RightMark Audio Analyzer
Subjective impressions should be verified by unbiased electronics. I will do that with the unique software suite called RightMark Audio Analyzer 6.2. It measures a number of standard parameters of the audio section by reproducing and recording a set of special signals. You can refer to the official user manual to find detailed info about the use of RMAA for testing purposes.
To measure the performance of the two outputs of the ASUS Xonar Essence STX I had to use 6.3-->3.5mm and RCA-->3.5mm adapters so that I could have the same equipment as in the previous reviews. To remind you, in order to test the line outputs I use a very short cable connecting the sound card’s input and output. The headphones amplifier is tested using a special splitter that is connected to both headphones and the line input of the Creative X-Fi Elite Pro sound card. For details about our methods of testing headphones amplifiers you can refer to the Scythe KamaBay Amp SDA-1000 review.
I want to say my huge thanks to the Xonar Essence STX developers for not saving on the analog-to-digital converter. As a result, this card’s line input is no inferior to those of the X-Fi Elite Pro and Xonar D2. The quality of the line input is highly important for tests: the result cannot be better than permitted by the recording device. And if you use the line input of another sound card installed into the same computer, you will surely get a worse signal-to-noise ratio.
Before discussing the test results I want to note that the Xonar Essence STX proved to be highly sensitive to the purity of power supply despite the additional power connector and numerous high-quality capacitors. In order to get acceptable results on the line output I had to roll the power cord into a small coil or there would be a lot of parasitic noise in the spectrums. The headphone output was less sensitive, its spectrums being clear all the time.
Line Out quality on Xonar Essence STX
The specifications say that the line input has a SNR of -118dB whereas the line output is twice better than that. Thus, it is the line input that must have limited the results in terms of noise and dynamic range. But what about the distortions? They do not meet the declared specs and are inferior to those of the Xonar D2 and X-Fi Elite Pro! I then performed a series of measurements of the headphone output and was perplexed even more:
Headphones Out quality on Xonar Essence STX
Paradoxically, the headphone output proved to be much better than the line output and by far exceeded the specified level of distortions. Thus, the high level of distortions in the line output is not due to a low quality of the line input. There must be some other reason. I suspect the feedback resistors because my replacing the opamps had no effect on the measured characteristics. The distortion spectrums agree with my supposition.
Headphones Out distortions
Line Out distortions
Here we can see that the right and left channels of the line output differ greatly in terms of harmonics. The headphones output shows the same, making me suspect a large variation in quality between the passive components in the I/U conversion stage, but there is also a considerable increase in the highest harmonics in the line output’s right channel. Therefore, that variation is not limited to a single defective part.
I guess the measurement results explain the drawbacks in the reproduction quality of the line output in comparison with the headphones output. Now, it is interesting to compare the parameters of the headphone amplifiers under load. Subjectively, the C.E.C. HD53R-80 was often better than the integrated amplifier. Will this be confirmed by the measurements?
The conditions and methods are the same as described in our first amplifier review but with one difference: the measurement results had only been correct at an output voltage of 1V due to an error in the normalization of the recorded signal. The problem was corrected in RMAA version 6.2.2 but the death of my Creative X-Fi Elite Pro prevented me from rerecording the results at a lower volume.
As I have written above, connecting one sound card’s output to another card’s input almost inevitably provokes more noise if the cards are installed in the same computer. That’s why the dynamic range of all the devices should be considered as similar. This does not apply to the other measured characteristics.
The harmonic and intermodulation distortions of the Essence STX’s integrated amplifier at the 32Ohm load are no match to specialized amplifiers although exceed the capabilities of sound cards’ line outputs. The distortions are especially poor at low frequencies, although the output resistance is comparable to that of the Creative X-Fi Elite Pro.
By lowering the volume by a couple of decibel the distortions can be reduced by 50%. When the volume is reduced by 9dB (not shown in the table), the headphones output of the Xonar Essence STX becomes as good as the line output of the Xonar D2 in terms of intermodulation distortions, but 5 times worse in terms of harmonic distortions.
TPA6120A2 distortions spectrum at 32Ohm load
Fortunately, it is the second and third harmonics that contribute the most to the total harmonic distortion whereas the fourth and higher harmonics fall below -100dB.
In fact, Xonar's integrated amplifier at low impedance loads can only be praised for good channel separation, which is not achived with line output of any sound card I tested. By the way, I want to remind you that the high channel crosstalk at measurements under load results from the Y-shaped splitter made from a thin screened cable. Therefore you should disregard the absolute values of the Stereo Crosstalk parameter.
The TPA6120A2 specification says that its intermodulation distortions get much lower at a load of 64 Ohms and higher. At a load of 600 Ohms the total harmonic distortion is not higher than 0.0003%, so I have no doubt the integrated amplifier is going to show its worth with high-impedance headphones.