Review and Measurements of the PS Audio Stellar Gain Cell DAC

[watchnerd]

Do they listen and decide? Do they fool themselves into circuits that sound like it has extra because of barely audible noise and extra meat and detail that is really sourced from nominally audible distortion? If so not much different than SET amps or such.

I think, though, that when people buy tubes, especially SETs, buyers are knowingly dipping in to the intentionally colored, eccentric end of the pool, governed by all the variables in outcomes therein, especially if using NOS tubes.

So I think, in that case, making designer products is fair to the audience; the buyers know what they're getting into.

With DACs, it's a different customer expectation, especially when using commodity chips. That's where the fail lies, IMHO.

 

[GrimSurfer]

I find % at least as useful in relation to perception. I am old though and that is all we had when I first started "get the distortion less than 0.1% then no-one can hear it".
That is -60dB so not considered particularly good in DACs but was all but unattainable in everything else except some amplifiers at some frequencies.
It still is all but unachievable in speakers and headphones at most frequencies...

This is an interesting post, @Frank Dernie, because it touches on the role that conditioning has on audibility.

Listeners back in the day were accustomed to shutting out or overlooking the relatively high noise floor of vinyl, not to mention the THD of period amps and considerable wow and flutter. So the introduction of systems with distortion less than 0.1% distortion didn't seem bad at all.

Advances in audio made in the 70s, combined with the introduction of digital music (in the form of CDs), did introduce audibly better sound to living rooms. Any person over the age of 40, therefore, has likely been conditioned to equate this level of fidelity, enjoyed in relatively quiet spaces, with "good sound".

Portable devices, IEMs, and recording practices have conditioned younger listeners to accept low DR, low FR, and low THD as being "good sound" regardless of the ambient noise environment.

So there is a frustrating subjectivity to all of this. We're all more or less capable of "hearing" the same thing but append different values to what constitutes good sound.

 

[HammerSandwich]

And when was the last time you read a news article about the politician that delivered on their promises and was beyond reproach in their personal life as well? Not that I actually believe there are any... but even if there was, it wouldn't be news.

That last sentence rather derails the argument. Something so rare is news, simply by definition.

 

[HammerSandwich]

Management: "Do I have to tell you everything? Reverse the Sprout, we have that already. In this one make the analog perform better than the digital, just not by too much."
Engineer: "But the analog section is pretty bad actually... we'd have to hobble that DAC chip entirely."
Management: "See, I knew you could figure it out! Now get to work... we need to move to production by Thursday... there's a show coming up!"

This is pretty cynical, but you may be on the right track. If someone were speccing a $1600 pre/DAC and already had a $6k DAC-only, decisions might be swayed by the potential for upgrade sales.

Yikes! Now we need a double-blind product-development process.

 

[AndrovichIV]

The reviews are nothing short of mind-blowing for me. I still had some residual confidence in professional critics but now there’s no point in even reading them. This is terrible news for the hobby and industry. Even if the intention was 100% integrity (of course it isn’t) it shows the subjective impressions and floral language are merely wasted pixels of creative fiction writing.

I wish I could like twice your comment

 

[MediumRare]

This is an interesting post, @Frank Dernie, because it touches on the role that conditioning has on audibility.

Listeners back in the day were accustomed to shutting out or overlooking the relatively high noise floor of vinyl, not to mention the THD of period amps and considerable wow and flutter. So the introduction of systems with distortion less than 0.1% distortion didn't seem bad at all.

Advances in audio made in the 70s, combined with the introduction of digital music (in the form of CDs), did introduce audibly better sound to living rooms. Any person over the age of 40, therefore, has likely been conditioned to equate this level of fidelity, enjoyed in relatively quiet spaces, with "good sound".

Portable devices, IEMs, and recording practices have conditioned younger listeners to accept low DR, low FR, and low THD as being "good sound" regardless of the ambient noise environment.

So there is a frustrating subjectivity to all of this. We're all more or less capable of "hearing" the same thing but append different values to what constitutes good sound.

Great point. Let’s also add that mastering frequency response (RIAA curve?) and even the musical “instruments “ have changed. I was at an EDM performance last night and it would be physically impossible to encode that sound on vinyl let alone reproduce it.

 

[Aprude51]

I may repeat: Gilbert/Blackmer-Cell VCAs are found in the audio chain at several places, the most prominent ones being mixing desks, (mastering) limiters/compressors, and protection circuits in many active studio monitors. Nobody freaks out about their "horrid" distortion profile there, so why should one in this case?
While I like the measurement-centric approach of ASR I'm getting more and more unhappy with the comments and interpretations of those measurements with all their "we-are-oh-so-politically-correct" attitude, anything that doesn't measure stellar and does cost a bit being considered a "rip-off" and "epic engineering failure" (notably those who are not competent electronic design engineers themselves should be careful with their wordings) . That is not to say that sometimes there are devices that really are failures and should be called such, justly.

I think you’re unfairly characterizing Amir’s comments about the use of a VCA. He says it’s unlikely that they can achieve “zero” distortion with one, as is claimed, and points out that in a discrete design there are more variables to control for. I don’t see him ever say that they’re not used effectively in other audio applications.

I think the issue here is pretty obvious: PS Audio created the “gain cell” because it allowed them to make marketing claims that have an intuitive appeal to audiophiles. They weren’t particularly concerned with the measured performance of their implementation.

 

[anmpr1]

(Seriously, just disregard this reply... I never wrote it)

Ok. But only if you disregard that I never read it.

 

[AudioSceptic]

I find % at least as useful in relation to perception. I am old though and that is all we had when I first started "get the distortion less than 0.1% then no-one can hear it".
That is -60dB so not considered particularly good in DACs but was all but unattainable in everything else except some amplifiers at some frequencies.
It still is all but unachievable in speakers and headphones at most frequencies...

Yes, a few (many?) of us remember 0.1% becoming a hi-fi "standard", and percentages were fine in the range 0.1 down to 0.01. Below that it gets more difficult, though, and I think using dB becomes increasing more useful, especially when so much else is expressed that way.

 

[LMDA1]

One of these days will appear a horse's head on Amir's bed.

Panther’s head?

 

[Cahudson42]

Go over to the ST70 review and look at the SINAD graph, To me, it looks a lot like this one. PS audio has created a 'tube amp' sounding DAC, which will tube-flavor anything following.

This apparently appeals to the PS 'monied' customer base, which grew up with stuff like the ST70..

I also think the observation may be correct that with everything below -70db cut out, that these same customers now notice more detail above that level. So we see the great subjectivist reviews as a result.

 

[amirm]

I may repeat: Gilbert/Blackmer-Cell VCAs are found in the audio chain at several places, the most prominent ones being mixing desks, (mastering) limiters/compressors, and protection circuits in many active studio monitors. Nobody freaks out about their "horrid" distortion profile there, so why should one in this case?

I assume you mean vintage analog mixing consoles. Regardless, VCAs can have far higher performance than what we see here. Here is an example: http://thatcorp.com/2181-series_Trimmable_Blackmer_IC_Voltage-Controlled_Amplifiers.shtml

PS Audio specified and I measured THD+N that is 10X higher:

Someone editing in a modern digital system like Pro Tools using a ADCs/DACs, is not using VCAs.

Even if they were, what is on the recording is what is on the recording. We don't want to overlay the same distortion effect on every piece of music we play. It is like saying if a painting is release with a tint pink, we should wear pink glasses to view that, and every other painting.

Also, pay attention to what this thing is doing:

THe problem is not just harmonic distortion that is setting SINAD. There are also the other artifacts I note. Here is the FFT of the VCA part I mentioned above:

Notice the lack of anything else other than harmonic distortion.

Sitting here, I don't know how anyone could have done a single FFT, saw what I say and thought, "this is clean enough; let's ship it." What is high-end if it is not precision in design and overengineering?

 

[amirm]

While I like the measurement-centric approach of ASR I'm getting more and more unhappy with the comments and interpretations of those measurements with all their "we-are-oh-so-politically-correct" attitude, anything that doesn't measure stellar and does cost a bit being considered a "rip-off" and "epic engineering failure" (notably those who are not competent electronic design engineers themselves should be careful with their wordings) . That is not to say that sometimes there are devices that really are failures and should be called such, justly.

That is not what is going on here. My objections are objective. I just reviewed a $500 device and gave it the same failing grade: https://www.audiosciencereview.com/...urements-of-furutech-adl-gt40-dac-phono.9279/

It has a phono stage that has a frequency response that is anything but flat.

When a device finishes in bottom 10 of nearly 200 DACs tested, what reaction do you expect us to have? Cool and just walk past the review?

Before I do any of these reviews, I search for others. There is so much praise and flowery word about this product and its gain cell. Surely all of that is painting a perfect device. Having some fury here against such impression left in the eyes of readers is merited and fine.

We had a similar situation with Schiit. Riots everywhere against the work here. At the end, Schiit quietly bought a proper analyzer and started to produce better products. If we had been quiet about it, this would have never happened.

So no, I don't mind people taking notice and expressing their opinion about the results of the review.

Watch this video of Paul with nearly 11,000 views:

He says it is the most transparent he has heard. You think that is true? If not, then we need to speak up.

 

[JJB70]

It's a DAC, it has a defined function in the audio chain. The Apple comparison is germane to the discussion as it demonstrates that the core function of converting a digital signal to analogue can be done for the price of a big mac meal. Adding the various input and output functionality, a nice case etc will bump the price up, but to nothing like $1700. If you are going to charge so much for a part of the chain which was commoditised years ago then you really need to to offer something very special for the product to have a reason to exist IMO.

I just can't see the fascination in DACs and really question why anyone needs one given that so many amplifiers now have digital inputs and devices have on board DACs. OK this is a bad one, but even if it really was stellar would it really make much difference to anything? Yes you can measure differences in DACs but the measured performance of very average DACs is so far ahead of the rest of the audio chain that it is probably the least productive part of the chain to spend on if you want to boost SQ. Once you achieve transparency in a DAC, which can be done for peanuts, then any further improvement in measurement may be reassuring and indicate excellent design but it won't alter the listening experience.

I think that the real issue is that the Audio industry is convincing a lot of people to spend $$$$$$$$$$$$$$s on DACs which are completely unnecessary. People slam audiophile cables yet seem happy to applaud expensive DACs (when they work well, not this one) yet I struggle to really see much difference between the audio culture around cables and DACs.

 

[Blumlein 88]

That's why I can't help seeing these products as nothing more than a glorified and ridiculously overpriced from of circuit bending. It seems like a colossal waste of development resources when the same result could be had by deliberately messing up the circuit in any dirt cheap DAC.

snip

I agree with this, but DSP bending of sound won't sale. Colossal wastes to bend circuits in ways that aren't adjustable sell like hotcakes if you spin it halfway right.

Also, often times state of the art would follow on with a description of the limits of some area. Not that it was superlative, but simply a list of what is possible and what is not. Even this other definition of the PS audio gear doesn't fit, and much more is possible. They are no where near the envelope for state of the art.

 

[GrimSurfer]

I just can't see the fascination in DACs and really question why anyone needs one given that so many amplifiers now have digital inputs and devices have on board DACs.

They're even worse performing, if you can believe it.

The all-in-one packaging of audio devices only serves one market: Those who want sound, and want it now, at price less than a good BBQ. They work on the correct assumption that most people will be happy if the thing sparks up, and are not designed to meet the needs of discerning listeners.

This isn't some snobbish statement but a reflection of the following:

You can have it quick, have it cheap, or have it good. Pick any two.

 

[Krunok]

I may repeat: Gilbert/Blackmer-Cell VCAs are found in the audio chain at several places, the most prominent ones being mixing desks, (mastering) limiters/compressors, and protection circuits in many active studio monitors. Nobody freaks out about their "horrid" distortion profile there, so why should one in this case?
While I like the measurement-centric approach of ASR I'm getting more and more unhappy with the comments and interpretations of those measurements with all their "we-are-oh-so-politically-correct" attitude, anything that doesn't measure stellar and does cost a bit being considered a "rip-off" and "epic engineering failure" (notably those who are not competent electronic design engineers themselves should be careful with their wordings) . That is not to say that sometimes there are devices that really are failures and should be called such, justly.

Well said.

 

[Blumlein 88]

IF (and a big IF) the sound is actually audibly colored, then what you have is a processing effect that can't be turned off or adjusted. So irrespective of whether a user likes the effect, I can't imagine that it would be liked on every recording...

While true, what happens is a customer makes a commitment to a given device based upon how it sounds on some favorite music, or some reviewer's praises or some marketing spin. Then they judge the music quality based upon the sound of the device, not the other way around.

 

[SIY]

They're even worse performing, if you can believe it.

I have a review unit sitting in my lab right now which combines streaming, signal processing, and amplification. And the DAC portion has analog outs as well. Performing beautifully so far... so these thing CAN be well engineered.

 

[SIY]

While true, what happens is a customer makes a commitment to a given device based upon how it sounds on some favorite music, or some reviewer's praises or some marketing spin. Then they judge the music quality based upon the sound of the device, not the other way around.

I suspect the large majority of purchasers aren’t even detecting any actual sound differences. They are quite content with the story.