Review and Measurements of PS Audio PerfectWave DirectStream DAC

[LTig]

Somewhere in those top two tiers of DACs, you're just not going to hear a difference anymore. Possibly below that level, depending on your hearing, training, amp, and speakers or headphones.

The above statement is the problem. If you want to make the assertion about what is below that very arbitrary level, that you can possibly hear it, that is just throwing FUD. To be valid in the methodology of science, you would need to document a few cases under rigorous listening testing that you could indeed hear something that is detrimental.

The limits of human hearing are already very well known. I can see no FUD here.

Here are the aditional problems when you dig deeper into the meaning of a single number that ranks it.

1. Equipment A and B rank the same in that number. One has higher harmonic distortion and lower noise while the other has higher noise level but lower harmonic distortion. Does that number mean they are necessarily equally bad and someone would hear that in both?

It depends on the circumstances.

2. Or take a variant of the above case when equipment A is in the “pass” category in that arbitrary chop off and equipment B is below it. All of B’s numbers come from noise most of which is in the inaudible range with very little harmonic distortion. Most of A’s numbers come from harmonic distortion with a very low noise floor. Is it reasonable to say in the interests of science that A is a “pass” while B is a “fail”?

It depends on the circumstances.

 

[Blumlein 88]

Here’s what their product page states:

“Classic PCM based DACS, including many of today’s DACS that can also process DSD, tend to cover up some of the subtle musical details buried deep within digital audio music; a problem inherent in their architecture. DirectStream solves this problem by employing a pure DSD single-bit approach for both PCM as well as DSD media. This means that your entire library of music can finally reveal all the music and subtle low-level details buried deep within its core.”

Hmm...

I think this is a fair statement. Well in a certain sort of way. The subtle low level details are certainly buried deep, way deep within a core of noise. It takes extraordinary effort by the listener to overcome. One just has to believe enough and all will be revealed. But not to people like Amir. He thinks a DACs job is to just lay it all out there in the open for any Tom, Dick or Harry Pearson to hear with no effort. Where is the mystery and imagination in that?

 

[LTig]

But you aren’t solving that problem either, just replacing it with “How much SINAD improvement will it take to lift that veil for my use”? 80db? 90db? 100db?

It depends in the circumstances. Tell them here and you'll get an answer.

If cost was no problem, all of this would be moot. Even buy the highest rating you can afford is bad because that may lead to people spending more than what they need. That is like saying buy the highest HP car you can afford.

Tell your use case here and you'll get an answer.

If someone had the resources to buy an Okto DAC, pair with Benchmark, all the decoding stuff that is needed to handle the home theater and somehow incorporate room correction etc, etc, sure no one would have to worry about noise or distortion but are they going to find it necessarily better than the $1500 AVR that measured poorly commensurate with the costs?

If not, then this has no more practical world use than the esoteric gear reviewers.

That's not true. For example you can get some very good DACs for much less cost than most of the esoteric stuff. For around $1000 you get an RME ADI-2 DAC with lots of additional features and an excellent SOTA performance. No need to look for Total DAC or PS. You can invest the remaining 5 to 10k$ into excellent active speakers and be done. If this is not practical advice then what is?

Or read this review and look at the comparison with Kadas Tone board for 100$. Buy the board, put it into a nice case and you're guranteed to have better SQ than PS. If this is not practical advice then what is?

 

[Guermantes]

@audimus , you keep complaining about this site being an "echo chamber" with members more interested in "charts and jargon that constitute tech-porn" and unable to relate that to "real world" applications.

There are very, very few people doing third-party measurements of consumer audio equipment in comparison with a multitude of reviews based on subjective impressions. Everyone with a set of ears is an expert in what they like. I think this site is justified in getting technical since there is so little data out there of that nature. The tech-porn is an antidote to the affect-porn.

Yes, context is important and some of your concerns about how measurements realistically reflect the user's experience are valid. These are questions that often come up and produce interesting and productive debate on ASR.

 

[amirm]

Let me see if I can put performance of the PS Audio DS DAC in much simpler terms. Let's take a 1 kHz 24-bit, sine wave and starting with -50 dB, keep lowering the digital signal to see how far it can go:

EDIT: Forgot to mention originally that the output of the DAC was limited to a bandwidth of 22.4 kHz. So it excludes high frequency noise.

We start OK but by -70 dB we are getting a bit wiggly. By -90 dB (far right), any resemblance to a sine wave is all gone and replaced with noise. -90 dB is 1 bit or 6 dB shy of what CD can do! So the notion that the DS DAC brings out detail in music that is not normally heard is just not right.

Let's see how good it can get using the Matrix Audio ESS X-Sabre Pro MQA DAC:

See the amazing difference? Even at -90 dB the Matrix DAC is sailing easily with no noise or distortion.

But wait.... it gets better! Let's keep reducing the amplitude of the signal with Matrix DAC and see how far it can go:

Even at -115 dB it has a better output than PS Audio DS DAC has at just -90 dB!!!

Conclusions
There is no way, no how the PS Audio DS DAC is a resolving DAC. It can't even do justice to 16 bit audio let alone 24 bit audio. Its output gets swamped with noise that is far louder than the signal itself at -90 dB or 15 bits. Actual useable range is even lower than this then.

Compare that to the incredible performance of the Matrix Audio X-Sabre MQA Pro. As the name indicates, it uses an ESS integrated DAC chip surrounded by superb engineering. The result is that it does justice to any resolution audio we can throw at it.

And oh, the Matrix Audio X-Sabre DAC costs one third of PS Audio DS DAC! It has a gorgeous enclosure and incredible build quality you can be proud of. No one got cheap giving you subpar parts as is the case with PS Audio DAC.

And oh, be careful in looking at any evidence of resolution using FFT spectrum. FFT sharply reduces the noise in measurements uncovering signals deep in the noise. Your ear does a bit of that but is not remotely as good as an FFT. And regardless, why have a DAC that needs that kind of noise reduction to sound good? Doesn't your brain get tired of all that noise????

 

[amirm]

Here are the aditional problems when you dig deeper into the meaning of a single number that ranks it.

What is all this talk about a single number? We have divided the entire range of DAC SIAND into just four buckets. Great, good, OK and poor. And that is exactly what they are.

 

[direstraitsfan98]

Why do you use so many exclamation marks and question marks and emojis in your posts?

 

[AudioSceptic]

Let me see if I can put performance of the PS Audio DS DAC in much simpler terms. Let's take a 1 kHz 24-bit, sine wave and starting with -50 dB, keep lowering the digital signal to see how far it can go:

View attachment 34485

We start OK but by -70 dB we are getting a bit wiggly. By -90 dB (far right), any resemblance to a sine wave is all gone and replaced with noise. -90 dB is 1 bit or 6 dB shy of what CD can do! So the notion that the DS DAC brings out detail in music that is not normally heard is just not right.

Let's see how good it can get using the Matrix Audio ESS X-Sabre Pro MQA DAC:

View attachment 34486

See the amazing difference? Even at -90 dB the Matrix DAC is sailing easily with no noise or distortion.

But wait.... it gets better! Let's keep reducing the amplitude of the signal with Matrix DAC and see how far it can go:

View attachment 34487

Even at -115 dB it has a better output than PS Audio DS DAC has at just -90 dB!!!

Conclusions
There is no way, no how the PS Audio DS DAC is a resolving DAC. It can't even do justice to 16 bit audio let alone 24 bit audio. Its output gets swamped with noise that is far louder than the signal itself at -90 dB or 15 bits. Actual useable range is even lower than this then.

Compare that to the incredible performance of the Matrix Audio X-Sabre MQA Pro. As the name indicates, it uses an ESS integrated DAC chip surrounded by superb engineering. The result is that it does justice to any resolution audio we can throw at it.

And oh, the Matrix Audio X-Sabre DAC costs one third of PS Audio DS DAC! It has a gorgeous enclosure and incredible build quality you can be proud of. No one got cheap giving you subpar parts as is the case with PS Audio DAC.

And oh, be careful in looking at any evidence of resolution using FFT spectrum. FFT sharply reduces the noise in measurements uncovering signals deep in the noise. Your ear does a bit of that but is not remotely as good as an FFT. And regardless, why have a DAC that needs that kind of noise reduction to sound good? Doesn't your brain get tired of all that noise????

So the PS Audio is worse than the Matrix Audio by 30 dB (~5 bits). It's amazing how the PS at -80 looks so similar to the Matrix at -110. The PS is shot by -90 but if it can manage -85, I'd bet it looks like the Matrix at -115. That's some margin.

 

[mkawa]

speaking seriously, if you want noise, especially in H2 and H3, stick a tube buffer into your chain at line level or as your power amp. a poor dac implementation is not a what you want to invest thousands in..

 

[Stephensank]

Sorry, but you show a lot of prejudice there- Any *properly* designed/built tube buffer or linestage is capable of beyond -120dBV below 1V. I know, because I've built such units numerous times.
Certainly will say, though, that 95% of "tube dacs" are simply ss dacs with a token tube stage at the end of an opamp chain, and there simply to mask the damage of poorly executed opamp circuits via euphonic coloration.

 

[Guermantes]

What is all this talk about a single number? We have divided the entire range of DAC SIAND into just four buckets. Great, good, OK and poor. And that is exactly what they are.

I think the criticism is that the SINAD number is being presented as a single score card for the device, or at least being used by others for that purpose. Much like the star ratings of magazine reviews, I suppose. In fact, if you convert the 4 buckets into stars, you could arrive at the same type of evaluation: e.g. the PS Audio DirectStream would be 1 out of 4 stars. The complaint is that this is not an indicator of "real world" user experience of the device but purely an engineering metric.

 

[Sal1950]

Let me see if I can put performance of the PS Audio DS DAC in much simpler terms. Let's take a 1 kHz 24-bit, sine wave and starting with -50 dB, keep lowering the digital signal to see how far it can go:

Thanks for that. It illustrates the differences in a much better manner than just the numbers present to those of us technically challenged.

 

[MZKM]

I think the criticism is that the SINAD number is being presented as a single score card for the device, or at least being used by others for that purpose. Much like the star ratings of magazine reviews, I suppose. In fact, if you convert the 4 buckets into stars, you could arrive at the same type of evaluation: e.g. the PS Audio DirectStream would be 1 out of 4 stars. The complaint is that this is not an indicator of "real world" user experience of the device but purely an engineering metric.

You can make the claim that all well-functioning DACs sound similar, so how else would you rank them other than subjectively?

Take a $100 DAC and put it in a nice enclosure, boom, you just made a $6000 PS Audio DAC equivalent. Sure it sounds good, but why pay $5900 more just for looks?

 

[mkawa]

Sorry, but you show a lot of prejudice there- Any *properly* designed/built tube buffer or linestage is capable of beyond -120dBV below 1V. I know, because I've built such units numerous times.
Certainly will say, though, that 95% of "tube dacs" are simply ss dacs with a token tube stage at the end of an opamp chain, and there simply to mask the damage of poorly executed opamp circuits via euphonic coloration.

oh, i didn't mean that to be a knock on tube circuits, but the opposite: tube buffers have been shown to be great for adding the distortion modes that people respond to positively, ie, as an analog filter. coming from the guitar amplifier perspective, we use tube stages to get that cool floaty sound from les pauls and old teles with tons of timbre. my understanding is that this measures out as H2 and H3 distortion and that this is what the audiophile crowd on the PS audio forums are referring to in what they like about the "perfectwave"

 

[Stephensank]

Apologies then. In the case of this PS dac, though, the damage from WILDLY mismatched & POOR quality transformers is vastly beyond even the most colored tube stage that ain't just flat out defective.
Quite obviously, PS chose the low 2.8/1.4Vrms max output deliberately, as they found that the max level they could get out of the horribly chosen xfmrs, and yet it can't even to THAT level, shown clearly by Amir's -8dB threshold of saturation distortion. There is really NO excuse for using such a pathetic excuse for output iron.

 

[August]

Why do you use so many exclamation marks and question marks and emojis in your posts?

I think this is a restraint.In order to prevent swearing

 

[mkawa]

hah! i was a ball of trouble in my teens when i first started playing with audio electronics (and took out of a lot of frustration on my guitar, to the detriment of the neighbors..), and at one point i was taken to a spiritual healer who claimed he could see auras. my memory of this incident is that i was sat down in front of him, he stared at me for a while, took out some stones, gave them to me to hold and then started reading my aura. he waved his arounds around my head like.. 6 inches away? for about 10 minutes and then he said something like "i sense that you're very sick" (spoiler: i was). he recommended some carefully chosen stones and then did some breathing exercises (as in he did them) while he waved his hands around my head for 10 minutes. when he was done he said "hope that helps" and waved me off for the next person.

i'm not sure, but i think he was paid somewhere in the 1-200$ range for this service.

in short, still a better deal than this dac.

 

[Michael_B]

hah! i was a ball of trouble in my teens when i first started playing with audio electronics (and took out of a lot of frustration on my guitar, to the detriment of the neighbors..), and at one point i was taken to a spiritual healer who claimed he could see auras. my memory of this incident is that i was sat down in front of him, he stared at me for a while, took out some stones, gave them to me to hold and then started reading my aura. he waved his arounds around my head like.. 6 inches away? for about 10 minutes and then he said something like "i sense that you're very sick" (spoiler: i was). he recommended some carefully chosen stones and then did some breathing exercises (as in he did them) while he waved his hands around my head for 10 minutes. when he was done he said "hope that helps" and waved me off for the next person.

i'm not sure, but i think he was paid somewhere in the 1-200$ range for this service.

in short, still a better deal than this dac.

Well played, sir.

 

[HammerSandwich]

I think the criticism is that the SINAD number is being presented as a single score card for the device, or at least being used by others for that purpose.

That's certainly a drawback.

I'd be quite satisfied with a 90dB-SINAD DAC that has a single H2 spike with nothing else above -120dB.

 

[tobes]

Apologies then. In the case of this PS dac, though, the damage from WILDLY mismatched & POOR quality transformers is vastly beyond even the most colored tube stage that ain't just flat out defective.
Quite obviously, PS chose the low 2.8/1.4Vrms max output deliberately, as they found that the max level they could get out of the horribly chosen xfmrs, and yet it can't even to THAT level, shown clearly by Amir's -8dB threshold of saturation distortion. There is really NO excuse for using such a pathetic excuse for output iron.

I don't think the transformers are the reason output is kept low in the DS dac (the transformerless DS Junior has similar low output).
The DS dacs are set up to operate as preamps with digital volume control. However they can become audibly noisy if used with too much digital attenuation (audible hiss). Having lower output means less digital attenuation can be used for typical listening levels with typical power amps. The DS dacs also have a 20dB analog attenuator which helps minimise digital attenuation in some situations.
Not saying this is good - it indicates a lack of bits to play with - just that there may be a reason for the lower output.