OP
- Thread Starter
- #781
Thanks. Let's see how many in PS Audio forum vote for the red as their preferred, better sound.
-
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
- Thread starter amirm
- Start date
Thanks. Let's see how many in PS Audio forum vote for the red as their preferred, better sound.
Red is certainly much more exciting
Here is a $600 Apogee Element 24 plotted against Matrix (Apogee=red, Matrix=blue). Looks nearly identical, so it can't be that hard to get this right?
OP
- Thread Starter
- #783
Nice to see. These boutique DAC designers need to wake up and take precision seriously.....
- Joined
- Feb 23, 2016
- Messages
- 20,701
- Likes
- 37,442
These guys like Ted trust their ears don't they? hehehehehe.
You want real comedy, when you use Pkane's excellent Deltawave software, you get to add gain to a signal until you hear it. So in those charts he plots you can listen to the files with about 85 db of gain. One you hear a pretty clean 1 khz tone with just a bit of noise mixed in sort of like tape hiss. In one you hear substantial noise, with some hum, and if you keep listening very closely there are hints of the tone in there from time to time. To defend the noisy one vs the clean one is comically ridiculous. Someone with access to the other forum need to post these two files with the boost already in and let people listen to them.
You want real comedy, when you use Pkane's excellent Deltawave software, you get to add gain to a signal until you hear it. So in those charts he plots you can listen to the files with about 85 db of gain. One you hear a pretty clean 1 khz tone with just a bit of noise mixed in sort of like tape hiss. In one you hear substantial noise, with some hum, and if you keep listening very closely there are hints of the tone in there from time to time. To defend the noisy one vs the clean one is comically ridiculous. Someone with access to the other forum need to post these two files with the boost already in and let people listen to them.
Thanks !
(And sorry for not answering. It was 2 AM here when you asked which measurement...)
Now, you killed all hope of a decent performance !
That probably means that the transformer, if that's the cause of this, was really added with no other goal than modify the sound, since that's not even to balance the signal that they included it.
I've read they use the transformer in the reconstruction filter, that they've made passive, with no opamp...
Last edited:
GrimSurfer
Major Contributor
- Joined
- May 25, 2019
- Messages
- 1,238
- Likes
- 1,484
It will never happen, but some of us have been honored by having screenshots of our posts featured on the PS Audio Forum:
https://forum.psaudio.com/t/another-review-of-the-ds-dac/13027/357
While flattered by the attention, I think they missed my best satirical pieces.
Totoro
Member
It’s not XOR right? So could he any combination.Incompetence, arrogance, or criminality? Which to choose from?
Measure more of these products. Especially the power amp and preamp.
That’s certainly impossible to achieve (as measurements are relevant) and I trust capable designers have something else to do.
Point is more if there is a one on one relationship between exceptional measures (resp. poor) with the perceived quality of the sound.
Sometimes, I cannot find that relationship to be aligned with my own experience / ears.
ASR allowed me to discover superb DAC’s I may have ignored otherwise. I bought them, listen to them and I like them very much: Sabaj D5, E1DA.
Some products have very good measurements, but I have technical issues with them (Topping NX4).
There are some products with superb measurements, which I like very much, who are regularly suffering from negative comments because of what the designer says (Chord). Some others are considered as fantastic because of measures, functions and rigor of the designer (RME). I have a Chord Hugo2 and a RME ADI2. Even after having read all comments in the forum, I still enjoy more H2 + Rupert Neve Amp versus RME ADI2 when used with headphones. Impossible to hear any difference when I focus on it, but when I listen I have just more fun with one of the 2.
I am owning a TotalDAC. Whatever measures, strange behavior of the designer and horrible comments here, I am not complaining with the sound. With ASR I discovered the Matrix Element X. Fantastic! (Not so much the app nor the configuration process. But works perfectly with Roon). They are now on my main music system and I am spending more and more time with the Element X. I cannot say however that thanks to its fabulous measurements I am rediscovering my records vs TotalDAC.
Finally, during the past years I have regularly listened to PS Audio DAC and I almost always didn’t like what I’ve heard. So some years back, the choice for TotalDAC. Thanks to ASR, I may now understand why I was missing something when listening to PS Audio.
But it also means that negative measures don't imply the same when it comes to sound. At least in my personal experience.
1. My point is basically what you said. Sound = measurements, more or less.
2. You buy way too many DACs for someone not running a review site. Seek professional help.
3. Embrace the truth and buy speakers with digital inputs and integrated DACs. Problem solved!
BillG
Major Contributor
We're largely immune to hearing distortion up to a point - perhaps you never reached that point...
We're largely immune to hearing distortion up to a point - perhaps you never reached that point...
I have had as well a lot of negative experiences with all kind of audio stuff, be it loudspeakers, amps, or sources. I would not consider that I have exceptional ears. But it happens frequently that I am really suffering hearing some audio stuff I don’t like - so maybe I am not fully immune from distortion at the end
I have a TotalDAC (bad measures) and I also had the opportunity to compare to PS Audio (bad measures) as well as Element X, RME (excellent measures), and the perception of the negative mesures is very different:
- I really didn’t like the PS Audio as I felt something was missing. And Amir is explaining why with the measures
- I cannot see a highly superior sound coming out Element X vs TotalDAC, so the very different measures are here not translated into different quality of sound - at least to my ears.
I was not convinced some years back, but now it is much more clear to me that differences in sound are more coming from headphones or loudspeakers than the source (provided you don’t pick a crappy one). And I am also considering what @beefkabob is saying: future may be active speakers.
Last edited:
I think Rupert Neve never made a secret of the fact he liked to have an amp adding some harmonics, which adds some "body" to the sound.
Kind of a personal signature.
There is nothing wrong with that.
That's what all music instruments do.
In ideal world, though, when it's about the reproduction system, it should be possible for the end user to add this "salt" to taste.
Kind of what you see in the device I measured in the attached link (by the way, I think Focusrite was initially Neve's brand)
https://www.audiosciencereview.com/...hd-in-the-perspective.9085/page-3#post-232575
2. You buy way too many DACs for someone not running a review site. Seek professional help.
I will consider ; -) In fact, the point is that now I have a lot of friends happy having these DAC at their place.
It is not very easy to say. Large part of the performance is probably due to analog stages and maybe partially due to DSP. But it is not very easy to say from this data what kind of split there is. And not really possible to test because there's no direct path to the conversion stage.
I personally very much enjoy my T+A DAC8 DSD (now running at DSD256, earlier before EC modulators at DSD512). Also Holo Audio DACs (Spring, Cyan DSD and Spring 2, all running at DSD256). RME ADI-2 Pro's the normal and AE versions running in DSD Direct mode at DSD256 (massively more accurate than with PCM input). And now recently also the EVGA NU Audio card running at DSD256 (also massively more accurate than with PCM input).
But just one example where it matters how the DAC is driven from DSP is Holo Spring 2 with PCM inputs...
Here's 1k -120 dB input TPDF dithered at 1411.2/24:
Low level linearity error of the R2R ladder creates distortion.
However, if we run the same, but within the ladders' linear range and use noise shaping to improve the linearity further, same 1k -120 dB input but now fifth order noise shaped at 1411.2/20:
We get quite clean low noise output!
We can also see how the DAC behaves when we drop down to 16-bit, keep the same rate with two different noise shapers...
With same fifth order noise shaper we begin to see a little bit of increase in noise floor after 60 kHz:
If we then switch to ninth order noise shaper:
And we close to fifth-order at 20-bit results again...
Last edited:
About the "benefit" of adding distortion:
It also depends of the frequency profile of distortion:
Adding heavy harmonics at 40Hz will make the bass more present, and especially on small loudspeaker that could hardly reproduce 40Hz, but will happily play 80Hz or 120Hz. This adds to the bass level significantly, so one may perfectly like the sound better.
Doing the same at 500Hz or 1kHz may be very different, since it will add distortion in frequency range where our hear is most sensitive and we may find that disturbing and more easily identify it as it is: distortion.
Those "high end" audio need to differentiate from competition. They probably do that by adding their "signature", be it distortion or other things.
Obviously, they won't measure perfectly.
But measurements, in my opinion, should still be showing SOTA design on which some sauce has been added. Like the Focusrite in my previous post, which is excellent when used flat.
Not the kind of garbage we see in some.
It also depends of the frequency profile of distortion:
Adding heavy harmonics at 40Hz will make the bass more present, and especially on small loudspeaker that could hardly reproduce 40Hz, but will happily play 80Hz or 120Hz. This adds to the bass level significantly, so one may perfectly like the sound better.
Doing the same at 500Hz or 1kHz may be very different, since it will add distortion in frequency range where our hear is most sensitive and we may find that disturbing and more easily identify it as it is: distortion.
Those "high end" audio need to differentiate from competition. They probably do that by adding their "signature", be it distortion or other things.
Obviously, they won't measure perfectly.
But measurements, in my opinion, should still be showing SOTA design on which some sauce has been added. Like the Focusrite in my previous post, which is excellent when used flat.
Not the kind of garbage we see in some.
Last edited:
It will never happen, but some of us have been honored by having screenshots of our posts featured on the PS Audio Forum:
https://forum.psaudio.com/t/another-review-of-the-ds-dac/13027/357
While flattered by the attention, I think they missed my best satirical pieces.
This one is a gem:
I’ve sat on the sidelines of this thread and have become progressively more annoyed and disappointed.
There is no difference between Amir’s approach to reviews and beating a Piñata. We all know the outcome of his reviews but we don’t know how many times it needs to bashed.
I’ve visited his site and there seems to be an excessive amount of slagging products. His fanboi’s routinely pile on in a feeding frenzy.
You need to go out of your way to create a cesspool of negativity that’s his modus operandi and clearly reflects his personality.
What I don’t like is the apparent guise of science to routinely damage hard earned reputations of many excellent companies and their products. I doubt that informed audiophiles pay much credence to his approach and clever smoke and mirrors to further some perverse agenda.
For him, it’s about cool hard cash or a least a bribe guised as a review sample. There are reasons why no manufacturer would offer up a device for his agendas. This explains why he routinely has to travel for reviews.
What ever happened to the main reason why we enjoy this hobby…the reproduction of music that approaches the ultimate goal…a live performance.
Measurements in themselves are a guideline and they tell me nothing about how the device sounds.
Unfortunately he doesn't measure the reconstruction accuracy at all.
A bit off-topic for this thread, but @amirm can probably move this ADI-2 related post to a correct place if necessary. Anyway, Matrix DAC has been compared on this thread several times already, so maybe it's not too bad to discuss comparisons.
At 44.1/32 PCM input we can see that first image around 352.8k is at about -52 dB level or so, so the reconstruction accuracy is equivalent of about 9-bit resolution. This is "stair stepped distortion" telling about imperfect reconstruction of the analog waveform.
With same 44.1/32 source data, but converted to DSD256 (ASDM5 modulator, DSD filter set to 50 kHz) input images are not visible anymore (I know they are below -192 dB now). Some uncorrelated noise shaping noise left at about -85 dB, 30+ dB difference, and now uncorrelated. Wide band noise level is now equivalent of 14-bit resolution.
In audio band, when levels are matched (set PCM volume control to -3.5 dB and then output volumes are equivalent, this difference is due to DSD spec allowing short term +3.15 dBFS peaks), the performance is very similar, with DSD256 giving a little bit shorter harmonic series and 2 kHz element from 19+20k IMD test disappearing. Disappearance of the 2 kHz is precisely due to lack of those images (intermodulation between negative and positive image).
It would only be easy if the measurements don't correlate to what's audible. Is that what you are suggesting?
Similar threads
