MEASUREMENTS: What does a US$500/watt stereo amplifier sound like? (Vitus Audio SS-010)

Behold!

That white item on my floor my friends is a Vitus Audio "Signature Series" SS-010 25Wpc solid state Class A into 8-ohms stereo integrated amplifier. It came out circa 2007 and had a suggested price of US$13,000 at the time (hence the ~$500/W stereo channel calculation in the title). Of course, the audiophile press had a crack at reviewing it. For example, here's the Stereo Times review from June 2007 and here's one from HiFi+ presumably from around then. I have seen no measurement results on this device and in fact the printed manual has this as a statement of their philosophy:

Vitus Audio is dedicated to musical performance rather than technical details and high performance measures. We believe that many manufacturers are blinded by the current available theory, instead of looking for new answers. At Vitus Audio everything works until proven not to. This way we always try to push the limits and explore the unexplored. As a result of this many of the solutions we use today are considered to be "impossible to ever get working", however they have convinced many people and reviewers world wide of the opposite. 

Before any of our new topologies are implemented in our products, they are carefully reviewed by our dedicated listening panel. This way we can guarantee the correct level of musically (sic) performance and robustness.

I added the bold highlighting as I thought this statement was rather surprising (worrisome?). Seemed a bit backwards from the orthodox concept of science driven by hypotheses that are supported by experimentation rather than just believing "everything works"! Suppose nobody bothered to test or measure... Does that mean it still "works" as a default position? By default, are theories then "true" until proven otherwise? Hmmmmm.... Let's just agree to disagree and move on.

At this price, it would be unusual if the reviewers did not like the device and the whole back-story around the company and designers. Interesting nuggets I got from reading the reviews are that internally this device is wired with VA's supposedly special "Andromeda" wiring (you can read more about this here) and the designer has a thing for beefy non-toroidal transformers. Thanks to my friend who was moving and needed a place to "store" the audio system, I borrowed this "handmade" amp for 2 months to try out. Let's just say my friend did not purchase at the manufacturer's suggested retail price.

Since 2007, Vitus Audio, the Danish company that makes this amp has upgraded and released a few newer models of course. This model according to HifiEngine was produced up to 2011. The Vitus Audio SS-010 will output 25W Class A into 8-ohms. There is also a Class AB setting which is also limited to 25W. The main difference is that in Class A it sucks ~250W and in AB ~50W (no surprise, Class A is inefficient). I didn't bother spending much time listening to Class AB mode which sounded fine. There's also the typical vague specs: SNR >110dB, THD+N <0.01% - presumable these are at rated power.

Vitus Audio SS-010 Rear panel. Two inputs - one set of RCA unbalanced, and one XLR balanced.

Perhaps most "impressive" is the fact that this baby is around 80lbs in weight. That's a heck of a lot of metal, much of it is the very solid enclosure and heat sinks (the dark stripes on each side - good dissipation of heat given Class A). It gets only slightly warm after a couple hours of use so it's clearly more than adequate to dissipate the heat provided that ambient temperature and airflow is adequate. Just as impressive is a hunk of metal called the remote control that can double as a police officer's baton in its ability to inflict physical damage. The remote is rechargeable.

When turned on it "thunks" and if you use the volume control, it also emits a short "thunk" with each step up (of which there are 47) probably due to the resistor relay system used. You can change the setting to be a power amp and defeat the volume control (it warns you before doing so and there's a pass code to do this - good safety feature). Unfortunately, I noticed the amplifier does produce a mild but audible transformer hum when turned on. Not loud enough to distract when playing music but audible from my listening position about 10' away on those quiet nights in the sound room.

As you can imagine, I listened to a number of albums on my system with this amplifier in place. Here's the equipment chain for reference:

Windows Server 2012 R2 computer --> gigabit ethernet system --> Logitech Transporter --> balanced XLR --> Emotiva XSP-1 preamp --> balanced XLR --> amplifier (my Emotiva XPA-1L monoblocks or this Vitus Audio as power amp) --> 4' Canare 4S11 speaker cable --> Paradigm Reference Signature S8v3 speakers

The amplifier comes with its own generic-looking black 16 AWG IEC power cable (I guess VA sees this as adequate for a very expensive amp). All bass management turned off from the preamp. Subwoofer disabled.

The AV system circa late-November 2016: Vizio P75-C1 up on the wall now :-).

Remember that using a 25Wpc amplifier means you're going to need efficient speakers! My Paradigm S8's being rated around 92dB/W isn't the most sensitive by far but was able to play loud enough for my typical listening levels (note though that I'm talking about typical music reproduction like from CD's; highly dynamic audio like home theater soundtracks would not be this amp's forte). From the moment I turned the system on with this amplifier in place, it was obvious that this amplifier's gain is lower than my Emotiva XPA-1L Gen 1 monoblocks ($700 x 2 MSRP back in 2013) given the same pre-amp input (by about 8dB). I also noticed marginally more hiss from the speakers when using the Vitus Audio than the Emotiva and I wondered whether internally it's designed with "true" balanced topology despite using XLR inputs.

So, what does $500/W sound like?

Let's start with the easy part... Subjectively I had fun listening to the amplifier in my system. It looks great, the weight and construction is very "solid" and the control buttons are easy to use. The amber LED display is small, simple, but functional. It's good that the display can also be dimmed. There's a nice illuminated "VA" logo that can be switched on/off for taste.

I ran the amp through a number of albums I've been listening to lately for comparison with the Emotivas. As you can imagine, with significant difference in sensitivity, switching cables then trying to match volume means there was no easy way to A/B compare with good confidence. My impression when listening to Sting's 57th and 9th (2016, DR7 - I like the track "50,000") suggested that it has enough dynamic ability to handle typical modern compressed masterings at adequate playback levels through my system. My impression was that there was indeed a "smooth" quality to the sound which IMO is not necessarily a "good" thing... I guess I appreciate a more "clinical" sound which in my mind speaks to the ability for a sound system to resolve details present in the recording (the HiFi+ review admitted that this amp wasn't the highest resolution amp reviewed - I agree). Listening to Leonard Cohen's You Want It Darker (2016, DR13!) demonstrated very good channel balance with Cohen's voice presented solidly "front and center" with nice layering of the instrumentation and separation from the background singers. On a synth-driven album like the new Yello Toy (2016, DR11!), the "surround" effects were presented nicely with an impression of 360-degree envelopment sitting in the sweetspot (eg. have a listen to the track "Limbo").

Acoustic music like Lucie Horsch's Vivaldi: Recorder Concertos (2016, DR12, Decca) likewise was presented nicely. However, with the average lower volume of classical recordings compared to rock/pop, I did wonder whether there was a bit more strain in reproducing the orchestral dynamic peaks. Also, as suggested above with the increased background hiss, the noise floor wasn't as good compared to the Emotivas. I'm sure many "serious" audiophiles are interested in playing classical music with elite amplifiers like this, so honestly, my impression suggests these should be paired with very sensitive speakers (like 100+dB/W horns and the like) in order to allow full dynamic expression and not to push the power too far to the limits. The noise floor remains an issue however.

What about objective measures?

Pulling out my 8-ohm power resistor rated to 300W then, let's run a few measurements through the dummy load and have a look at how these amplifiers perform... First, I wanted to have a look at the noise floor during silence (the Transporter was streaming a track full of digital "0"s):

As you can see, the Vitus Audio is indeed a bit noisier than the Emotiva overall. There's also an unfortunate noise peak out near 50kHz with the Vitus Audio amp.

How about the overall frequency response using a sweep from 20Hz to 48kHz (24/96 test signal)?

Interesting and impressive flat frequency response for the Vitus Audio amp - only -0.5dB way up at 40kHz. You see the Emotiva drops down to -0.5dB at both extremes of human audibility at 20Hz and 20kHz. Realize of course that in the big scheme of things, this isn't a problem because speaker-room interactions will have a huge impact on the sound in ways much more profound... Check this out:

Previously, I have written articles on room measurements and correction filters and applying these filters to Logitech Media Server. What you see in the light blue tracing above is the frequency response the amplifier has to produce to "equalize" for the room at the listening sweet spot (as processed with BrutefirDRC/LMS on the server computer and sent to my Logitech Transporter)! Remember folks just how much "imperfection" there is in the frequency domain when that sound emanates from the speakers and bounces around the room by the time it reaches the listener. Needless to say, this is hopefully a poignant and concrete reminder of the importance of speakers and the room you listen in, plus how you manage the room acoustics...

As for basic distortion measurements, let's start with having a look at a 1kHz 0dBFS 24/96 signal passed through an 8-ohm power resistor at 1W:

As you can see, I've overlaid the Emotiva XPA-1L's graph over the Vitus Audio and slightly offset it to the right so we can see the relative levels of the harmonics. It's obvious again that the Emotiva is quieter and the harmonic distortion is significantly lower as per the THD measurements.

Now, let's give it a bit more juice and look at 5W:

Notice that in both instances, the odd order harmonics (3kHz, 5kHz, 7kHz, 9kHz) are more prominent than the even order harmonics. This is the typical pattern with solid state amplifiers as opposed to tube amps where even order harmonics tend to be more prominent and said to be a type of "euphonic" distortion. Since I am one who prefers to achieve "high fidelity" rather than subjective "euphonia", personally I'm happier when distortions are lower or absent; in this regard, obviously the Emotiva is much cleaner...

Since the graphs above were all done with the right channel, I wanted to make sure the left channel for the Vitus Audio and my left Emotiva monoblock wasn't much different:

I plotted the graph above in log scale as well to have a look at the lower frequencies. This graph is a direct overlay without any shifting. As with the right channel comparison graphs, the Emotiva clearly achieved a lower noise floor overall with less harmonic distortion. The ~51kHz noise is still present through the Vitus Audio amp's left channel.

Here's another comparison but using a more complex 19 & 20kHz intermodulation signal:

Again, the previous observations and relationships remain. On the whole, the Emotiva amplifier has a lower noise floor. Harmonics and intermodulation products are not as pronounced as the Vitus Audio. The ~51kHz noise obviously still there with the Vitus Audio (circled) and easily examined when plotted with linear frequency scale.

Finally, for completeness, since Vitus Audio makes a big deal about Class A sounding much better than Class AB, I thought it would be interesting to run it in AB mode and see what the noise floor and distortion results look like at 5W:

IMO insignificant difference. Marginally higher value on the THD in Class AB mode but nothing to be excited or concerned about. Baseline noise level and general harmonic structure with higher odd-order harmonics also retained as expected. As I noted above, I didn't really spend much time listening to this amp in AB mode. I did play a couple of songs and didn't think there was much if any difference. We don't know how much bias is used in the amplifier design though. If it uses a high bias, we could still be operating in Class A mode at 5W I suppose. Although not shown in the spectrum from 20Hz-20kHz above, the Class AB FFT still had that 51kHz ultrasonic noise present.

Note that I did not go higher than just 5W in these comparisons. I really did not feel like pushing the borrowed amplifier and I also wanted to keep it simple when analyzing the FFT using my Focusrite Forte ADC. Even though I'm not pushing the amps to the limit, I do believe it's really those first few watts that's important when it comes to sound quality. At 1 and 5 watts, we can at least compare apples-to-apples measurements at typical power output when playing music. Refer to sites like Audioholics for full bandwidth tests, dynamic bursts (the Emotiva monoblocks are said to have 90,000uF storage capacitance - not sure about the Vitus Audio), damping factor, etc... Also, worth watching the video here for discussion of how much power you might need. Finally, I do recommend reading this article from them about the sound of amplifiers; many good points and thoughtful discussion items to consider. Remember that although the discussion can be complex, when it comes to "real world" listening, just be mindful that we don't want to overly trivialize nor magnify the sonic differences that might exist between reasonably spec'ed amps (as in general the audiophile press IMO are apt to do).

By the way, out of interest, I used a couple Audio-Technica AT2035 microphones to record the sound in my room playing segments from Eiji Oue & Minnesota SO's Rachmaninoff: Symphonic Dances... (Reference Recordings, 24/96) and Eva Cassidy's Songbird ("Fields of Gold") using the Vitus Audio and the Emotivas playing at the same amplitude level. Despite the relatively high noise (not the best microphones), the recordings did sound slightly different between the two amps when flipping back and forth using foobar's ABX comparator. A bit more "warmth" with the Vitus Audio, more "clear" with the Emotiva... Far from a definitive test, but interesting data point for me since there's no way to easily A/B compare directly without fumbling with switching cables and volume readjustments.

In conclusion...
So what does this $500/W amplifier sound like? Well, not bad of course (it would take gross incompetence for an amplifier >$500 to sound bad these days!). But from listening and in the objective measurements, I can say definitively that the Vitus Audio's noise level is not as good as the much less expensive Emotiva monoblocks which MSRP at 1/9 the price. And at the measured conditions of 1W and 5W into a dummy 8-ohm load, the harmonic distortion is likewise higher with the Vitus Audio.

For me, they both provide enjoyable listening experiences but obviously for higher volumes and "aggressive" music, hitting those peaks was effortless with the Emotivas and my speakers. Likewise, on extremely dynamic material typically with lower average volume (especially movie soundtracks), the two Emotiva monoblocks capable of packing 250W into 8-ohms each really fill the room with clean sound (especially bass frequencies with the subwoofer turned off).

The real question is the usual one when it comes to audiophilia and price. What is the value of a "handmade", heavy, solid metal, boutique device of this type to you? Compared to the Emotiva monoblocks, to imply higher value to a device with 1/10th the output power and 9 times the price makes no sense from the perspective of "sound quality" as defined by distortionless high power output and lower noise floor based on objective measurements (and just putting one's ears to the speakers).

As I see it, a luxury item like an amplifier with an audacious asking price >$10,000 is about those "non-utilitarian functions". Like the aspiration to achieve "pride of ownership" (or more bluntly "bragging rights") when it comes to a luxury car, fancy watch, expensive handbag, designer coat, etc... It's important to realize that these luxury items might not run faster, tell time any better, hold more things, keep you warmer, or in this case amplify an electrical signal cleaner and with less distortion. How these things are "valued" is based on socioeconomic status and psychological factors, not the underlying "utilitarian" quality (ie. "better sound") - no matter how some folks seem to confound or miss this rather obvious fact.

PS: I wasn't initially going to bring this up, but it just kept nagging at me as I was writing up this post. Have a look at this "Pictures" page from the printed user manual:

Notice where the images come from. Still copyrighted HiFi+ magazine pictures! I'm not saying there's anything terribly wrong with this, but I am concerned about the obviously poor "optics". Over the years, I have expressed concerns about the mainstream audiophile press and the intertwined relationship with the Industry. Already I think for many of us, glossy magazines like HiFi+ appear to be nothing more than the advertising arm of the Industry. What then should the consumer think about a situation like this where HiFi+ not only writes a long review on this product, the review is linked to on the Vitus Audio website as advertising testimony (see review link above), and the company even uses the magazine's pictures in the product's manual sent to each owner. So where are the boundaries? What does journalistic independence (and perhaps integrity) look like in this hobby? When a magazine prominently provides images in the user manual, does that already imply that the magazine endorses the product before consumers get their hands on one?

---------------------------

Wow! December is here already? Where did this year go?! It really is amazing as I get older how quickly time just flies by. Looking around this past week, it doesn't look like much is happening in the audio world. As you can see in the picture above of my sound room, I did get that 75" Vizio P75-C1 4K TV and have it all set-up in my media room.

Early impression - impressive :-). In a dark room, black is black with the Full Array Local Dimming (FALD) lighting system. Gamma looks nice (as compared to the Sharp a few weeks ago). Good to have both HDR10 and Dolby Vision HDR standards supported although Dolby Vision material remains lacking (I hope they'll start embedding DV metadata in UHD Blu-Rays at some point, DV streaming on Netflix looks great despite limited bitrate). I'm running the beta firmware (3.0.6.4) at this time which works well for HDR10 playback off an inexpensive Amlogic S905X-based TV box using a custom LibreELEC/Kodi OS build. I figure I'll wait until the next formal firmware release before I spend time with more detailed calibration. The 10-bit panel looks great for gradients when displaying high bitdepth HEVC material. If Vizio continues to release high quality but reasonably priced TVs like this, they'll be like the Oppo of the video world. I guess Christmas has come early in my household!

I remain impressed by the real advancements happening in the world of video display. It has been a good year I think for the proliferation of 4K and HDR technology primarily... This is as opposed to stereo audio technology which clearly has matured for many years now and what we're IMO seeing is primarily maintenance of high quality at lower price points (like the Raspberry Pi streamers) and what I believe are needless "pseudo-solutions" like MQA with little potential.

Have a wonderful week ahead folks as we run into the final stretch of 2016. Hope you're all enjoying the music (and video!)...