• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Review and Measurements of PS Audio PerfectWave DirectStream DAC

For sure, in that case, Joe explained that he wanted a special sound and reverb and he selected the room for that.
That gives that kind of an "explosive" sound.
I like this album since its launch... 35 years ago :-S
Click to expand...

… and there were simply no digital reverbs available...
 
Herbert said:
… and there were simply no digital reverbs available...
Click to expand...
Well, what about the Lexicon 224, Sony DRE 2000, AMS RMX 16, Quantec QRS, Eventide SP 2016, Yamaha Rev1 and Roland SRV-2000, as a few examples?
I think the first ones I've seen live were the PCM70 and the Sony.
 
Then I got something wrong, because I remember Jackson wrote this in the liner notes.
But maybe the reverbs mentioned could only be implemented in an analog environment
(back in 1984) or did not have the matching bit-depth or sampling ferquency to the 3M Recorder...?
 
greenpsycho said:
I have a bit of an odd question pertaining to the PS Audio DS DAC - I have the Junior, and I always liked it, while at the same time I realize it may not be "hi fidelity" but at the time I purchased it, it was one of few out there with networked roon integration and all the other features I needed.

Anyways, after this review, I picked up a balanced Grace SDAC for comparison. I volume matched them and am using an XLR switcher fed into my headamp (bryston BHA-1). This allowed me to seamlessly switch between the two to really call out the differences. The one odd thing I noticed is that the PS Audio dac is slow. Yes, its delayed compared to the grace dac (not much, but some), but its also playing at a slightly slower speed. Its not always apparent, but I noticed it on a track by Daft Punk and I'm honestly a bit stumped. It was slight, but definitely there and noticeable in parts with fast beats. I wish I had a way to measure or record what I'm hearing, but I don't

Does anyone have any thoughts or clues as to why this might happen?
Click to expand...

Differences in the clock, but it would have to be extreme to be audible.

A simple way to check it would be with a smartphone if you have one. Get one of the free tuner apps people use for guitars and such. Play a 440 hz tone thru both devices and let the phone check the tuning. This would give you an idea of the relative speed between the two devices. The phone should be consistent in speed even if not accurate.

You could also do like someone suggested, record on a phone, a long song or even a 440 hz tone. See the length you get with each device and see the difference in speed.

You also could record higher frequency tones and measure down to the few ppm range. But that is more complex (and dangerous to tweeters) so I would suggest trying the above methods first.
 
mkawa said:
i don't understand -- how can one ship a product like this in good conscience?
Click to expand...

I suppose the short answer to your question is ‘subjectivity’.

There are many out there who have trained their ears to actually appreciate the effects of some amount of distortion. If the customers of these companies enjoy the coloration these DAC units add to their music - on top of the far greater coloration that their amplification and speakers/headphones already provide - and are willing to pay $5K, $10K or $15K to experience it, who are we to say they’re wrong?

In some measure of fairness to these companies, even when Amir does his listening evaluations, very rarely does he say that one of these high dollar, poorly measuring units sounds horrible. And that’s coming from one who, like many of us, has an ear that’s trained to appreciate transparency. Different strokes for different folks, I guess. These companies will never be seeing any of my money, but I don’t think it’s fair to insinuate that they’re necessarily unethical.
 
Bronco said:
I suppose the short answer to your question is ‘subjectivity’.

There are many out there who have trained their ears to actually appreciate the effects of some amount of distortion. If the customers of these companies enjoy the coloration these DAC units add to their music - on top of the far greater coloration that their amplification and speakers/headphones already provide - and are willing to pay $5K, $10K or $15K to experience it, who are we to say they’re wrong?

In some measure of fairness to these companies, even when Amir does his listening evaluations, very rarely does he say that one of these high dollar, poorly measuring units sounds horrible. And that’s coming from one who, like many of us, has an ear that’s trained to appreciate transparency. Different strokes for different folks, I guess. These companies will never be seeing any of my money, but I don’t think it’s fair to insinuate that they’re necessarily unethical.
Click to expand...

This should provide a clue why some manufacturers provide few, or no, specifications. A product not meeting specs could be legitimately returned. Many products not meeting specs could result in class action.

So no or limited specs should be a very strong indication that a product mightn't be as good as the normative description (which is purely subjective and easily defended legally) might indicate.
 
OK, one more measurement :) to settle an argument Ted is making that the DS DAC resolves to 20 bits/-120 dB. He shows such an output using Picoscope. That display shows that the output is dwarfed by spurious tones but his graph is very hard to read. So I thought I replicate it using my audio analyzer.

Here is how the DS DAC performs with a -120 dB signal:

PS Audio PerfectWave DS DAC 1 Khz -120 dB FFT Audio Measurements.png


I bet if I did not point out where the 1 kHz input tone is, you would never find it! There are so many spurious tones including a bunch from the power supply that are much louder than the signal itself. This, despite the fact that I used Toslink for input (i.e. fully electrically isolated) and XLR Balanced output (which should be immune to ground currents/noise).

The FFT I use has 256,000 points so lowers the noise floor so don't think that the actual noise floor is that. It is not.

Despite the very low level of our signal, my analyzer is so sensitive it is able to clearly show harmonic distortion, with 5th harmonic being the highest.

No way can anyone with a straight face say this DAC is resolving to 20 bits/-120 dB signal. We don't "watch FFTs." We listen to music and that music is absolutely polluted by distortion and noise well above our DAC.

You want to see how good this can get in a properly engineered DAC? Here is the Matrix Sabre-X MQA Pro again with identical test:

Matrix Audio Sabre-X MQA Pro DAC 1 Khz -120 dB FFT Audio Measurements.png


This is stunningly cleaner! Power supply noise is now below -150 dB and whopping 30 dB or more lower than our signal itself. Harmonic distortion, if you can even spot it, is just the third-harmonic at incredible -165 dB! Our best case hearing dynamic range is 116 dB so you can't hear any of these artifacts. (And can only hear the main tone if you play using 110 dB of amplification.)

Noise floor now (with the addition of FFT gain) is so low it actually falls off the bottom of the graph!

Let's put these side-by-side so the comparison is more clear: (clock on image for larger size)

Matrix Audio Sabre-X MQA Pro DAC 1 Khz -120 dB vs PS Audio DS DAC.png


How could anyone call themselves "audiophile" yet want to take the DAC on the left versus the one on the right?

Conclusions
It is pure marketing nonsense to say that PS Audio PerfectWave DirectStream (DS) DAC can resolve to -120 dB and hence has 20 bits of resolution. It does not whatsoever. It reproduces that signal at wrong level (despite the averaging in FFT), surrounds it with tons of spurious tones that are even higher amplitude at it, demonstrates non-linearities in the form of harmonic distortion and has a ton of noise to boot.

And remember, this is at 1 kHz where the DS DAC has its lowest distortion. The highest amplitude signals in music tend to be pretty much in lower frequencies where the DS DAC has worst performance.
 
amirm said:
OK, one more measurement :) to settle an argument Ted is making that the DS DAC resolves to 20 bits/-120 dB. He shows such an output using Picoscope. That display shows that the output is dwarfed by spurious tones but his graph is very hard to read. So I thought I replicate it using my audio analyzer.

Here is how the DS DAC performs with a -120 dB signal:

View attachment 34598

I bet if I did not point out where the 1 kHz input tone is, you would never find it! There are so many spurious tones including a bunch from the power supply that are much louder than the signal itself. This, despite the fact that I used Toslink for input (i.e. fully electrically isolated) and XLR Balanced output (which should be immune to ground currents/noise).

The FFT I use has 256,000 points so lowers the noise floor so don't think that the actual noise floor is that. It is not.

Despite the very low level of our signal, my analyzer is so sensitive it is able to clearly show harmonic distortion, with 5th harmonic being the highest.

No way can anyone with a straight face say this DAC is resolving to 20 bits/-120 dB signal. We don't "watch FFTs." We listen to music and that music is absolutely polluted by distortion and noise well above our DAC.

You want to see how good this can get in a properly engineered DAC? Here is the Matrix Sabre-X MQA Pro again with identical test:

View attachment 34600

This is stunningly cleaner! Power supply noise is now below -150 dB and whopping 30 dB or more lower than our signal itself. Harmonic distortion, if you can even spot it, is just the third-harmonic at incredible -165 dB! Our best case hearing dynamic range is 116 dB so you can't hear any of these artifacts. (And can only hear the main tone if you play using 110 dB of amplification.)

Noise floor now (with the addition of FFT gain) is so low it actually falls off the bottom of the graph!

Let's put these side-by-side so the comparison is more clear: (clock on image for larger size)

View attachment 34601

How could anyone call themselves "audiophile" yet want to take the DAC on the left versus the one on the right?

Conclusions
It is pure marketing nonsense to say that PS Audio PerfectWave DirectStream (DS) DAC can resolve to -120 dB and hence has 20 bits of resolution. It does not whatsoever. It reproduces that signal at wrong level (despite the averaging in FFT), surrounds it with tons of spurious tones that are even higher amplitude at it, demonstrates non-linearities in the form of harmonic distortion and has a ton of noise to boot.

And remember, this is at 1 kHz where the DS DAC has its lowest distortion. The highest amplitude signals in music tend to be pretty much in lower frequencies where the DS DAC has worst performance.
Click to expand...

This is possibly is why Ted claimed (elsewhere) that the ear can filter white noise... perhaps he knew what would be found?

There is so much noise across the audio spectrum at -140 dB that it resembles white noise. But that distracts one from the noise that rises to potentially audible levels elsewhere. :rolleyes:
 
Last edited:
@amirm Can AP record a WAV file of a measurement? Sine recorded at -100dB could be more illustrative as "some" graphs :) But tests on such low volumes are good only if one is using DAC as a preamplifier.
 
miero said:
@amirm Can AP record a WAV file of a measurement? Sine recorded at -100dB could be more illustrative as "some" graphs :) But tests on such low volumes are good only if one is using DAC as a preamplifier.
Click to expand...
Sure.

They are both in .wav format so dropbox can play them in place. I can hear both on my Topping DX3 Pro at max volume using my headphones:


PS Audio DS DAC: https://www.dropbox.com/s/2w45f5tb3...0 dB Toslink XLR_09_27_2019_11_51_06.wav?dl=0

Matrix Sabre-X MQA Pro: https://www.dropbox.com/s/vyckudyrshpl5ie/Matrix Sabre-X MQA Pro DAC -120 dB Toslink XLR.wav?dl=0
 
Is there a dB noise reduction equivalent for 256k points?
 
Yes. Here is the amount of "FFT gain" (measurement noise reduction) for different number of audio samples used:

32K = 42 dB
256K = 51 dB
1 million = 57 dB

The actual dB is a few dB different since it also includes the effect of the FFT Window. But is not material in grand scheme of things.
 
amirm said:
Yes. Here is the amount of "FFT gain" (measurement noise reduction) for different number of audio samples used:

32K = 42 dB
256K = 51 dB
1 million = 57 dB

The actual dB is a few dB different since it also includes the effect of the FFT Window. But is not material in grand scheme of things.
Click to expand...

So the -120dB fundamental is buried about 30dB under the actual noise floor and masked by relatively massive mains leaks. Nice.
 
You must log in or register to reply here.

Similar threads

amirm
  • Poll
Fosi Audio ZD3 Balanced DAC Review
8 9 10
Replies
181
Views
41K
TonyJZX
T
amirm
  • Poll
NEOHIPO DC10 Stereo DAC Review
2 3 4
Replies
75
Views
22K
Daviede
D
M
Custom test for DAC performance
Replies
19
Views
1K
mike7877
M
E
DACs with transformer isolated co-axial SPDIF input?
Replies
6
Views
701
Randyman...
Randyman...
R
Measurements of a $5 DAC
3 4 5
Replies
88
Views
16K
bachatero
bachatero
Back
Top Bottom