Saturday, 12 June 2021

REVIEW / MEASUREMENTS: Pass Labs / First Watt SIT-2 stereo Class A amplifier. [And Stereophile steps in with MQA yet again...]


Greetings all. For this post, as you can see above, I've got something exotic to review / measure here at chez Archimago. It's a Pass Labs / First Watt SIT-2 stereo amplifier which was made in limited quantities back around 2011-2013.

This amplifier comes with a nice backstory; that of the silicon carbide (SiC) "Static Induction Transistor" which you can read about here (aka V-FET). The thought is that these transistors behave in a "triode-like" manner, hence perhaps they will be able to convey a "tube-like" sound. Furthermore, this device follows the general audiophile philosophy of "simplicity and minimalism" (quoted from the manual) as principles that would result in a device sounding "good". As such, this amplifier has only a single SIT gain stage, operates in Class A, built with minimal components, employs no feedback, and as expected, accepts single-ended input only.

Online I found this user manual and Japanese product flier for those wanting a bit more information from the company. The cost of this amplifier when it first came out was around US$5000.

Externally, it has the typical First Watt amplifier look with large heat sink fins on the sides:

As a Class A device, I noted that this draws about 135-140W constantly when plugged in using the Kill-A-Watt meter with little to no variation during playback. This is lower than the 200W reported in some of the literature I've seen. Assuming the Kill-A-Watt result is correct, I wonder for products like this whether subtle design changes are implemented during production which could change performance without notice.

The box gets pretty hot after an hour. I measured it at around 40°C in ambient 20°C by 2 hours of use in my basement in the evening. It's uncomfortably warm to touch but not going to burn fingers so the heat sinks are doing a good job.

Out back, we see that this particular unit has serial number 17XX (anonymized to protect the innocent of course). I wonder how many in total were made.

RCA pair for single-ended input. A selector for "10K" vs. "100K" input impedance to match with your source (interesting use of XLR connectors as jumper settings). For most of my testing here, I had it set at 10kΩ since that is the most "direct" setting which presumably results in best performance. The 100kΩ choice adds extra buffer circuitry that will improve compatibility if you're using high output impedance source devices (those exceeding 1kΩ, like some tube preamps specifically). I will check to see if I can detect a difference between the two settings later.

Two pairs of multi-way speaker binding posts that accept banana plugs, IEC power connector, power switch, and fuse round out the rear complement.

Physically, this box measures 17" wide x 14.5" deep x 5.2" tall not including the feet or slight protrusion from the rear connectors. I measured the weight as 27.5lbs (12.5kg). Overall, it's a pretty large, hefty box which I think does look like a "serious" audiophile component without any glowing-tubes level of bling. The construction is pristine, and I appreciated the rounded edges which added to the impression of excellent workmanship when lifted and moved.

I. Measurements

Alright guys and gals, let's get objective.

Here she is on my test bench warming up and getting ready to go. Notice the two bluish power LEDs which are tastefully bright without being blinding. As usual, I'll measure a device like this after some warm-up; in this case, I gave it about an hour to get to 35-40°C on my temperature probe. Since I'm borrowing this from a friend, I'm going to be relative gentle in not pushing it beyond reasonable limits for the tests.

As usual, I will use the RME ADI-2 Pro FS as ADC and Linear Audio AutoRanger in the measurement chain. For anyone who cares still, I used a thick "audiophile" power cable as discussed here. To optimize noise level, I will run the ADC and AutoRanger devices off lithium batteries. Signal source used is the Topping DX3 Pro V2 with excellent high-resolution output.

First up we have the voltage gain. First Watt says that the gain is +18dB. Here are my measured values for this unit:
        Left channel gain: +15.9dB
        Right channel gain: +16.0dB

Not a high value compared to most other amps which typically start >20dB. In fact, this is almost exactly the same as the Pass Amp Camp Amp measured around this time last year.

Let's have a peek now at the damping factor across the audible frequencies:

As you can see, I measured it across a 4Ω load. Note how low this is! Good that it's flat. On average, we're looking at just below 1.6x damping which means the output impedance is around 2.7Ω. Obviously be mindful of this and try to pair it with appropriately high impedance speakers for flatter frequency response. First Watt says it's compatible with 4-16Ω; I'd advise >8Ω if you can.

Using my standard power resistor loads and the old Sony SS-H1600 bookshelf speakers which I have used with other amplifier measurements, let's have a peek at the frequency response with output level from the amp at 1V into each of the loads:

We see a nice and flat frequency response into the resistive loads with around -0.5dB at 20Hz and -0.2dB at 20kHz. The low damping factor allows the Sony speaker to bounce around quite a bit from +2.0 to -1.5dB across the audible spectrum reflective of its impedance variation (the Sony is a 6-8Ω speaker load).

My friend asked me to check if the amp is able to maintain "absolute polarity". There are many ways to check but here's one way where I just captured an impulse response from a Topping DX3 Pro V2 DAC --> SIT-2 amp --> RME ADI-2 Pro FS:

Since I know from previous measurements that both the Topping DAC and RME ADC (set to linear phase filter) maintain absolute polarity, we can see that the impulse has not been inverted as it goes through the SIT-2 amplifier. The marked +/- inputs and speaker outputs on the amplifier are therefore "non-inverting".

Using that 1V sweep into 4Ω, we have our first look into the amplifier's distortion (THD) characteristics across the frequency spectrum:

With the cursor over 1kHz, we can appreciate perhaps the most important thing to know about the nature of this amplifier. It's a low-order harmonic distortion generator. ;-) Basically, we see very high amounts of 2nd and 3rd order harmonics even just at 1V into 4Ω. That's already 0.5% THD at only 0.25W!

Interestingly the harmonic distortion is flat throughout the frequency spectrum and we see that the more audibly objectionable high-order distortions are actually very low. In fact, 4th order is down at 0.002% and harmonics above that are down at 0.001% (at 1kHz).

Let's dig down into the actual FFTs at various power levels into the 4Ω load; 1kHz fundamental:

Notice the consistent "cascade" pattern with 2nd > 3rd > higher harmonics across these snapshots from 2.5mW to 2.25W into 4Ω.

We can see the same pattern into 8Ω as well:

Due to the high 2nd and 3rd harmonics, clearly this amplifier is not going to look very good on distortion-power curves which do not differentiate between the relative harmonic amounts:

The graph above shows higher distortion than what First Watt published on their website; again, I wonder if the design and performance characteristics of these products change over time unannounced and what differences there are between development prototypes and production units. I'm seeing 1% THD+N at just over 1W into 8Ω and around 0.6W into 4Ω. Since I'm borrowing this amplifier, I didn't push the output too far towards the 10W spec since distortion is getting quite high already where I stopped measuring (4.5W into 8Ω with 3.1% THD+N, and 6.25W into 4Ω with around 10% THD+N).

The official specs from Pass rates this amplifier as 5% THD at 10W into 8Ω, 1kHz. Sure, it could be something like that.

I was quite pleasantly surprised when I looked at crosstalk (0.5V into 4Ω) however:

This is my simple crosstalk measure where I put a 0dBFS 4kHz sine in the left channel, and 300Hz in the right channel, looking for seepage of the signals into the other channel. We see an excellent crosstalk performance of >100dB between left and right sides. The two channels share the same power supply, but otherwise function quite independently. Again, notice the high amount of harmonics.

With high THD, we expect high intermodulation distortion as well (they're both reflective of amplifier nonlinearities). So let's run a few standard IMD tests into that 4Ω load. Output level at 2Vrms or around 1W which has been a "standard" in my tests:

As expected, the intermodulation distortions are high.

Here's the synthetic TIM ("Transient InterModulation") signal consisting of a 1kHz square wave with 12kHz sine created at 192kHz samplerate (96kHz bandwidth), measured again into 4Ω:

The TIM result is relatively high at -72.5dB, 2V output. As expected, those distortion sidebands around the 12kHz signal will drop, in this case down at -85dB when we pull the volume down to 0.5V output. This kind of result is similar to the Amp Camp Amp which achieved -70dB into 4Ω at 2V. While the Pass amps are measuring quite high on this "slew-rate induced distortion" test, don't think that this kind of anomaly only affects solid state amplifiers, the tube Melody Onix SP3 Mk II had a -80dB result. Contrast this with the Class D Hypex NC252MP with a result down at -120dB, basically in the noise floor on this same test.

As a single summary measurement of distortion I've been using to "grade" amplifiers, here's the Triple Tone TD+N captured at 2V into 4Ω; left and right channels:

Clearly, we see quite a bit of harmonic and intermodulation products showing up in the FFT. A Triple-Tone  (48Hz, 960Hz, 5472Hz) TD+N of -30.5dB (2V output into 4Ω, average of channels) is a rather low result in the context of most other amplifiers and about the same as the Amp Camp Amp which scored -30dB.                      

Notice that there's some "asymmetry" in the Triple-Tone FFT with the right channel showing more low-level distortion that the left which is reflected in the numbers.

Here's the square wave off the oscilloscope:

This is a signal played back from the Topping DX3 Pro DAC ("non-aliasing", 24/384). That small leading-edge overshoot is actually from the DAC itself and the amplifier reproduced this as fed. Channel balance is great and no complaints with overall morphology.

Since this amplifier is supposed to have a very wide bandwidth, linnrd was curious to see the square wave FFT. Here's the RME ADI-2 Pro FS running at 768kHz sample rate with a 1M-point FFT showing amplifier output all the way to 300+kHz.

While square waves only consist of odd harmonics, notice that the even harmonics are showing up here as a result of the high harmonic distortion. We see roll-off into the high frequencies which I believe is mainly attributable to the DAC/ADC system itself.

Let's make sure the ultrasonic spectrum doesn't contain any excess noise:

Looks clean as expected for a Class A device like this. I'm playing a mixed 5kHz and 93kHz signal, notice that both levels are similar in amplitude confirming the high frequency bandwidth of this amplifier. Although the FFT is low-resolution, we don't see any ultra-high-frequency content all the way out to 1.2MHz (unlike what we typically see in Class D amps with their ultrasonic switching noise).

Finally... As I mentioned in the description of the amp above, there is an impedance jumper which allows the input impedance to be switched between 10kΩ and 100kΩ. The Topping DX3 Pro V2 DAC which I used has an output impedance of 100Ω, thus compatible with either setting. Let's see if there's a measurable difference.

First, is there a measurable frequency response difference?

Nope, with a low impedance output source, basically no difference at all which is to be expected.

How about distortion?

Basically no difference either. That extra bit of circuitry to increase the input impedance to 100kΩ did not add any distortion anyone should be concerned about. Having said this, since we are using highly sensitive instruments (rather than human ears which can tolerate a heck of a lot more distortion differences anyway), if we're being absolutely (and insanely) obsessive, the 10kΩ setting gave us 0.405% THD versus the 100kΩ setting at 0.414% which is not significant in any way, but sure, by all means stick with the 10kΩ setting for the lower distortion number if you can. ;-)

We can also look at an FFT with different conditions to see if this same tendency holds - here's a 0.5V, 1kHz signal into 4Ω load:

Visually I can't see a real difference and although the numbers vary slightly in realtime, there seems to be a very small edge to the 10kΩ jumper setting that's something like 0.1dB better than the 100kΩ for THD(+N).

While I don't know if any subjective listener has ever claimed to hear a difference, I think it's neat to demonstrate that there's basically no problem either way for peace of mind. So for sure if you have a high output impedance pre-amp of >1kΩ, feel free to use that 100kΩ jumper with absolutely no need to be concerned that distortions have been added.

One last thing on the FFT graphs above. Notice that while the 60Hz North American mains hum is low, the 2nd and 4th harmonic "spurs" at 120Hz and 240Hz seems to be the strongest power-related noise poking up. Desirable or not, this amp likes to give us harmonics. ;-)

II. How does it sound?

So here's my system with the SIT-2 being used as the amplifier. As usual, I've got my Paradigm Signature S8 v.3 front speakers connected. Listening with all subs (Paradigm SUB1 and Polk PSW111) turned off and no DSP since I don't have time to dial in the settings for this amplifier.

Back in late 2019, I mentioned this amp in my friend linnrd's system, so a big thanks to him for letting me borrow this! I had the Pass SIT-2 for about a week to test and listen here at home. I listened to it on the 1st evening, spent 2 nights with testing, and listened to it again for another couple of nights before returning the little beast - I think this worked out as a nice balance between subjective and objective evaluation. As I've expressed before, I don't think there's a need for "long term listening" as the differences in sound, if they be significant, will be heard quite quickly... Over time, all that acclimatizes are the audiophile's ears rather than actually hearing anything new!

As you can see in the image above, I used my Roon-based server streaming to the Raspberry Pi 4 "Touch", DAC duties performed by the RME ADI-2 Pro FS Black Edition, feeding the low noise Emotiva XSP-1 preamp with balanced cables. The unbalanced output from the preamp then goes to the Pass SIT-2 and speaker wires are my DIY cables. A technical detail with the system is that the Emotiva XSP-1 preamp has a relatively high output impedance at ~950Ω so I listened with both the 10kΩ and 100kΩ jumper settings on different evenings and did not notice a difference.

First, let's just talk about the very obvious main amplifier limitation in a system like this. This amplifier does not deliver much power. The low voltage gain also means I had to push the preamp volume higher than normal, basically similar to the Pass Amp Camp Amp (ACA) last year. With room gain, my speakers are rated at about 92dB/W/m so with the claimed 10W of power, I was still able to push volumes to a pretty good level. Something I noticed was that there wasn't an obvious need for "warm up time" with the SIT-2 in order for the sound to stabilize compared to the ACA. I thought the sound was quite stable when turned on at least within 5 minutes. Also, compared to the ACA and the more recently measured vacuum-tube Melody Onix SP3 Mk II, the SIT-2 had a quieter background with less hum or hiss at normal playback volume. This is very important if paired with extremely high sensitivity speakers in low ambient noise listening rooms!

These are not "rock 'n roll" amps in a system like mine. Yeah, I can turn up and enjoy an album like AC/DC's The Razor's Edge (I used the more dynamic older DR11 mastering), but when the drums escalate and the masses of guitars and vocals congregate, it's not hard to notice the lack of dynamics, especially restrained bass impact. This was the same with that awesome flute (LeRoi Moore) & percussion (Carter Beauford) at the start of Dave Matthews Band's "Say Goodbye" (Crash - one of the best rock recordings of the '90s IMO) where the dynamics sound constrained and don't quite achieve the usual emotional impact in my room compared to the Hypex amp even with the subs turned off.

Having said this, with more modern loud pop where there's not too much dynamic variability or gradual "build up" on amplitude throughout the song, it sounded fine. For example, a popular album like Billie Eilish's When We All Fall Asleep, Where Do We Go?, well-known tracks like "Bad Guy" sounded crisp, delivered a wide good soundstage, and good details at normal listening levels. Music like "Whispers" on Deadmau5 At Play Vol.3 with numerous digital effects sounded nice also with the "surround sound" spatiality well preserved.

Within reasonable playback levels, bass delivery was OK and I didn't hear anything unusual or problematic. Rather, I did think that the mids and upper frequencies sounded very sweet. For example, the horns on "Tour De Force" (Dedicated to Diz, Telarc) were full and "present" with these amps more than I had noticed before. On some amplifiers, there can be a harshness on this track, not so on the SIT-2, very nice.

Subjectively, what I think many audiophiles love about the sound with amps like this is with vocals. Female vocals in particular like Melody Gardot, Rebecca Pidgeon, and of course the lovely Diana Krall sounded great ;-). What I found interesting was the subjective smoothness of the female voice. There's an "ease", "dimensionality", even "presence" for lack of more precise wording. Enunciations were clear, tonally sounding natural, with a good impression of detail. Jane Monheit's "presence" as she delivered the solo opening on "Over The Rainbow" (on Come Dream With Me) was simply palpable in the sound room plus I liked the nice shimmering of cymbals later in the song. By the way, Monheit's new album Come What May (2021, DR10) is quite enjoyable as well if you like this type of "typical" audiophile female vocal genre.

As I write this, I of course know what the measurements are showing. To me, this amplifier delivers the "euphonic effect" that I think many audiophiles like about minimalist, no feedback amps with high but pleasant harmonic distortions. The "euphonia" I suspect appears to be due to harmonics in the upper mids, "Presence" and "Brilliance" frequencies being "filled out" even though they were not in the original recording (ie. not quite "accurate" reproduction of the source material). Furthermore, the speaker's impedance curve will play a huge part in how well it "matches" to this amp due to the high amplifier output impedance.

I didn't think the effect was as noticeable with male vocals although it complemented music like "The Sound of Falling Snow" (雪落下的聲音) from the Story of Yanxi Palace (延禧攻略) soundtrack where the ethnic instrumentation really bloomed (here's a YouTube video of the song). Voices like Johnny Cash, Andrea Bocelli, and Philip Bailey came through with good clarity and fullness even if I didn't think the effect of this amp was as impressive as with the ladies.

III. In summary...

As I said in the opening paragraph, this is a rather exotic piece of audiophile gear. Arguable, this is about as exotic as it gets for solid state devices with a unique, rare transistor implemented in a distinctly "audiophile" minimalist manner in Class A, with no feedback.

As usual, here's the summary "AMOAR" scoreboard I've used over the years for amplifier reviews:

Perhaps this is what it looks like when "numbers don't matter" in amplifier design? ;-)

We're seeing very high output impedance (low damping) with resulting frequency response anomalies when presented with an actual speaker load. Significant amount of non-linear distortion is obviously causing that high -30dB Triple-Tone Distortion Factor score (I typically round the number down, it's around -30.5 average). The amount of power available is also limited and in fact I was unable to find any output level with  <0.1% THD+N from 0.1V up to at least 5V into 4Ω and 8Ω loads (it does get close to 0.1% THD+N down at 0.25V output into 8Ω, or around 10mW).

Despite power limitations and large amounts of distortion, at least the harmonic distortion is low-order and generally "benign" subjectively which could be perceived positively although some people may judge the effect as detrimental as well; such is the nature of human perceptual abilities and preferences. Notice the distortion level on the THD+N vs. voltage/power graph increases gradually instead of sudden severe clipping which probably also makes the sound less objectionable as volume is increased and during highly dynamic passages.

What this means is that it's up to the audiophile to make sure to match the performance of an amp like this with the speaker he/she is using. Obviously, it's best to pair this amplifier with higher impedance speakers. I would say aim for >8Ω to improve effective damping to smooth out the frequency response more. Look at your speaker's impedance curve to approximate how the frequency response may fluctuate.

While my measurements show that distortion with this particular unit is higher than the Pass website distortion vs. power graph, I trust that nobody's going into this purchase thinking they're buying a very low distortion "high fidelity" amplifier! This is not the kind of amplifier to show off your "hi-res audio" music to impress friends who are more interested in clean, accurate reproduction!

With the low power rating, high speaker sensitivity is another characteristic you need to look at. In this regard, linnrd's pairing of this amp with his Avantgarde UNO horn speakers with active bass unit and sensitivity of >100dB/W/m is wise.

A closer look at the Pass SIT-2 and the little DIY Hypex nCore on top.

If you have a chance, have a listen to the SIT-2 for an "audiophile experience" as another example of what high 2nd and 3rd harmonic distortion amplifiers sound like. As a solid state box (the Amp Camp Amp being the other solid state one I've examined with high distortion), this machine has the benefits of better reliability, won't wear out like tubes, and doesn't require much if any warm-up time.

Furthermore, this is an amplifier with an intriguing back story: Nelson Pass design, minimalistic philosophy, unique transistor technology (specifically Pass' limited-run SemiSouth Static Induction Transistors), and just plain rare. In this regard, scarcity also probably means that this amplifier will keep its intrinsic value as an "investment" whether monetarily or to satisfy one's psychological edification. And these qualities can be independently appreciated regardless of the sound it makes! Remember years ago we discussed the importance of "non-utilitarian" factors that affect desirability and price; terms like "high fidelity" proper, "high resolution", and "accuracy" are simply not what this is about whether some Golden Eared Audiophiles want to admit it or not!

I see that this amp has been discussed over the years such as in this thread and this one. Subjective reviews / opinions have been provided by Home Theater Review and Enjoy The Music when the amp first came out. More recently, around 2018, Pass / First Watt released a SIT-3 model which appears to be a bit more complex; less "minimalistic", but less distortion.

Thanks again linnrd for the opportunity to try this baby out - it was fun listening to this and putting it on the test bench to "calibrate" what I was hearing! Indeed, this is quite the "weird beast" as you described it to me once.


To end... Let's talk about MQA once again. Feel free to ignore this section if you're sick of MQA talk.

By now, in 2021, I trust there's no longer a need to say much more about the technical abilities (or just as relevantly, criticisms and inabilities) of MQA. Although P.T. Barnum supposedly offered the proverb "There's no such thing as bad publicity", in reality there are certainly instances where this is not true. Much of the publicity around MQA surely has not been good, especially recently with GoldenOne's YouTube videos on the CODEC. I suspect this has been the kind of "bad publicity" that MQA does not want, especially when viewed among audiophile circles.

This gets us to the recent Jim Austin editorial in Stereophile - "MQA again" which is targeted at the videos. It's fascinating that within this "perfectionist-audio pond", the audiophile press continues to step up to bat for a CODEC that is clearly not "perfectionist" in any way, shape, or form.

Austin continues to repeat the claims of "deblurring"; yet MQA has not explained how this is possible in all these years. There's even the obvious but still amazing admission by Austin that "In the interest of making the sound better, it alters the sound the mastering engineer and musicians signed off on." Isn't this seriously problematic? How do we know it's "better" when it alters the "signed off" sound? So what in the world does "Master Quality Authenticated" even mean if the DSP is allowed to modify the sound without the artist/production team's authentic input? MQA: aren't you a little concerned about what Austin just said there!?

Let's not get into the "seaweed" with the specific numbers and test signals GoldenOne sent in for encoding to Tidal as he has expressed his findings well already. Let's think for a moment why we use test signals. We do so because they help us determine the limits of something - whether it's some hardware or in this case the effect a CODEC/DSP has on the known signals sent into it. Since this is a "lossy" compression system, the MQA encoder has to decide what to keep and what to throw away; what is deemed "music" and what is not. Wouldn't it be good for audiophiles to understand exactly what these limits are?

Inherently as audiophiles who hang out in the (again, as Austin puts it) "perfectionist-audio pond", we'd like to achieve audio playback with no compromises. We want systems capable of "perfectly" reproducing potentially audible signals whether from real acoustic recordings or commonly synthetic sounds which have never existed before. Heck, some of us might even want to ensure that the system reproduces likely inaudible signals >20kHz which hi-res allows with super-tweeters and such! As hobbyists seeking this level of high performance, possibly spending lots of $$$ in the pursuit (just look at the price of some of the stuff Stereophile reviews), of course there will be many of us curious enough to explore the limits of this CODEC and how it would influence the sound quality.

Did MQA honestly think that by strategically employing magazines like Stereophile and The Absolute Sound in its PR campaign, audiophiles would simply fall in line and be satisfied with what these people reported? I think MQA has grossly underestimated the independence of audiophile hobbyists and magazines have lost touch of their actual ability to influence audiophiles over these years.

This article to me seems like a desperate attempt by Austin/Stereophile to act as some kind of "influencer" to change the MQA dialogue which has gone grossly awry against the company. I think all this article has done is perpetuated the "bad publicity" against MQA and in the process continues to demonstrate the bias that the magazine has against the interests of the community versus the desires of MQA Ltd. to achieve some level of success (by taxing music lovers for their needless "service").

As I said in one of the comments below the Jim Austin article, I believe that MQA barely was of benefit back in 2014-2015 when it was first introduced. In 2021, its proprietary scheme - one shrouded in nebulous claims by a company that actively enforces their opaque practices - simply has no value for the perfectionist-audio hobbyist nor music lovers in general IMO.

I'm certainly listening if anyone can provide even a single benefit MQA provides that cannot be done just as well if not better than existing, free, solutions of plain MQA-less FLAC/ALAC which have been embraced already by the other lossless music streaming services (the exception being Tidal of course, and we're still waiting on Spotify HiFi's exact technical plans).

While Austin ended off the editorial characterizing GoldenOne's tests as a "lost opportunity", I think the real opportunity that has been lost over the years has been to show that the magazine is able to perform a little bit of independent investigative journalism. To find courage to go beyond what they have been fed by the MQA company. For example, John Atkinson was able to get some of his music encoded years ago. Why not insist on trying some test signals then to understand the limits or claimed benefits of the encoding/decoding processes? I'm sure audiophiles would have been fascinated by what "deblurring" meant and how it could be demonstrated. Stereophile would truly have been the magazine for both subjective and objective evaluation of audio technologies if it had the balls to dive deeper and call out the BS (pun intended) if there's nothing to find before hobbyists had to take it upon themselves to investigate.

It should be obvious to everyone that if Stereophile is supposed to function as a reliable journalistic enterprise, then it must be truly impartial. Despite Jim Austin's words that "If I'm partial to anything, it's fairness", I'm not sure this is what we're seeing here. I'm struck by the echoing of MQA-inspired language used ("post-Shannon developments in sampling theory", eh?) and text devoid of independent thoughts or suggestions of "leg work" done by Austin to examine the veracity or significance of the disparate viewpoints other than basically agreeing with MQA against points made by GoldenOne. This article sounds more like the work of a tentative "lobbyist" and the magazine simply as a friendly advertising platform for the Industry. There's nothing wrong with this so long as the magazine is honest about it and the biases inherent in these roles openly disclosed.

I think it would have been better for Jim Austin to simply have remained silent rather than offering an article like this. Surely MQA Ltd. can handle themselves, and have offered their rebuttal. With any luck, let's hope this is the last Stereophile editorial we might see about MQA in awhile... Ideally forever...


While writing this SIT-2 amplifier article, I saw that Paul K (pkane) has released a VST plug-in called "PKHarmonic - Audiophile Harmonics Generator" where you can dial in the amount of harmonic distortion you want to add to your music playback! So for the Pass SIT-2, based on what I see in the measurements, try throwing in some -50dB 2nd, -70dB 3rd, and -100dB 4th harmonics to start (these numbers correlating with the distortion levels at 62.5mW into 4Ω or 125mW into 8Ω) and go up from there to see if you like the "tube-like" sound.

If playing with the software satisfies your curiosity and saves you a few bucks by not buying a new tube(-like) amp/preamp, make sure to donate to Paul's work!

Stay safe friends. Hope you're having a great June. A shout out to the audiophiles attending T.H.E. Show 2021 in Long Beach this weekend - have fun!

As always, I hope you're enjoying the music.

linnrd got confirmation from Nelson Pass that only 150 of these SIT-2 amplifiers were ever made (by hand mind you)!!! Well then, that's way rarer than I thought. I guess it'll be a bit more difficult to find one of these locally to listen to. ;-)


  1. Hopefully some day you could measure and review some low cost audio like the chinese Douk amp ( Got one at amazon returns site along with a tube pre-amp ( a sony blu-ray player and a set of Micca bookshelf speakers ( for a total cost of < $100 cdn. How does it sound? - very good as a near-field system connected to my iMac. Rather impressive, voices are correct - wide soundstage, good depth, and surprising bass from a 4.5" woofer. No bling, but it does sing...Would it replace my previous $10k system? No, but i'm enjoying listening to it. good work - keep it up!

    1. Hey gnickers, interesting suggestions. I'll look into some them certainly if I run in to the products.

      An amp like the Douk Audio G3 Pro should measure better than the old single-TPA3116 Yeeco which was measured here:

      Wow! Impressive that you got the whole package for <$100. I'll need to look for an outlet like that around here in Vancouver!

  2. I would love to hear one of Nelson Pass amps matched with appropriate speakers. Within the last few years I have found myself drawn to female vocalists and am very curious about the notion of Class A and the sound that I hear so much about.
    On the MQA and Stereophile again, i just see further proof that magazines/websites like Stereophile have become nothing more than trade journals, meant to put forth manufacturers new equipment for our consideration with little actual, objective investigation into the manufacturers' claims. Just reading the first paragraph with "MQA has once again floated to the surface of the perfectionist-audio pond—not belly-up as some have hoped but forced there by relentless pursuit by anti-MQA predators posing as impartial jellyfish.", was enough to let readers know that Jim Austin's remarks were going nothing more than the continued acceptance of MQA claims without any hint of doing any serious investigation of them.

    And to end with "GoldenSound's tests are a missed opportunity." is really quite ironic. The only missed opportunity I saw was once again MQA refusing to allow any type of objective testing of their claims, or simply answering any of the any questions about those claims. Maybe instead of chastising those asking legitimate questions, Stereophile should take thwe "opportunity to display some journalistic integrity, and ask those questions for their readers. Anyway, I'm thinking that first with Amazon HD, and now Spotify and Apple jumping into the lossless streaming pond, MQA days are likely numbered.

    1. Hi Joe,
      Yeah, agree, MQA's days are numbered and while we can't say at this point if this will be one of the last articles, I think we are closer to a eulogy with all the other streaming services moving on and just able to offer the listeners full lossless without (too many) games - even if at this time ensuring "bit-perfect" playback might still be an issue on devices.

      Currently I'm listening to Amazon Music HD and this is certainly more than good enough for me to sample music in excellent quality. For stuff I love, I'll still look around for the CD (most of the time) and rip into my music server for Roon replay. Buying the physical product and going to concerts (one day) are probably still the best way to support artists. Absolutely no need to care whether the DAC or player decodes MQA or just "renders" or other nonsense like that for the end user... Much less spend extra money for this hassle!

      Yeah, I got a kick of Austin's 1st sentence as well. LOL. Austin must be looking at this upside down. MQA is totally belly-up and the swimming bladder is dysfunctional at this point, maybe constipated or something but still alive although clearly struggling.

      I don't know if the audiophile magazines operate in any kind of cyclical fashion or if true "journalism" can ever come back. Maybe it's too late and nobody really cares; the only way magazines can get the $$$ they need to operate is from manufacturers and this will always apply a bias to the whole venture.

    2. Sir, you clearly have no understanding at all of M-QuAck. M-QuAck is among the very greatest of mankind's inventions. It makes something out of nothing; the dreams of the alchemists realized at last, and the unseen source of much boodle in Bob Stuart's retirement fund. Any fool can see the value. It is authenticated! No need to spend money on expensive DACs when any $200 M-QuAck certified DAC is authenticated and reproduces the original studio engineer's experience perfectly. You sir, are incapable of appreciating perfection, unlike the esteeemed Stereophile staff -- granted delusional, half-deaf or more, often drooling, ignorant and often imbeciic, as they are. You sir are an objectivist-skeptic; the very worst kind. M-QuAck shall conquer yet. Plenty of senile, delusional, and deaf golden-ears reviewers out there ;-) Warm regards Terry NYC

    3. Thank you Brother Terrance from NYC, mea culpa for I have sinned and now observeth the light. Thou hast admonished my wayward course and set mine destination right. As a result, I have summoned the fastest homing pigeon in the friar's loft to carry this important communiqué to the Most Highest MQA Council in Cambridgeshire - it reads thus:

      The Confessions of Archimago

      Forgive me, Father Robert, for my sins.

      I have cavorted with thine adversary, the Lossless Old Nick and shall forever be in thine debt. I know thou givest great providence and I have not appreciateth thy bountiful provisions of pure, authentic, unblurred, sonic orgasmic climax. I know thou hast boundless mercy and by approaching thine throne, thou can forgiveth thy humble vassal.

      I therefore meekly submit myself fully to thine calling of enriching thy coffers and the betterment of thy spawn. May MQA live forever, enshrined in the music of the Spheres. May thy spawn and stakeholders never suffer in the wretched poverty and toil of unceasing labor, nay ever be ensnared as we lesser beings within our mort coils.

      I am ready to receive thy medicinal penalty in the hopes of thy kind salvation and escape from the ultimate humiliation of excommunication - though I knowest that I am deserving of such a hideous fate.

      Send forth thy commands with haste. I shall be expecting no less than unceasing lashings of thy Virginal Jbaras, undertake the arduous sacraments of the Lindberg-Lyd, and a multitude of Hail Meridians.

      With patience and trembling, I shall await the unfolding of thy Wise and Perfect discernment...

      Thy humble bondservant,

    4. applause ! ; superb ! ;-) btw, on a serious note it is telling that co-inventor, Professor Peter Craven, a respected collegue of Michael Gerzon, rapidly distanced himself from M-QuAck once the flummery started. best, TerryNYC

    5. :-)

      I noticed that as well. Even though Craven's name is brought up once awhile usually by Stuart in his interviews, we don't hear anything direct from him these years...

  3. Another great effort on your part. There are benefits to the lock down, but enough is enough and we all need our lives back. My middle son, Jon, is taking me to a Cubs game in Chicago on the 21st. A costly day trip from Atlanta and back, but it will be fun as I have been a fan since 1952 and the days of losing.

    I feel like "the science" behind MQA and is a muddled mess of misinformation much like covid. Even the bright minds here in the CDC were of no help. I am surprised that some drank the "Kool aide" who are scientists, so I am glad to see you are still pushing the envelope of "enlightenment". It seems to have fallen into the realm of politics rather than science. Sad. Has the industry really lost its way as to what "improvement" really means?

    I will just take some well recorded 24/192 and be very happy. I love DSD and SACD as well and am over worrying about the ultra-sonic issues. I can't hear it and it is not bothering any of my playback gear.

    As a side note, I have read the late, great, Al Schmitt's book, "ON THE RECORD" twice now and found it interesting that he said that they now record at 24/192, but he loves tape and LPs, but knowing where he was so blessed to start the beginning of the music business with Mono, he has embraced the new technology, but uses mics as tone controls and does little EQ on the board. It is a great read. I also bought his DVD of his recording Master Class and how he does a big band recording. Super fun to watch and listen.

    Stay safe.

    1. Thanks for the note Jim,
      Great ideas there! I will need to have a look at Schmitt's material at some point since I really like those technical and practical pearls! As audiophiles, we spend so much time on what "could" or "might" make a difference that we can easily become stuck in arguing over the impracticalities...

      - "Do USB cable change sound?"
      - "Should I buy this $4000 speaker cable?"
      - "Which is better... S/PDIF or USB?"
      - "Hey buddy - Class A is better than Class D!"
      - "My DAC has a 1kHz THD+N of -105dB, that one's better with -110dB... I need it"
      - "Which bit-perfect audio player sounds? Foobar vs. Audirvana vs. Roon?"
      - "I really need that linear power supply to make the streamer better!"

      These are all silly examples of the countless arguments we debate about over the years yet IMO the science is pretty clear about what is or is not worthwhile! Seriously guys, you don't even need an AP machine to tell you this stuff in 2021...

      So, on and on we go especially in the "professional" magazines when at the end of the day, even something as ridiculous as MQA gets by those who should know better and should recognize the stench of a "product" that has no value. How sad is that?

      In that way, unfortunately, yeah, things gets politicized. Monetary / financial / capitalistic / ego factors intrude upon what is true. The "fog of war" between different factions and tribes - "subjectivists" vs. "objectivists" - obscure what really should have been just "rational audiophiles" who appreciate the value of both sides and can see that contentious things like price and preference are decided not by any one factor - even science taken to extremes in some things will not provide wise council in all human affairs.

      So too COVID, right? Balance needs to be found. Safety and risks in life both must be sought.

      Keep up the great work on the recording / production side, Jim. As the one who is recording and doing the production of the music itself, you become the one who "authenticates" and creates. That's more powerful than any critic audiophile who just renders an opinion whether "objective" or "subjective" leaning.

  4. When a dedicated objectivist claims for the Nth time that science knows all, and our ears cannot be trusted, the First Watt amp comes to mind. I mean, isn't the subjective listening experience what it's all about? Who cares how a piece of gear measures, as long as it sounds pleasant, natural and releases a significant amount of endorphins into our bodily system?

    On the opposite side of the court is MQA. It's unbelievable to me that this fraud is still in business. Even more unbelievable that some people can't see right through the dishonesty and reluctance to explain their methods clearly. C'est la vie, I guess...

    1. Yup, agree Duck.

      For me, there are things in this world which are "knowable" in the sense that I appreciate the *facts* about them. For example, I think it is important to have manufacturers provide the key measurements like the damping factor, power ability, some idea of the distortion amounts, etc... This helps us mix and match especially at a time in history when in reality we will not be able to go into a local showroom to listen to what we might be interested in.

      But don't make "science" the only "god" to whom we must judge things which are also "affairs of the heart" like the passions and joys around owning a "desired" item. Nor trash something which has actual properties that some might love!

      No, this SIT-2 amp will absolutely not be for everyone... But it would be crazy for me to say stuff like this is "garbage", post passive-aggressive pictures of the product, or tell linnrd that he wasted his money, etc. I'd be a fool for saying / insisting such things!

      If we speak with facts, respect reasonable opinions (ie. snake oil is still never good!), debate with respect if it's clear what we're debating about, then I think as hobbyists, straddling this balance between objective/subjective is a lot of fun!

    2. I agree Archimago. Measurements are important to check manufacturers' claims and as a means to help me to decide what I may want to purchase considering the lack of local brick and mortar stores. Also they are a nice check against the purely subjective reviews that are available. Trust my ears? Sort of. Though I'd suggest more than my ears are in play when listening to music. Trust somebody else's ears? Depends who they are and my familiarity with their biases (likes and dislikes).

    3. Measurements are very important as they can show manufacturer's engineering defects once standards are accepted. It is possible for something to sound too analytical for some, but if something does not measure as the best, that doesn't mean someone may not like it or prefer it. One amp may be a better match for a pair of speakers, the room, or the listeners preferences.

      I have bought 3 pieces of Schitt gear; their $99 DAC and 2 of their headphone amps at $99 and I am very impressed at what $100 can still buy today. Their $99 dac sounded much better than the dac in my 2003 Sony DVP NS 755 DVD SACD player that stopped playing SACDs. It still spins fine and now the Schitt has made it better. It is not the quality of my 2 Project S2 dacs, but they were way more money and well worth it.

      It is still important that testing go on. I remember a test that revealed a flaw in a $5K CD player that still "sounded fine" to the reviewer, but testing found the flaw...a manufacturing defect. An unsuspecting owner would probably never have known as the reviewer did not catch it.

    4. Hi Joe,
      Yeah, ultimately, when it comes to ears, it's when an audiophile listening with his/her own ears that matters on the subjective end. I guess there are people I would trust, but verification is essential there.

      Hey Jim,
      Oh absolutely, a $100 DAC that is made well should be excellent these days. Technological maturation is a real thing and there comes a time when we're simply not going to be able to significantly hear a difference. While folks will have opinions that one is better than the other, I bet a serious listening test - a blind listening test - will attest to this even though voodoo audiophiles try their best to complain and make excuses.

      Nonetheless, devices will have other attributes and selling points. For example, that Schiit is produced in the USA is potentially a valuable characteristic to favour.

  5. I am still in the straight wire with gain camp. I have absolutely no use for gear that adds colouration or whatever. Also, I do think energy consumption from audio gear is a relevant criterion in this day and age.

  6. PKHarmonic needs quite a bit of work still and I've had problems getting it to work properly. But I can recommend GSatPlus, which is also free.
    Set Character to Warm, Stages to 1, Odd to 0 and Even to 45 and you'll get around 1%THD made up of even low-order products.

    The problem with plugins like this is that they rarely display the sort of dynamic behaviour you find in an Amp. Amps will obviously distort more as the source level rises, but these plugins apply the same distortion throughout. And my measurements of a relatively high-distortion amp show that the dynamic behaviour can be quite complex, with the level of individual distortion products following intricate patterns as the overall level changes.