Quote:
The UHS-6S sounds better than the O2 to my ears. (Got them right in front of me - will do the long term black-box tests next.) The UHS-6S actually has a better suite of measurements which are more meaningful than the existing O2 measurements (which really are very cookie-cutter type measurements):
http://www.leckertonaudio.com/2011/01/uha-6s-harmonic-distortion/
I like seeing across the audio band, THD+N, D2, D3 at
low/realistic power levels (1mW and 10mW) into varying loads (32ohms and 300ohms).
The astute objectivist easily sees how much more meaningful Nick's measurements are compared to others'.
Based on this objective data, people should be going gaga over this amp. For $279, it even includes a USB/Optical/Coax DAC, and it's truly pocket portable too. And as I've said, it subjectively sounds good too - the best portable amp I've ever heard short of that new $$$ ALO piece.
I hardly want to blast Nick for providing more meaningful data than pretty much everybody else on the planet, but I'm really not convinced that what's there is more comprehensive and meaningful than what's available on the O2.
First of all, 10mW is indeed overkill for many headphones, but it's definitely not for some planar magnetics, which have impedance not that far off from 32 ohms. It's also on the low side for some other headphones, at least for a loud peak. Elsewhere he lists power levels for 1% THD+N, but you need to do some guessing with regards to where the knee is. Are you faulting others' measurements for not including more information at lower levels? Would you expect an amp to have higher THD at low output levels than higher output levels (note: not THD+N, since noise will dominate at lower levels) ? I thought a point of benchmarks was to stress the device, knowing that performance under less stress is going to be better.
Secondly, there's nothing on reactive load drive, stability, overload or clipping conditions tests (which could be actually relevant, given the modest output levels it's capable of), or anything of that nature.
Frequency response and range are given but not graphed. That's fine by me. Phase is not mentioned, but there should be low deviation considering the FR.
There's also nothing on channel balance. Sure, channel balance is very good, but it's unlisted and can definitely be a big deal. For that matter, there are no crosstalk results—granted, pretty much every design of this nature should have plenty low crosstalk.
There are also no multitone tests! Of course, with THD+N that low, I wouldn't expect much IMD either. Seems like a fairly serious omission though.
I think the big draw here is plotting D2 and D3 vs. frequency? If you asked me, I'd count these as cookie-cutter type measurements too, not that I negatively associate any of the standard measurements. This isn't like transducers, where there may be sharp anomalies at certain frequencies. (If we were looking at transducers, the first thing other than FR and maybe CSDs I'd want to see is THD / D2 / D3 / etc. vs. frequency.) Anyhow, if THD+N is already low, this already tells you that D2, D3, D4, D5, D6, ... are all low too. Other graphs already indicate that it's usually D2 dominating, sometimes with D3 about the same level, so it's not like there's probably some secret D7 spike that's hidden under the rug. There is THD+N vs. frequency for different loads for the O2 available, as well as spectra for 20 Hz and 20 kHz tests at different loads—in addition to the 1 kHz tests and IMD tests, I would consider the frequency range pretty much covered, considering that we're not looking at transducers.
Anyhow, the issue with the lineup is just output power levels: that's it. And even though these are relatively low, they're plenty for most headphones and most tastes. I agree that more people should be interested.
Absolutely nothing in the design goals about it being a good sounding amp. Effectively what he was saying was that good numbers = good sound.
I'm not that familiar with every posting on this subject, but is it necessarily true that he's saying that good numbers = good sound (yeah, that's the natural implication, but can we say more)? It could just be a design exercise to meet all the criteria, leaving other people to judge whether or not the design was successful in terms of sounding good.