As has been pointed out before, Sennheiser appears to have designed their HDVD800 to optimized the presentation of their flagship headphones the HD800. I thought it might be interesting to some folks as to how some of this was done. Note that the numbers in the following presentation are not intended to be super accurate, but it should be enough to get the idea across.

I went to Innerfidelity's site to get a copy of their HD800 Impedance curve. I then roughly estimated the values so that I could get it into Excel. The result was the following:

As you can see the impedance is not constant from 10Hz to 20KHz varying anywhere from 350 to 650 ohms. Normally, this should not matter if the output impedance of the amplifier driving them is very low (e.g., < 1 ohm) since the full value of a given voltage at any frequency will show up directly on the headphones. E.g. a 2 volt signal at 90 Hz will show up as 2 volts on the headphones and a 2 volt signal at 6000 Hz will also show up as 2 volts regardless of the fact that the impedance of the phones at one point is double that at the other. This can be seen in the following chart showing the amount of amplifier signal voltage showing up on the headphones when the amplifier's output impedance is around 0 ohms.

Although amplifier output impedance is a bit more complicated, for this discussion it can be viewed as a simple resistance in series with the headphone's impedance. As the amplifier's output impedance is increased through design, less of the signal generated in the amplifier gets to the headphones. The real issue here is that the amount of the signal's voltage that DOES get to the headphones is dependent on both the frequency of the signal as well as the frequency dependent impedance of the headphones and the output impedance of the amp.

The following chart was created based on an amplifier's output impedance of 47 ohms (like the HDVD800) feeding a HD800 headphone. The vertical axis is the percentage of the amplifier's internally generated signal voltage that actually makes it to the headphones.

Now in the upper Bass/lower midrange nearly 94% of the amp's signal voltage is still getting to the headphones. However, at the same time though, in the upper mid range/lower treble only 87% of the signal is making it to the headphones. Also note that this 7% difference is relative to the internal amplifier's signal. Since the real difference is based on signal levels at the headphones, the actual difference being heard on the headphones is closer to 9%.

So by creating a headphone amp that has an output impedance of 47 ohms, Sennheiser has reduced the relative level of the upper midrange/lower treble signal to be much lower than it would normal be by itself for a set of HD800 on a different amp (such as a Burson, e.g.). And isn't it this range of frequencies on the HD800 that many folks considered "bright"? Sennheiser has basically "turned down the treble" without affecting phase.

The reason that this was rather interesting to me is that the T1 that I have also has a similarly shaped impedance curve (only twice as high at nominal 600 ohms). This higher output impedance on the HDVD800 also helps to tame the brightness issue of the T1 in the same way as the HD800 (although the impact probably isn't as significant since the T1 are twice as high an impedance as the HD800 are so the percentage differences are lower).

Just for interest, I also plotted what the relative signal levels would be if the amp had an output impedance of 300 ohms (i.e., similar to the headphone impedance). Chart follows:

As you can see, the higher the output impedance is, the more extreme the variations become. In this case the output impedance is about the same as the impedance of the headphones so only about 50% or so of the amp's signal is actually getting to the phones. But even worse, the variations of signal level now approach 26% across the headphones themselves totally warping the frequency response of the phones. This is the type of effect that can be had by putting low impedance headphones on the HDVD800. the last chart above shows an example of a headphone being close to the same value of an amp's impedance, although in reality, low impedance headphones tend to have less variation in their impedance curves than the higher impedance phones typically do.

Obviously, if the impedance curve of your phones was fairly flat, their frequency response would not vary much with different amp impedances (other items like damping factor could though).

And one last item. These examples were based on the assumption that the output impedance of the HDVD800 of 47 ohms is consistent. I.e., if it is only nominal and varies with frequency, obviously things can get far more complicated.

Anyway, hope this tidbit is of interest to some.

Edited by wisemanja - 7/22/13 at 8:13pm