Pass Labs HPA-1 Headphone Amp Review

[Drakkar Noir]

I am actually happy for @amirm 's success and influence as a reviewer, an equally happy that some of the big players like the one @WayneC represents here are taking notice in this trend. Should not be so difficult for a company like Pass Labs to change a screw, connector or wire here and there to make their products more competitive in the measurements department. We all stand to benefit from this.

I for one know PS Audio and Schiit are currently releasing new and updated products that they claim measure well, no doubt because of reviews on forums like this one.

Sucks that amongst the useful information provided here there are those that are only too happy to pile on the bashing of some imperfect component for no good reason. You know who you are.

 

[HiFiBob]

The HPA-1 isn't junk. It's a well-engineered and solidly built product that doesn't have "ultralow distortion" on the Marketing Requirements Definition spec sheet. I don't think it's accurate to invoke P. T. Barnum. Changing a screw, connector, or wire here and there isn't needed and won't improve measured performance. Pass Labs built the amp this way for specific reasons. Whether you agree with those reasons is a separate question.

I could post the reverse-engineered schematic from diyaudio.com sometime and go through some of the design details and rationale. Probably not since most people won't care. There's a few things about the circuit I disagree with. I know from the diyaudio discussion that Jam feels any changes to "improve" performance would detract from the sound. That's a personal subjective opinion and I can't argue with it.

For instance, the unbuffered front end is heavily loaded by two resistors which set the overall open-loop gain to a low value, giving IIRC about 18dB NFB. That could easily be changed to give more feedback but then it would be a different design. It's not bad engineering, just a conscious decision on Jam's part.

About the only thing I think is really wrong with the amp is the design of the output stage bias spreader circuit. The way it's implemented, the amp takes a long time for the output stage idle current to settle down. This could have been done better, IMHO, and would probably do away with the always-on recommendation. Electrolytic capacitors only have a few thousand hours operational lifespan, so you can easily wear them out in a few years if you never turn the amp off.

There are a few other compromises which I think have to do with the parts that were available in the Pass Labs stock room. Overall, I think it's a pretty elegant circuit.

 

[HiFiBob]

By the way, I didn't come here to cheerlead for the HPA-1. I think it's a nice-sounding amp, but to my tin ears not the magical experience you'd expect given all the hype and breathless subjective reviews. I just happen to have sentimental feelings about discrete transistors that have nothing to do with sound quality. The HPA-1 does sound a little different from other amps. There may be people with more refined hearing than me who legitimately appreciate that difference, or maybe they just have their own sentimental attachments to the thing. It's not my place to question their preferences.

I would say it's equally misleading to market an amp as "better" because it has 101dB SINAD versus 97dB. That's just nonsense. I keep wondering what the point is of ranking components in order of inaudible spec differences.

But I really don't want to argue. Fighting is so tedious.

 

[Drakkar Noir]

The HPA-1 isn't junk. It's a well-engineered and solidly built product that doesn't have "ultralow distortion" on the Marketing Requirements Definition spec sheet. I don't think it's accurate to invoke P. T. Barnum. Changing a screw, connector, or wire here and there isn't needed and won't improve measured performance. Pass Labs built the amp this way for specific reasons. Whether you agree with those reasons is a separate question.

I could post the reverse-engineered schematic from diyaudio.com sometime and go through some of the design details and rationale. Probably not since most people won't care. There's a few things about the circuit I disagree with. I know from the diyaudio discussion that Jam feels any changes to "improve" performance would detract from the sound. That's a personal subjective opinion and I can't argue with it.

For instance, the unbuffered front end is heavily loaded by two resistors which set the overall open-loop gain to a low value, giving IIRC about 18dB NFB. That could easily be changed to give more feedback but then it would be a different design. It's not bad engineering, just a conscious decision on Jam's part.

About the only thing I think is really wrong with the amp is the design of the output stage bias spreader circuit. The way it's implemented, the amp takes a long time for the output stage idle current to settle down. This could have been done better, IMHO, and would probably do away with the always-on recommendation. Electrolytic capacitors only have a few thousand hours operational lifespan, so you can easily wear them out in a few years if you never turn the amp off.

There are a few other compromises which I think have to do with the parts that were available in the Pass Labs stock room. Overall, I think it's a pretty elegant circuit.

My comment about improving measured performance is more about it being not so difficult to obtain, IF you wish to go that route. I mean, some of the more recent brands are achieving this constantly and with what I assume are limited funds. I'm sure Pass Labs is able to afford the best measuring equipment.

As I've stated before, to me this makes very little difference. I listen to mostly rock music and my favorite headphones with the Pass are the very euphonic Audeze LCD-3 and Grado RS1, so yeah. These headphones make Pearl Jam sound even more like Pearl Jam. Doesn't matter to me if I can't listen to Eddie Vedder's nose hairs move with this setup. That's what it all comes down to, isn't it? better performance would mean better detail retrieval. To me the Pass is good enough, great even.

If there is any sloppiness in the build or mishandling during shipping to Amir, Pass will hopefully fix this.

 

[oleg87]

For my money, measurebating over astronomical SINAD ratios far beyond human hearing is far more irrational than chasing whatever "euphonic" colorations get subjectivists going.

Unless you're looking for something that doubles as a low-cost lab instrument, transparent, low-cost headphone amps of moderate power levels are a solved problem. No value Pass or any company that can't fab high-performance op-amps can add there.... feature set, that's a different discussion.

 

[DonR]

For my money, measurebating over astronomical SINAD ratios far beyond human hearing is far more irrational than chasing whatever "euphonic" colorations get subjectivists going.

Unless you're looking for something that doubles as a low-cost lab instrument, transparent, low-cost headphone amps of moderate power levels are a solved problem. No value Pass or any company that can't fab high-performance op-amps can add there.... feature set, that's a different discussion.

Ditto for DACs. Once the audibility threshold is passed, the price becomes a sticking point for me particularly if there is no other objective or subjective embellishments.

 

[Xulonn]

As I've stated before, to me this makes very little difference. I listen to mostly rock music and my favorite headphones with the Pass are the very euphonic Audeze LCD-3 and Grado RS1, so yeah. These headphones make Pearl Jam sound even more like Pearl Jam.

Wow! So you are saying that these headphones and the Pass amplifier produce a sound signature that accurately duplicates rock concert venue PA systems?

That's quite an accomplishment!

/s

 

[Drakkar Noir]

NO. I'm saying I find rock n roll very pleasing and exiting with this setup. Very simple.

 

[amirm]

For my money, measurebating over astronomical SINAD ratios far beyond human hearing is far more irrational than chasing whatever "euphonic" colorations get subjectivists going.

There is a time to make that argument, and there is a time not to. SINAD at high values indicates noise floor. With headphone amps and sensitive IEMs, noise level matters hugely. Another thing that is king is amount of power. Without that, you are going to distort and in anything but "euphonic" way. Given this massive box and heavy weight, there is no reason to not have more power than this box outputs. These are the things that damn the product, not the pure distortion story.

Here is again the power story:

My *minimum* bar for power at 300 ohm is 100 milliwatts. Desktop products go way past this routinely including this little $109 box from Schiit:

That is 3X more power. This is hardcore, reliable, objective data that directly translates into subjective listening.

So please park the talking points at the door. This box is simply without a market in high-end or even middle of the headphone market. Folks routinely buy insensitive headphones these days and need a clean, noise-free, powerful source to drive them.

 

[amirm]

I appreciate that Nelson Pass employs American workers and provides them health insurance and benefits. You can build your own HPA-1 clone. Good luck making it the same quality for under a thousand dollars, and that doesn't account for the costs of labor, marketing, customer support, and keeping the factory lights on.

So that is how it works? My $3,000 spent on something other than this box like college education for kids, a vacation, etc. would not support American labor but this box does? And what about the fact that there are at least three American companies with products that cost a fraction of this with better performance? Money spent with Schiit and JDS Labs doesn't count? But spending thousands on this box does? Good grief.

 

[amirm]

I've read all the reviews of the HPA-1. They're all very positive, except Amir's.

How many negative reviews do those people post? Close if not zero, isn't that right? Pass could take a brick, glue RCA connectors on it and these guys would rave about it! This is the power of branding and non-critical evaluation of audio masquerading as "review." This review is reason enough to stop reading those reviews, not second guessing mine.

Remember, I have tested more headphone amps than all of those people combined. The count is easily in hundreds. I actually own probably 50 to 100 one right now! I not only measure each one, I listen to each one using proven, critical methodology that brings out weaknesses of them. I don't just listen to some random new album with each review. This is why all up, I only give thumbs up to only 1/3 of all devices I test:

I suggest waking up and not going by non-critical tests conducted by these "reviewers" and yourself for that matter. I push products to the limit to find their weaknesses. That you or others think everything sounds great is neither here, nor there. You don't have the measurements I have, and comparative data to demonstrate your opinion is valid.

 

[amirm]

It legit makes me sad to see people hating on this amp, and I'm not even a Pass fanboi. I do get tired reading about cookie-cutter Chinese gear all the time. Pass Labs' design philosophy is different. It's a big world, full of injustice. There are plenty of more important things to lose sleep over. IMHO and YMMV.

The "philosophy" is to give you a big shiny box with way too little performance -- both objectively and subjectively. If this was a glass vase, sure, you would just look at it. But it is not. While many including myself value looks as well, high-fidelity and limitless performance has to come at these prices. I should not be able to plug in a headphone, crank up the volume and have it distort. What on earth am I paying for if it is not for avoidance of that? My amplifiers driving my speakers are insanely expensive. But they output 1000 watts into 4 ohm. That is the limitless power I need to drive my speakers. So when someone asks me why I bought them, I can point to a reason. That reason doesn't exist here.

To be sure, I am not here to judge whether any product should or should not be in the market. You all do. I present data and provide metrics that show the limits of the product and tell its true story. And that story says you are paying for metal and not performance here. And in audio, performance better be priority #1 or you are in the wrong hobby.

 

[HiFiBob]

"Fighting is so tedious."

 

[amirm]

I wish Amir had arranged to have the amp inspected at the factory before publishing such a scathing review.

If the company had supplied me with the unit, I would have done exactly that. But when members send something to be tested, I do that and post the results. You need to be pushing companies to publish comprehensive measurements like I do so there is no doubt as to validity of the testing. Until they do, this situation will remain.

 

[Xulonn]

NO. I'm saying I find rock n roll very pleasing and exiting with this setup. Very simple.

And I thoroughly enjoyed rock music in the late 1970's played through my pair of big custom JBL bass reflex speakers (from Hank Hong of Honker's Sound Company in Berkeley, CA). The drivers were the famous JBL S8 drivers with 15" woofer, and powered by a McIntosh receiver. This speakers below are JBL C50's - with the same S8 drivers and crossovers.

Hank was a friend and business associate of Nelson Pass ( who had started Threshold Audio two years earlier, shortly after getting his degree in physics at the University of California, Davis.) Technically, Nelson Pass knows physics and electronics, and is far more qualified than characters like Danny Ritchie (business degree) or Paul McGowan ("self-taught", no degree). On a related note, there is lots of interesting discussion about the evolution of the audio scene and businesses in California at the Audiokarma.org thread on Pacific Stereo stores history HERE. Nelson Pass was - and is - an integral figure in the California audio scene, and has designed a number of excellent components that are now considered classics- including the excellent classic Adcom GFA555 power amplifier.

I also owned, in the early 2000's, and enjoyed very much a pair of very efficient Klipsch Forte II's powered by a flea-power custom 2A3 single-ended amplifier with NOS Marconi tubes from Italy's Attilio Caccamo of Tektron.

I am sure that neither of those systems would measure well, but they both brought me great pleasure. It is always good to frame comments about one's own personal; experiences and pleasures in the first person, and not as generalities. I am very aware that the sound waves that strike my tympanic membranes do not get passed accurately to my consciousness - the brain "converts and interprets" those sound waves in ways of which I am not aware, and cannot objectively evaluate. However, well-designed and implemented double blind ASX testing can remove some of the most obvious influences that alter my perception of the sounds I think that I "hear" - and can allow me to know if I really do hear "differences".

If someone thinks that $3.7k is a fair price for a headphone amplifier sans DAC without balanced input and output, and not enough juice for "power hungry" headphones, go for it. As for me, I'll stick with My $350 Topping DX7s DAC/headphone amp which doesn't have those limitations, and spend the $3.3k I saved elsewhere.

 

[HiFiBob]

Amir:

Per JA's distortion graph in Stereophile, the HPA-1 clips at about 8.5VRMS out into 30 Ohms. The amp has a robust output stage and power supply, so should produce sufficient current and voltage to meet its claimed power output of 3500mW into 20 Ohms.

With a gain of 8dB (x2.5), you cannot actually "push the product to its limit" with the 2VRMS output of your analyzer. Your readers may not understand that and might draw the wrong conclusion about the amplifier's power rating.

To grok this amplifier, you need to prepare yourself psychologically for the concept of 0.1% distortion. The ear is remarkably insensitive to 2nd harmonic distortion in real-world signals. Nelson Pass has written plenty about this. You just don't need ultra-low distortion for audible transparency.

The claim is that sighted listening tests are worthless. So I choose to ignore your subjective evaluation. I do wonder if your impressions of the amplifier's sound were colored by knowledge of the measurements.

About made-in-China versus made-in-USA, it doesn't seem appropriate to discuss that topic here. The issues are obvious to anyone who reads the news.

I don't feel the need to address your other comments. I'm satisfied with everything I've posted here. Thank you for giving me the opportunity to respond to your review of the HPA-1.

 

[Alcophone]

If someone thinks that $3.7k is a fair price for a headphone amplifier sans DAC without balanced input and output, and not enough juice for "power hungry" headphones, go for it. As for me, I'll stick with My $350 Topping DX7s DAC/headphone amp which doesn't have those limitations, and spend the $3.3k I saved elsewhere.

The DCA Stealth is a power hungry headphone. It sounds beautiful out of the HPA-1. It's plenty loud with the volume level at noon. With some songs I get so into it, I turn it to 2 or 3pm, but very rarely because that's hearing damage territory.
More than enough power for this application, it would seem. But there's a graph on the internet says it doesn't have enough power, so what do I know.

I had the DX7s. I didn't enjoy it very much and sold it long before I got the Stealth. Glad it works for you. Did they fix the issue were it stops working at some point until you disconnect power for long enough?

I also got the SparkoS Aries ($3000) around the time I got the Stealth and HPA-1 because the SparkoS opamps let me enjoy the Burson Fun and Gustard H20. The Aries is well respected among audiophiles. Someone preferred it over the Benchmark HPA4.
The Aries is also a single ended amplifier with balanced inputs for convenience, and strangely it needed them to sound just okay to me with the Yggdrasil, while the HPA-1 sounds much more pleasant with just the single ended output of the balanced Yggdrasil. I tried my best to like the Aries, trying most of the headphones in my inventory, but it wasn't happening. I returned the Aries and kept the HPA-1, which reliably provides a beautiful sound bubble to stick my head in and relax.
But sure, it must have been the $500 price difference that convinced me.

Would I prefer to spend 10% or less and get something I enjoy as much? Yes. Sadly, even Amir hasn't cracked the code yet of posting a number that correlates with "sounds good to me". The latter is the performance I care about. I don't worry about the AAF (Analyzer Acceptance Factor).

But I'm always curious and open minded. If someone is crazy enough to send me an amp that will certainly trump the HPA-1, or a DAC to trump the Yggdrasil OG, I'll happily try it out. Unfortunately my cat isn't qualified to facilitate blind testing, so my feedback can be all too easily discarded.

If only we could measure brain states while listening to gear and post screenshots of those.

 

[Xulonn]

The claim is that sighted listening tests are worthless.

That is one of the lamest strawman fallacies that infects ASR. No one who understands science and agrees with the fundamental principles that guide ASR says that.

Amir and most, if not all, of the rest of the rational objectivists here at ASR know that there are real, easily to moderately difficult to distinguish audible differences in reproduced music and other sounds. With good electronics and proper component matching, differences exist mostly with respect to transducers and environments, and for audiophiles, that means speakers and listening rooms. Sonic anomalies related to speakers and their interaction with rooms are not just common, but ubiquitous on the world of audio.

We also know that there are human limits on audibility, and that scientific instruments can be far more sensitive with respect to sound than human hearing. Intelligent, aware audiophiles and audio professionals are aware of both the limits of human hearing, and where and when double blind ABX testing is the only way to verify claimed audible differences.

To claim that objectivists say that "sighted listening tests are worthless" is simply a strawman based on ignorance or denial.

 

[amirm]

With a gain of 8dB (x2.5), you cannot actually "push the product to its limit" with the 2VRMS output of your analyzer.

2 volts is the maximum DAC RCA output of almost all the DACs I test. It is not material what the product does beyond that level of input. You are not likely to have that kind of source. Heck, a lot of mass consumer products don't even do 2 volts out (think many AVRs).

The competitors to this product have 2 or 3 gain settings and hence, are able to deal with this or other output levels from source components.

To grok this amplifier, you need to prepare yourself psychologically for the concept of 0.1% distortion.

Nope. To grok this amplifier you have to be ready for it to go into clipping with insensitive headphones. And have noise with every sensitive IEMs/headphones.

 

[HiFiBob]

You're just wrong Amir. It's not a good review. But the thing is, nothing good ever comes from arguing with a narcissist holding a ban hammer.

LOL, I've said what I came to say. See ya.