First Impressions and Some Theory
I had some prejudices and even fears about the Auzen HomeTheater as I had used to have a headache from the hard sound of the Auzen Forte. Although I had eventually got accustomed to that “special effect”, I still had some unpleasant reminiscences. Music shouldn’t provoke a headache after all. To my greatest surprise, I did not have anything like that with the HomeTheater. Its sound was comfortable right from the start and the change of sound over time (the notorious warm-up effect) was far less conspicuous. After about a month of working at my computer daily, I noticed the sound to have become more subtle and detailed.
The second important difference of the HomeTheater from the Forte is that any opamps sound good with it. This had been the case with the Auzen X-Fi Prelude 7.1 and ASUS Xonar Essence STX but not with the Auzen X-Fi Forte 7.1 where the default LME49720 sounded bad in my tests but the replacement OPA2132P was even worse. The HomeTheater comes with the same opamp as the Forte, but sounds good with it and even better with other opamp models. This sound card lacks a well-defined voice of its own. That is, the resulting sound is largely determined by the opamp you install, therefore the Auzentech modules provided for the tests came in most handy. I spent a few months listening to diverse music with each of them and writing down my impressions. The continuing warming-up of the card was improving its sound, which made me correct my evaluations of many opamps over the course of the tests.
By the way, this sound card does not get as hot as the low-profile Forte. You cannot scorch your fingers on it. This must be due to the larger PCB whose metallization layers dissipate the heat from the power converter and headphone amplifier. On the other hand, this makes the other electronic components on the card hotter whereas heat is known to be the main limiting factor of the service life of electrolytic capacitors such as those you can see on the card. These are not the hardy solid-state capacitors but ordinary aluminum ones of the “civil” grade with a max case temperature of 85°C. That’s why I have doubts as to how long a Forte might last until the electrolyte begins to dry out in its capacitors. The HomeTheater is safer in this respect as you can easily touch even the hottest capacitors of its power converter without risking to get scorched.
There arose the same problem as with the Auzen X-Fi Forte 7.1 when I tried to measure the card’s characteristics objectively. There was a high level of noise when I used an external device with top-class ADCs. I guess it must be due to the power converter. And if I used the card’s own line input, the result was limited to the quality of the integrated WM8775 converter and no different from the result of the Forte.
Listening to those numerous opamps demanded a lot of time, even without accounting for the mentioned change in sound over time which introduced its inaccuracies. Comparing a pair of opamps takes a few hours because you cannot instantly replace them whereas the long-term sound memory of humans is not perfect. Besides listening to various genres of music, alternating the chips, I had to recheck my impressions because your subjective impression depends on your general state and mood and how tired your hearing is. So, I had to think of some trick to make the testing easier.
To remind you, my tests of the Auzen Forte 7.1 did not reveal any serious difference between the front output based on an AK4396 DAC and the rear/side outputs based on a CS4382A, despite the different opamps. The analog outputs of the Auzen HomeTheater HD are all serviced by one converter CS4382A and the difference boils down to the impossibility of easy replacement of the soldered-in NJM4580 opamps. If so, why not take the rear channels as the reference point?
In the Audio Creation mode the X-Fi chip allows sending audio to any of the four outputs or to all of them simultaneously. And switching the connecting cable takes a couple of seconds only. This made the testing easier and I compared the opamps again to correct my previous impressions. The NJM4580 turned out to be not as bad as its reputation. It yields correct imaging, i.e. the musical instruments are stationary in their locations throughout the width and depth of the sound scene and are perceived coherently with the reverberations produced. Not all of the tested opamps could boast that, particularly the default LME49720 in the front output.
But why, if the LME49720 is specified to be no different from the LM4562 which produced excellent imaging in the Auzen Prelude? To make things clear I took the opamp out of my Prelude and put it into the HomeTheater. Cross-checking the three chips (the third one was the NJM4580) left no doubt in me: the LME49720 was definitely different from the LM4562 although some similarity did exist. What does it mean then?
Same-type chips produced at different times and, probably, on different production facilities may vary in their audible characteristics. I guess this must be due to a difference in the input capacitance of the opamps. Take note that National Semiconductor does not even specify the input capacitance in its product documentation. My supposition is confirmed by the fact that the characteristic flattening of the sound scene could be observed with opamps that had field-effect input transistors. For such transistors, the nonlinearity of input capacitance is more important. Opamps with bipolar input transistors (AD826, NE5532, and NJM4580) were absolutely free from that problem. The distortions caused by the nonlinearity of the input capacitance are usually eliminated by equaling the input impedances, but are there any conditions for them to be unequal?