Quote:
Originally Posted by cegras /img/forum/go_quote.gif
Um, go to headroom. There is a wealth of frequency response data (for some amp) with headphones.
|
If you're talking of the «Build A Headphone Graph» section: The (seeming) deviations from amp to amp (or rather from single-ended to balanced drive) are the result of measuring variances.
Quote:
| Furthermore, tube amps fall within the general category of amps. |
Yes, of course, and I'm not pretending that they sound any worse than solid-state amps, not even clearly less neutral. They have been left out because they don't guarantee measuring data low enough to be considered below the hearing threshold.
Quote:
| It may be possible to say ss amps can all be made to sound the same, but since tube amps offer much more distortion and fall under the category of 'amps,' then the conclusion is that there is significantly more flavour to be had by switching amps (the differences being electrically justified) than by switching cables. Although the magnitude of the difference depends on how much you want to believe. |
I don't entirely agree on the magnitudes, because in my experience cables can make a similar degree of difference to my ears. As to tube amps' higher distortion versus solid-state amps' lower distortion: Interestingly the sonic differences among tube amps aren't any larger than the sonic differences among solid-state amps. Even in terms of neutrality they aren't clearly inferior, not in every case.
But you know what: I actually agree with you!
My reference for testing the neutrality of amps is the «direct connection» of a high-impedance headphone (HD 650 or now HD 800) to the Bel Canto DAC2's line out. Its 20-Ω output and high current-supply ability are ideal preconditions. For attenuating the signal I use a 500-ohm potentiometer. It does have some minor impact on the sonic balance, but additionally, for verifying the effective cause of possible differences, I can also entirely renounce it and use a low-level music passage as test signal. It has turned out that the tiny sonic-balance shift is mostly negligible or then compensated by a pair of parallel resistors of 1 kΩ for halfways restoring the damping factor, thus the sonic balance.
With this reference characteristic at hand I try to detect the sonic difference between the direct connection and the detour via headphone amp. I could also use a second headphone amp serving as «source» to make sure that the headphone isn't underpowered or otherwise compromized by the «misuse» of the line outs, but the results are effectively the same, and the cleaner and more detailed line-out signal has turned out to serve my needs better. I also tested the impact of the low load impedance represented by a 300-ohm headphone by means of a sound check via monitor headphone. The sound was stable, just slightly reduced in volume due to the voltage-dividing function.
With this test configuration I can passably test the sonic flaws of the amps. Of the about two dozen amps that have passed this test the Corda Symphony is clearly the most neutral and accurate, followed by the Opera. That doesn't mean the Symphony is 100% neutral, but it is the first amp that offers equal detail, resolution and transparency as the direct connection. However, earlier tests with less «modern» amps than these two have placed the Earmax Pro quite high on the neutrality list, despite its reputation of a very tubey sounding tube amp. It may have helped that I preferred modern, less musical and euphonic current-production tubes with it (such as Sovtek and Electro-Harmonix).
These experiences tell me that it's not so simple to fix sonic traits on harmonic distortion. On the other hand I don't believe in magic in the context of audio phenomena, so the differences are certainly measurable. The question is if the measuring data are really adequately interpreted by the conservative approach. Sure, it's no news that 2nd order harmonics or even-order harmonic in general are less obvious or do less harm to the sound. But what about the extremely low measurable distortion of the latest amp designs with nevertheless existing sonic differences? I don't mention other criteria such as frequency response or noise, because they are no real issue and can be neglected (at least in most cases).
As you have mentioned yourself, amps can react differently to complex loads such as headphones. I don't really deny it, but for the sake of my argumentation stategy also coudn't really agree. Because with classic measuring methods you won't detect the differences. I'm not speaking of damping factors, which could easily be measured by means of output impedance (curves), together with the possible sonic-balance shifts produced by them, defined by the load impedance curve. I'm rather speaking of their reactions to complex, dynamic signals. The reported impacts of better (more stable, quicker...) power supplies must have their equivalent in the measurings. So it's clear that static or regular signals such as sine waves or pulses aren't the be-all end-all when it comes to grasp the entirety of audio phenomena.
In an earlier post I have tried to present a new
approach from
«Stereoplay»'s test engineers with the goal to bring perception and measuring data in accordance. I like this approach, and if it turns out to be universally valid, it would be a great tool for detecting previously unnoted flaws in audio electronics. Moreover it would open new horizons for the cable-sound proponents.
Again some notes to the different magnitudes of measuring and perceived sonic differences among amps and cables: In «controlled tests» they don't seem to play a role; both device categories seemingly don't alter the sound at all.
.
