This review page is supported in part by the sponsors whose ad banners are displayed below


Then I checked on USB with Light Harmonic's LightSpeed cable. Now I got nasty crinkly distortion. One glance at my Audirvana panel suggested probable cause. I'd not reset its custom sample rate preferences from the 32/384 AURALiC Vega.


If you've read the Amanero specs you already knew that this USB transceiver module is fully 32/384 capable. As that's what my iMac saw (not the circuitry past it) Redbook files streamed as usual: at 32bit/352.8kHz. Once I reset upsampling to 44.1 --> 176.4, all was peachy. Audirvana's audio system panel strangely missed 352.8. Just to be sure I reset 44.1 --> 384. Same distortion. The DS5's D/A conversion can't process such data. Neither does it claim to. All was fine.


PureMusic's equivalent panel did show that the DS5's USB input is compadre with 352.8 just as you'd expect with 384. I don't know why the competing player skipped it.


As you saw this machine also is integer mode compatible. This will appeal to Macists because this feature puts to sleep even more background computing for better sonics.


Whilst I was busy not being patient I also quickly checked on DSD 64/128. Instant joy. This crossed all the 't's. Ami's DS5 worked just as advertised. No fuss, no muss. Time to let break-in run its course.


After proper aging as Ryu had called it—I quite liked the tie-in with fine cheese and wine—I moved to the desktop. Here my HP Z200 work station on Windows XP Pro likes to fuss and muss over driver installs.



But instant joy was the word again. To offset my Gallo Strada II's luxo leanings for the occasion, I moved in the sonically trick but on cost pedestrian Clone Audio 25i. This brought up a missing feature. The original clone lacked a subwoofin' pre-out. Funjoy since converts of one of his three inputs by request. But my unit doesn't do 'out'. The DS5 only does it once. Without a 1-into-2 adaptor this took away my Gallo TR-3D subwoofer's line-level hookup. Sniff. Note to Ami Musik. As desktop preamp your DS5's usefulness would go up considerably had you at least a second pre-out on 3.5mm stereo given how scarce your real estate is around back.

Streaming Spotify+ for a 'which should I buy' sampling of five Caribbean Jazz Project albums
I keep my work machine free of music files but routinely stream Spotify+ at 320mbps to discover endless new music. In went a Telos Audio USB leash. My RWA-modded Astell&Kern AK100 streamed up to 24/192 (yes!) via AudioQuest's Diamond Toslink. My 160GB iPod Classic streamed 16/44.1. Running all three sources simultaneously, the 'in' button afforded instant comparisons. Despite the often dissed optical connection (the Diamond uses glass fiber not plastic) hi-rez files off the AK100 led the pack.

Adapters double up on the outputs to feed both Clone Audio 25i and Gallo TR-3D subwoofer
Since 99.9% of my music is CD quality this was a mute point but audiophiles do love to obsess over such matters. Here's one I consider more relevant if one needs beaucoup attenuation: digital vs. analog volume. Players like PureMusic offer 64-bit dithered volume control in software. Ami's is 24-bit. Would this bit strip to throw away resolution or lean out the DAC chip's output stage by not driving it full scale? I could use either Clone's or Ami's volume. The former is a quality passive Alps, the latter the CS4398's on-chip interpolator.

With high 30dB voltage gain, the Clone at full tilt (bypassed Alps) forced the DS5 to operate at just 8 clicks up. That meant two LEDs on its SPL scale, 7 clicks for the first light, one more to hit the next. No comparison! Analog control crushed digital. To diminish the damage I had to back off the Clone to about 10 o'clock. That meant 7 full LEDs on the Ami. Now I didn't hear any losses. Operated as a fixed-out DAC without attenuation eliminates this inherent weakness. As digital preamp with obvious attenuation it all depends on overall system gain and desired SPL. The fewer LEDs you light up, the more your SQ will suffer. Obviously that's true also for headphones. I'd say the ideal loads are those which run the DS5 above five LEDs.

With the AK100 and for normal levels, Aëdle's VK-1 triggered 2. It got mega butch at 3. AKG's K-702 played 4 for mellow and 6 for hot. Audeze LCD-3 came on song at 3. I'd not exceed 4. Ditto for Sennheiser's HD800. MrSpeakers' Alpha Dog lived between 4 and 5. Without any distortion HifiMan's HE-6 got louder at full crank than I'd need with Buddha-Bar type tracks of limited dynamic range which sound louder. For once inefficiency pays dividends. Get ye a gas guzzler Simon sez.
Streaming 24/192 via optical and MrSpeaker™ Alpha Dog
This didn't mean more efficient 'phones sounded poor. Far from it. I just knew how they left stuff under the table. From my grab bin and cognizant of price matching the very best headphones with the DS5 were Dan Clark's rebuilt $300 and $600 Fostex planars and AKG's $449 K-702. I'm unsure why Ami's engineers didn't go for a quality analog pot* as Asus do for their $399 STU and three Essence One versions. Not only would sound quality be consistent regardless of app, an easy visual fix of where the volume sits wouldn't need twiddling to trigger a light show. If I were part of a MkII team I'd lobby for analog volume and a 2nd pre-out. Everything else is dandy. As a fixed-out converter in fact the DS5 really struts its stuff.
____________________________________________________________________________________________________

On this topic one also wonders whether the 6dB gain compensation for PCM (which at present will likely be 90-100% of actual usage) already occurs in the digital domain or is implemented with resistors.