PS Audio DirectStream DAC Mk.2 measurements

[AdrianusG]

Busy Saturday measuring the "NEW" improved Massive firmware (in red below).
We didn't like the sound of this new FPGA firmware. We prefer the previous beta version (in blue below) even thou they claim the new one has "massive" improvements.

FPGA firmware updates:
ANTERO 208 (3/2023)
MASSIVE 228 beta (6/2023)
MASSIVE 233 final (1/2024)

View attachment 345928
On a positive note. The user interface update has fixed many flaws and missing promised features from past versions and now allows us to turn off the power to the WiFi chip located inside the DAC. We don't want active WiFi noise in our DAC. This WiFi feature does not allow useful streaming or HTTP user settings. It was put in only for over the internet updates which can be done via the USB port. The over the internet update feature is still not active.

Oh come on now, you can't have it all for just 8000$ right!
you'll probably get that with one of the next even more massive updates, patience

 

[ahofer]

Oh come on now, you can't have it all for just 8000$ right!
you'll probably get that with one of the next even more massive updates, patience

Wait for the Mount Hyperbole update.

 

[WRWSTD]

Has anyone seen any measurements of the recent Mount Blue Sky update?

 

[Matias]

Has anyone seen any measurements of the recent Mount Blue Sky update?

It's a hopeless cause IMO. I am skeptical and will wait them to prove me wrong.

 

[Billy Budapest]

Busy Saturday measuring the "NEW" improved Massive firmware (in red below).
We didn't like the sound of this new FPGA firmware. We prefer the previous beta version (in blue below) even thou they claim the new one has "massive" improvements.

FPGA firmware updates:
ANTERO 208 (3/2023)
MASSIVE 228 beta (6/2023)
MASSIVE 233 final (1/2024)

View attachment 345928
On a positive note. The user interface update has fixed many flaws and missing promised features from past versions and now allows us to turn off the power to the WiFi chip located inside the DAC. We don't want active WiFi noise in our DAC. This WiFi feature does not allow useful streaming or HTTP user settings. It was put in only for over the internet updates which can be done via the USB port. The over the internet update feature is still not active.

Are you sure the SOUND is any different? If you look at your own frequency response measurements, all three firmware versions track closely until the ultrasonic region which is inaudible in any case. Plus, all perform similarly crappily in the bottom octave, although some add dB while others minus dB. I guess that could be audible, barely.

 

[mdsimon2]

Has anyone seen any measurements of the recent Mount Blue Sky update?

Source = https://forum.psaudio.com/t/dsd-mk2-latest-mountaintop-update-mount-blue-sky/40152/292.

I assume the slightly lower noise one is Blue Sky. Distortion is still high, as is mains hum. Also need to be very careful comparing FFTs with no parameters given. I wish they stated a quantitative improvement in dynamic range or residual noise as the FFT can be visually gamed with changes to averaging and FFT length. For example, the 3 FFTs below are from the same DAC and all indicate the same dynamic range / residual noise but use different FFT lengths (16K vs 128K vs 1M), but you could easily misinterpret the results to say the noise had decreased.

Personally, I find it hilarious they are claiming sound improvements such as a "blacker background" based on lower noise and backing it up with measurements. Imagine how black the background would be if they were listening to a DAC with an objectively low noise floor!

Michael

 

[617]

 

[jkr]

View attachment 418170

Source = https://forum.psaudio.com/t/dsd-mk2-latest-mountaintop-update-mount-blue-sky/40152/292.

I assume the slightly lower noise one is Blue Sky. Distortion is still high, as is mains hum. Also need to be very careful comparing FFTs with no parameters given. I wish they stated a quantitative improvement in dynamic range or residual noise as the FFT can be visually gamed with changes to averaging and FFT length. For example, the 3 FFTs below are from the same DAC and all indicate the same dynamic range / residual noise but use different FFT lengths (16K vs 128K vs 1M), but you could easily misinterpret the results to say the noise had decreased.

View attachment 418175

Personally, I find it hilarious they are claiming sound improvements such as a "blacker background" based on lower noise and backing it up with measurements. Imagine how black the background would be if they were listening to a DAC with an objectively low noise floor!

Michael

Suspect...

These measurements do not match what we measured with our AP. Wonder why PS Audio did not use their AP to do the measurements???

 

[ObjectiveSubjectivist]

Imagine how black the background would be if they were listening to a DAC with an objectively low noise floor!

Exactly.
If now they back their subjective impressions with measurements, please someone post on their forum a 200usd dac that do circles around their dac in terms of noise and distortion and ask how is the detail, soundstage bla bla bla with it...

I wonder how many irrational/illogical comments would show.

 

[ahofer]

I wonder how many irrational/illogical comments would show.

I don’t anymore. A lot.

 

[mdsimon2]

I find it fascinating they quickly killed the DS Mk2 and Paul is openly admitting the underlying architecture is flawed, and nothing can really be done to improve the noise performance.

Will the DS MKII ever see more development? Is this something Ted could do on his own as paid updates?

Correct and for the ten years or so we built that platform that was the case and produced a world class beautiful sounding DAC. But, as you know, progress moves on. Always will. After more than a decade of supporting that platform we felt it had reached its maximum potential. Indeed, there was...

forum.psaudio.com

Correct and for the ten years or so we built that platform that was the case and produced a world class beautiful sounding DAC. But, as you know, progress moves on. Always will.

After more than a decade of supporting that platform we felt it had reached its maximum potential. Indeed, there was more room on the FPGA to have offered some additional sonic magic, but the FPGA wasn’t what was limiting the technology’s potential. It was the follow on hardware configuration.

In this design chain, the output lvds of the FPGA went through a series of high speed video op-amps to increase its signal level, through a low pass filter, and then out through an audio transformer.

The signals themselves, based on lvds technology, were very low in level and once amplified had inherent noise which resulted in relatively poor measurement results in the -80dB range.

That’s when we reached a crossroads and chose to investigate an entirely new path, one still based on the core technology of a DSD stream converted directly into audio. The divergence is in what happens after the DSD signal is generated. As mentioned, in the DS MKII chain, the lvds output was further amplified and then LP filtered through an output transformer. In the new PMG Series, we take a very different path, one that not only sounds better but measured 20dB better as well.

The new path was quite a science project but worth the improvement.

Michael

 

[HalSF]

I found this thread to be extremely calming and soothing after just now reading a thread at the Steve Hoffman Forum started by a guy who was asking for advice about buying a “no-nonsense DAC,” and who, following several pages of high-end upselling and sneering at inexpensive DACs like the WiiM DAC the guy already owned, announced that he had bought …the PS Audio DirectStream MK2.