Originally Posted by ramicio
Not really rolloff, just plain sucky, weak, ill-defined bass. The amp just isn't powerful enough.
It wouldn't surprise me if the same thing applied to all of these other puny little pathetic tiny amps like the O2 and other battery or wall-wart-powered crap. Frequency response charts don't mean anything, especially when they're just measuring a simple tone played through the headphone.
So what do you
mean by "powerful enough" here? I'm not particularly sure if I see a relationship with amplifier size, but that depends on what you mean. Certainly for higher output power, you may need some larger components, heatsinks, but not as huge as some people make amps unless you insist on the more inefficient class A designs. There are a lot of factors involved with regards to size. This below:
Originally Posted by ramicio
Yes. It takes power to move a mechanical device large distances and change direction in a clearly-defined manner. Why do you think subs in cars are rarely ever just a few watts?
implies electric power to me though, watts. So are you talking watts or not?
For that matter, which headphones are we talking about? As we know, Essence ST / STX damping factor is not ideal for some low-impedance headphones; granted, plenty of people say they're just fine with ~22 ohms Denons and other more resistive models on those sound cards.
I wouldn't be so dismissive of frequency response information. These amplifiers are mostly linear in our operating range of interest (input signal bounded and not containing information above the audio band, output within a certain level, and the load being in a certain range of impedance). Hopefully it's all clear from linear systems theory how the frequency response (equivalently, impulse response) can describe the output behavior in such a system. Nonlinearities are small, as evidenced by the THD / IMD figures, so deviations from theory should likewise be small, as in any number of other systems out in the real world that are similarly modeled.
If you are going to make an argument based on controlling the driver in motion, back emf, etc., then sure. I wouldn't dismiss all of those out of hand either. That said, rather than accepting some hand waving, I would recommend this test, using music where you are hearing the "sucky, weak, ill-defined bass", if you want to investigate what's going on or are feeling generous and want to help us out:test (Click to show)
- Play the music out of STX -> record with STX.
- Play the music out of STX -> y-splitter to headphones and recording with STX.
- Play the music out of STX -> amp that doesn't have "sucky, weak, ill-defined bass" -> record with STX.
- Play the music out of STX -> amp that doesn't have "sucky, weak, ill-defined bass" -> y-splitter from amp output to headphones and recording with STX.
Grounding issues, sound card measurement caveats, and D/A and A/D quality aside, that should expose any weakness of the sound card not being able to handle the signal while driving the headphones. If there is enough of a difference to make the bass sound off, it should show up in the recording, right? The recording is pretty much the voltage seen by the headphone drivers. There's not another input into the system that should cause differences in sound quality, except stupid negligible stuff like slight temperature variations, slight "burn-in" effects from one test run to the next. If the voltage is significantly different between test runs, then we can analyze what's different and maybe figure out why.
P.S. so I'm not ragging on anyone in particular. If you don't mean bass rolloff as in the frequency response magnitude is less at lower frequencies, calling it rolloff is pretty misleading too, IMHO... If I knew what you meant, then some of these posts could be avoided. People talk about bass rolloff of speakers or headphones and mean that the FR is sagging at the low end. Or a DC blocking capacitor at the output is causing bass rolloff in terms of FR. Again, I thought the meaning was clearly defined.Edited by mikeaj - 11/13/12 at 4:11pm