Originally Posted by pete_mac
The latest audio-gd DI with the USB32 chipset is a clear improvement over the older DI when used as a USB to SPDIF interface. In fact, I can't think of another converter that would surpass its performance at its asking price (around $140 + postage for the unit, excluding the optional external power supply). Audiophilleo-like performance at a fraction of the price (and a class above than the cheaper TE7022 or TI/BB equipped units).
I spent a good part of the weekend comparing two converters. I set up two pc's side by side, both running win 7, 1 32 bit and 1 64 bit. from each of these i ran a usb cable. the 64 bit pc's usb terminated at the usb input on the Audio_gd di v3, the other was run to a Nuforce U192S. i ran a digital coax cable from each to the S/PDIF input on the Grant Fidelity TubeDac-11, and had copies of the same hires needledrops on both machines. i used foobar on both machines to play back. Per the manufacturer’s suggestions, for the NuForce I selected Wasapi event driven output while for the Audio_gd I used kernel streaming.
after a bit of practice i became adept at starting the same tracks simultaneously on both computers, and was able to a/b very quickly by simply swapping which coax cable was input in the back of the Grant Fidelity TubeDac-11.
Before I had worked out the timing of the start-ups, I thought everything sounded pretty much the same from the gd to the nuforce, but heard what I perceived to be a difference in the bass (specifically the gd sounded “stronger/deeper” in bass, but the nuforce exposed more nuance of tone/detail. But I kept becoming befuddled as I couldn’t always hear the differences. I spent a lot of time trying to nail down what I was hearing.
Then I stumbled on a method of executing the start-ups(the perfect lag time that allowed me to switch the cable and hear the music without missing a note going from the audio-gd to the nuforce smoothly. That turned it into an actual a/b comparison and allowed me to hear all the differences in all their glory!
Drum roll, si vous plais….There were absolutely no differences in the sound! Nada, zero, zip. The bass was exactly the same ( it was because I was not hearing it as a real a/b; so that I was hearing an out of sequence comparison – where the bass two measures later IS different: my memory could not retain the exact intonation when there was a gap and I was comparing two spots in a cut that were not concurrent. DUH! That’s why they call it ‘a/b-ing’ and not ‘a/d-ing’.
So on both the Beyer Dynamic T1 and the Sennheiser HD800 (two extremely revealing headphones), using the most revealing tube I’ve ever heard (Amperex usa 1960 white label 6922 usn with a damper on it) I could hear absolutely no change when I rolled from one to the other. The only time I did hear a change was when I removed the outboard power supply from the Audio-GD Div3; at which point there was a significant difference between the two with the gd not delivering what it did with the power supply module.
Testing this on another listener who owns a tubedac-11 produced the same results. He chose the nuforce as superior to the audio_gd when the gd was usb powered; but when it was powered by the separate unit he could not tell which unit he was hearing.
Both sound fantastic, btw.
So, Pete, I highly recommend powering the div3 using something other than the usb power; and although I have no idea of anything else at the same cost as the gd that will outmatch it’s sound quality, the Nuforce at $99.00 currently matches it. (btw, the nuforce is also tiny so that it’ll be great for portable use, and it displays what it’s playing (16/44; 24/96, 24/192 etc).
I am wondering from if the lesson to be learned from this is that, when it’s done right the re-clocking works and should not sound different. Could it be that the differences described by listeners between the various more expensive units are actually a colorization being enforced by them? Or in some cases perhaps the units are not performing up to snuff and it’s simply the jitter still at play?
I don’t know. I plan to study the feedback on more costly converters and when I can give one a try. As for now I am firmly comfortable recommending the USB powered Nuforce (I know, I know..it CAN’T sound ok if powered by USB…unless it can, apparently) as performing at exactly the same level with identical sonic output as the audio_gd.
Audio_gd div3 owners (of which I am one) – I am ready to take the heat. But don’t bother unless you have compared the two side by side, with all things equal.
btw - i listened to hi res needldrops of 45 rpm pressings of Brubeck 'Take Five', Bill Evans 'Moonbeams' and live at art d's, led zep I, bartok concerto for orchestra(reiner, classic 33 1/3), a great copy of natty dread, 33 1/3 at 16 bits, bowie's aladin sane title cut (jpn 33 1/3) tea for the tillerman (audio fidelity 33 1/3); sunshine superman and sunny goodge street(steve hoffman's remaster); some radiohead 45 rpm eps; eric dolphy out to lunch 45 rpm "hat and beard" and coltrane blue train title cut from 45 rpm).