Airpot Express - Upgrade amp or DAC?


I have a pair of KEF Q300s running through an Audiosource - Amp 100, fed by an ethernet wired Apple Airport Express. I'm running uncompressed AIFF files via iTunes.

What's my next updgrade, a DAC or amp? In comparing the Airport DAC to my Yamaha RX-V1900 DAC, the Airport sounds awful. I'm definitely leaning towards a DAC, but I wonder if people have other suggestions.

Regards,

Andrew
rooty-j
+1 Kijanki. And the AE can sound superb if you kill the jitter. I'm using a Z-Systems RDP 1 between my DAC and the AE and the results are surprisingly good.

That said, one you buy a DAC, your amp will be the limiting factor. Isn't this hobby fun? :)
Steve, I'm aware of changes you made to DAC1. These changes, as I remember eliminated jitter reduction ability of DAC1.

Effects of jitter are not even detectable buried deep in -130dB noise floor according to many different reviews. PLL on DAC1 is used only for the syncing while main reduction of the jitter is obtain by reclocking in asynchronous sample rate converter. Perhaps I don't have as good hearing as you but my Benchmark sounds identical with different sources. It sounds a little better with different op-amps (tried few) but even original NE5532 (TI) sounds pretty good, unless you had earlier version of this amp (thin sounding) made by Signetics before their factory burned down in 2001. TI version has different (larger) die.

Original question was about choice between upgrading amp or DAC. Choice is quite easy since AE has 2400ns peak to peak jitter on analog output and respectable 258ps peak to peak jitter on digital out.

According to Benchmark Media technical director John Siau DAC1 was not designed to sound warm but natural. He stated that warm sound (enhanced even harmonics) does nasty job on instruments (like piano or percussion) with harmonic structure more complex than simple overtones. Piano might even sound, on overly warm gear, like out of tune.

Benchmark DAC1 is very clean and revealing making many systems sound bright. That was the case with my previous speakers that had aluminum dome tweeter. Hyperion HPS-938 I use now are perfect match for the Benchmark and Rowland class D amp. Not only that it sounds perfect to me with extremely clean sibilants and amazing details but also because of AE I don't have to play games with expensive playback programs, computer speed, amount of RAM memory etc.
"Steve, I'm aware of changes you made to DAC1. These changes, as I remember eliminated jitter reduction ability of DAC1."

Only on the I2S interface, not on the S/PDIF input. I am a firm believer in not doing resampling in the DAC, so this interface provides that. Driven with a low-jitter source, this I2S is stellar.

"Effects of jitter are not even detectable buried deep in -130dB noise floor according to many different reviews."

Then why can I easily hear these differences? Why do other reviewers hear them?

For around $1K you could have a Metrum Octave.

Steve N.
Empirical Audio
Steve, Isn't Metrum Octave a NOS DAC? I would need reclocker for best results with AE. Many people go for NOS claiming ill effects of digital filtering (pulse response) but reclocker most likely involves digital filtering defeating purpose of NOS. Did you hear anything about Benchmark DAC2 HGC?

"Then why can I easily hear these differences? Why do other reviewers hear them?"

I don't know. It could be many reasons starting from not bit perfect sources, playback programs that do processing of the data, ground loops etc. or it is possible that my hearing is not that good or I am under placebo effect after seeing FFT plots. I've read reviews claiming audible difference but also read many reviews being unable to detect the difference. One thing for sure - almost everybody praised the clarity, some even calling it "ultra clean" what wouldn't be the case with jitter present.
"Steve, Isn't Metrum Octave a NOS DAC? I would need reclocker for best results with AE."

Yes, but you need a reclocker with upsampling DACs as well for best results with AE. They all benefit significantly from low-jitter sources.

"Many people go for NOS claiming ill effects of digital filtering (pulse response) but reclocker most likely involves digital filtering defeating purpose of NOS."

Depends on the reclocker. Older reclocking chips did have SQ impact, but the newer TI chip is really outstanding. I'll bet the DAC2 uses this one. Have not heard it yet.

Not being able to detect differences, even reviewers is quite common. All it takes is a typical active preamp and all of this will be masked in noise, compression and distortion. You will not hear the difference betwen AIFF and .wav for instance.

Steve N.
Empirical Audio