(switchable to a 6 dB/LED or 1 dB/LED resolu-tion). Input levels are controlled by a 3-positionrange switch and either the detented rotary con-trols or recessed precision trimmers.

It’s always been a treat to put Benchmarkgear on the test bench. The ADC1 did not disap-point; in fact, it dazzled me. Pulling up old testfiles of other manufacturer’s gear as a startingpoint, the performance of the last device meas-ured is visible on my test gear’s computer display.Running the same test on the ADC1 consistentlyproduced results that were much better thanthose of a broad array of previously tested gear,so much so that I often had to adjust the displayparameters to even see an entire sweep.

Frequency response is razor flat (+/- .01 dBacross the audio band), with a sharp cutoff justbelow the Nyquist point. Dynamic range is excel-lent, and the ADC1 measures perfectly linear towell below -90 dB below full output, and withina half dB to -130 dBFS. Crosstalk measurementsare excellent, and the noise floor is clear of anydiscernable artifacts. Harmonic distortion plusnoise (THD+N) numbers are simply remarkable(in the 0.000x percent range) under almost all in-

put conditions below clipping. That’s close to un-precedented in my experience, certainly so at theADC1’s price point. Intermodulation distortiontest results are also very impressive.

The bench results predict an excellent sonicperformance, transparent to the source. PSNturned to mastering engineers, Hank Williams

and Ken Love, and director of technology, JimKaiser of MasterMix mastering studios inNashville for critical listening tests. The trio com-pared the ADC1 to their high-dollar, “housestandard” converter package. They set up in the“neutral” listening environment of their sur-round mastering suite (listening only in 2-chan-nel, of course), using an EMM Switchman II ana-log controller which drove the Nova ApplauseS/Threshold main monitors and Bryston-drivenBag End dual 10-inch subwoofers.

The listening source was a 1-inch, 2-track,15 ips analog master from a new SonyNashville artist, played on a Mike Spitz ATR-102using the ARIA electronics. The song selectionwas a medium-tempo, solo vocal w/ anacoustic-oriented, light rhythm section. All con-verters were calibrated to a reference level of+4 dBv = -14 dBFS, and no program “peaks”exceeded -6 dBFS. This listening was not per-formed “double-blind”; however, the master-ing engineers did not know which converterwas being selected to relate to the source.

Williams was first up. At 44.1 kHz/24-bit, hefound it “easy” to reliably differentiate betweenvarious converters and the source, and to “iden-tify” each converter’s character. Hank describedthe Benchmark as “a bit laid back in themidrange” compared to the source and refer-ence converter. Williams found that switching to88.2 kHz/24-bit “closes the gap” between the

source and converters. At that point, he re-marked, “The Benchmark is very true to thesource and my reference.”

The setup was changed to have the choiceof the ADC1 or MasterMix’s reference A/Ddriving their reference D/A, and Love joinedthe party. Listening first at 88.2 kHz/24-bit,

Williams felt that “the gap slightly widened”between source and in this configuration. Lovecommented that the ADC1 was “slightly moreconstricted” than the reference, with “slightlyless definition.” However, both agreed that thedifferences were minute. At 44.1 kHz/24-bit,Kaiser reported the differences to be more ob-vious, “as one would expect.”

In conclusion, both Williams and Love de-scribed the differences that they heard as“very subtle” and “difficult to describe, asthey are so slight.” Williams said he “reallyenjoyed the Benchmark ADC1/DAC1 combi-nation at 88.2/24-bit, and would have noproblem using this equipment on a similarproject.” He actually used the word“shocked” to describe how close the pack-ages sounded at higher sampling rates, espe-cially considering the price differences.

I predict the Benchmark ADC1 will matchthe success of the DAC1. It’s a great package atany price, and a spectacular package at its mod-erate actual price.

[54 ]

Worth the WaitBenchmark ADC1 A/D Converter

Product information

ADC1 2-Channel A/D Converter: $1,775Benchmark Media Systemswww.benchmarkmedia.com

The Drawing BoardJ

ohn Siau, director of engineering forBenchmark Media Systems, says the clean, con-

sistent electrical performance of the ADC1 comes inpart from meticulous optimization of each gainstage: “There are a lot of calculations that havegone into each stage, as far as currents, as far asthe types of components that are used—and allthose things contribute.”

An AKM package was eventually chosen for theconverter stage. “We did a lot of evaluation on allthe available parts,” explains Siau. The AKM partwas selected in part for its low distortion, and inpart for its filter characteristics. “We’re not happywith some of the things a couple of other manufac-turers have done in terms of digital filtering.Particularly at 192 kHz, they’ve taken a lot of short-cuts in the DSP processing.”

Like the DAC1, the ADC1 employs Benchmark’sUltraLock anti-jitter circuitry, based around a noveluse of an Analog Devices sample-rate converterpackage, for a claimed 100 percent rejection of jitterfrom the clocking circuits. “One of the things thatthat does,” says Siau, “is it gives us consistent per-formance. No matter how good or how poor yourmaster clock system is, or whether or not you’rerunning from an internal clock source, or runninglocked to an external studio reference, the perform-ance is the same.”

Siau says he’s particularly pleased with the“very wide sweet spot” that the ADC1 exhibitsacross its range of gain settings. “You’ll achieve thesignal-to-noise numbers, you’ll achieve the distor-tions numbers, at a very wide range of gain set-tings,” he elaborates.

The linear power supply in the ADC1 is alsobuilt to perform under even extreme conditions.“We have a lot of immunity to what’s going on inthe power supply rails. I can do some pretty terriblethings to those power supply rails...and the ADC1 isstill clean.”

Siau also points to the circuit board layout ascritical to the ADC1s performance, “a 6-layer, inside-out construction.” Benchmark’s isolation techniqueswere developed to prevent crosstalk between chan-nels, between channels and the power supplies, andbetween the analog and digital circuitry. “That’s abig part of keeping clock noise out of the audio, andkeeping all the digital trash out of the audio.”

tech@psn

inthefield

pro sound news.com October 2005