CHANGSTAR: Audiophile Headphone Reviews and Early 90s Style BBS

Lobby => Amp and DAC Measurements => Topic started by: Hands on May 23, 2014, 11:39:30 AM

Title: PS Audio PerfectWave DAC MKI Measurements (FW 2.0.2 vs 2.4.4)
Post by: Hands on May 23, 2014, 11:39:30 AM
Test Hardware and Setup

- PS Audio PWD MKI - SPDIF coax input from JKSPDIF, also did some brief USB tests. Took measurements of FWs 2.0.2 and 2.4.4. Remember, MK1, not MK2. I couldn't get USB to work on 2.0.2, so those are all 2.4.4 measurements. Assuming SPDIF if it doesn't specify.

- DIY Gaming Desktop - i5 2500K locked at 4.5GHz, Biostar TZ77XE3 motherboard, GTX 780, Antec EarthWatts 650W Green PSU, Auzentech Bravura sound card, 4 140mm fans, goofy UV cathode lights in case, Fractal Design Define R4 case...overall, most likely not optimal for audio, but does well enough as an output in tests and represents a real-life scenario for many. Plugged into APC H10 power conditioner. I did run Fidelizer on it for the tests. Used USB 3.0 ports from the mobo, tested for which ones produce best results and use those. Generally does not cause issues.

- Creative X-FI HD USB (SB1240) - Takes line-out signal from DAC to line-in for recording purposes, powered by USB bus, plugged into Surface Pro. Unit is limited to 48/96KHz input/output.

-1st gen Microsoft Surface Pro - Records from SB1240 via USB. Running Fidelizer tweaks. Battery powered. Screen is touchable and thus is a more spiritually fulfilling experience. Produces cleaner and generally better results than running the SB1240 from the desktop as well.

Software

RightMark RMAA 6.3.0, ARTA

Misc. Notes

Creative SB1240 has a slight roll-off in the bass.

I do not guarantee 100% accuracy with these measurements. Good DACs can fairly easily max out the capabilities of my SB1240.

Links

Creative SB1240 - http://us.creative.com/p/sound-blaster/sound-blaster-digital-music-premium-hd (http://us.creative.com/p/sound-blaster/sound-blaster-digital-music-premium-hd)
RightMark Audio Analyzer - http://audio.rightmark.org/index_new.shtml (http://audio.rightmark.org/index_new.shtml)
ARTA: http://www.artalabs.hr/ (http://www.artalabs.hr/)
Title: PWD MK1 16/48 RightMark Measurements - 2.0.2 vs 2.4.4 - Filter 1
Post by: Hands on May 23, 2014, 11:50:06 AM
16/48 RightMark results using native mode and filter 1, comparing FWs 2.0.2 (green) and 2.4.4 (white). Both results exhibit a very slight bump in the FR above 10KHz. Some very slight differences between the two. Overall, quite good measurements.

TestFW 2.4.4 (White)FW 2.0.2 (Green)
Frequency response (from 40 Hz to 15 kHz), dB   +0.03, -0.07+0.03, -0.07
Noise level, dB (A)   -97.4-97.3
Dynamic range, dB (A)   97.397.3
THD, %   0.00390.0038
IMD + Noise, %   0.00630.0063
Stereo crosstalk, dB   -95.7-96.1

Attached pictures in order are:

1. Frequency response
2. Noise
3. Dynamic Range
4. THD
5. IMD
6. Crosstalk
Title: PWD MK1 16/48 RightMark Measurements - 2.0.2 vs 2.4.4 - Filter 3
Post by: Hands on May 23, 2014, 11:54:38 AM
16/48 RightMark results using native mode and filter 3, comparing FWs 2.0.2 (green) and 2.4.4 (white). Measurements are still good overall. Some of the slight differences in the last measurements are still present here. Some differences are within the normal margin of error.

TestFW 2.4.4 (White)FW 2.0.2 (Green)
Frequency response (from 40 Hz to 15 kHz), dB   +0.03, -0.07+0.03, -0.07
Noise level, dB (A)   -97.3-97.4
Dynamic range, dB (A)   97.397.3
THD, %   0.00390.0039
IMD + Noise, %   0.00600.0060
Stereo crosstalk, dB   -96.1-95.7
Title: PWD MK1 24/96 RightMark Measurements - 2.0.2 vs 2.4.4 - Filter 3
Post by: Hands on May 23, 2014, 11:59:40 AM
24/96 RightMark results using native mode and filter 3, comparing FWs 2.0.2 (green) and 2.4.4 (white). Overall results are good. The 24/96 measurements give us a better view at the slight differences between the firmwares. Let's see how other measurements compare.

TestFW 2.4.4 (White)FW 2.0.2 (Green)
Frequency response (from 40 Hz to 15 kHz), dB   +0.02, -0.08+0.02, -0.08
Noise level, dB (A)   -105.0-104.6
Dynamic range, dB (A)   104.9104.5
THD, %   0.00390.0040
IMD + Noise, %   0.00540.0049
Stereo crosstalk, dB   -99.0-98.6
Title: PWD MKI Jitter Measurements (FW 2.0.2 vs 2.4.4)
Post by: Hands on May 23, 2014, 12:06:56 PM
24/48 JTest (jitter) measurements. Images attached are as follows (SPDIF unless specified otherwise):

1. FW 2.0.2 - Filter 1, Native mode
2. FW 2.0.2 - Filter 3, Native mode
3. FW 2.4.4 - Filter 1, Native mode
4. FW 2.4.4 - Filter 3, Native mode
5. FW 2.4.4 - Filter 1, PWD upampled to 192KHz
6. FW 2.4.4 - USB input, Filter 3, Native mode
7. FW 2.4.4 - USB input, Filter 3, PWD upsampled to 48KHz
Title: PWD MKI -90.31dB Measurements (FW 2.0.2 vs 2.4.4)
Post by: Hands on May 23, 2014, 12:14:35 PM
Low-level, -90.31dB 1KHz sine wave tests. Files attached are as follows:

1. FW 2.0.2 - Filter 3, native mode, 16/48
2. FW 2.0.2 - Filter 3, native mode, 24/48
3. FW 2.4.4 - Filter 3, native mode, 16/48
4. FW 2.4.4 - Filter 3, native mode, 24/48
5. FW 2.4.4 - USB, native mode, 16/48
6. FW 2.4.4 - USB, 48KHz upsampled, 16/48
7. FW 2.4.4 - USB, native mode, 24/48
Title: PWD MKI 19KHz Measurements (FW 2.0.2 vs 2.4.4)
Post by: Hands on May 23, 2014, 12:19:29 PM
Single sine wave at 19KHz, at both 0dB and -3dB. Images attached are as follows:

1. FW 2.0.2 - Filter 1, native mode
2. FW 2.0.2 - Filter 3, native mode
3. FW 2.4.4 - Filter 3, native mode
4. FW 2.4.4 - Filter 3, 48KHz upsampled
Title: PWD MKI 13/14KHz Measurements (FW 2.0.2 vs 2.4.4)
Post by: Hands on May 23, 2014, 12:22:46 PM
13/14KHz measurements, at both 0dB and -3dB.

1. FW 2.0.2 - Filter 1, native mode
2. FW 2.0.2 - Filter 3, native mode
3. FW 2.4.4 - Filter 3, native mode
4. FW 2.4.4 - Filter 3, 48KHz upsampled
Title: PWD MKI Square Wave Measurements (FW 2.0.2)
Post by: Hands on May 23, 2014, 12:35:26 PM
1KHz square wave measurements on FW 2.0.2. For all 5 filters, all native mode, the left side is a -3dB 1Khz 16/48 signal, and the right side is a 24/96 signal. If my measurements are correct, the 24/96 filters are different and yet don't always seem to match their 16/48 counterpart.

PS Audio lists the filters as such in the PWD MK1 manual:

 1. AUTO - This will automatically choose filter MP Soft for 44.1kHz and LP Soft for any
 higher sample rates.
 2. Filter 1 - Minimum phase Apodising filter.
 3. Filter 2 - Minimum phase soft knee filter.
 4. Filter 3 - Linear phase Apodising filter.
 5. Filter 4 - Linear phase soft knee filter.
 6. Filter 5 - Minimum phase recursive Half Band symmetrical filter.

Images are attached in filter order. Didn't measure Auto.
Title: PWD MKI Square Wave Measurements (FW 2.4.4)
Post by: Hands on May 23, 2014, 12:59:58 PM
Square wave measurements for FW 2.4.4. I was wondering if other FWs might have implemented different digital filters, but that doesn't appear to be the case. Same as before with the first 5 filters posted in order (native mode). The last two attachments show what happens to filter 2 when you upsample to 48KHz and 96KHz on the PWD (what happens at 192KHz is a fairly obvious continuation).
Title: Re: PS Audio PerfectWave DAC MKI Measurements (FW 2.0.2 vs 2.4.4)
Post by: firev1 on May 23, 2014, 01:15:13 PM
The PWD jitter measurements look a little suspect to me, after all the talk about XMOS, I'm pretty surprised that jitter would still show up at the outputs. It also shows up in the regular measurements. Pretty surprising how firmware would make a measureable difference.
Title: Re: PS Audio PerfectWave DAC MKI Measurements (FW 2.0.2 vs 2.4.4)
Post by: Hands on May 23, 2014, 01:20:57 PM
Quote from: firev1 on May 23, 2014, 01:15:13 PM
The PWD jitter measurements look a little suspect to me, after all the talk about XMOS, I'm pretty surprised that jitter would still show up at the outputs. It also shows up in the regular measurements. Pretty surprising how firmware would make a measureable difference.

MK1 does not utilize an XMOS USB solution. I bet MK2 would measure better on all digital inputs too.

I don't know enough about DACs to understand why different firmwares might sound or measure slightly differently. The only thing I could guess they could change with such a small impact on measurements might be something like noise shaping, but you'd think they'd have mentioned that if they actually changed something like that. I really can't give much of an educated guess here. But, really, any code change at all could theoretically slightly alter the D/A process and performance. *shrug*
Title: Re: PS Audio PerfectWave DAC MKI Measurements (FW 2.0.2 vs 2.4.4)
Post by: firev1 on May 23, 2014, 01:51:43 PM
Oh I see, that would explain things , I guess the PWDmk1 does show its age. Yeah, for isochronous usb the code does makes a huge difference. I do remember reading about it once though I would be hard pressed to look for the article again. In a sense, looking at this makes me give respect to Nwavguy as an engineer(or whoever wrote the ODAC's usb code) and the Antelope Zodiac(isochronous) as well. Below is a impromptu measurement(does not follow my usual separate laptops procedure) of the Zodiac's jitter.

Amazing how the low resolution test shows its usefulness and how the 2.4.4 firmware appears to be consistently worse. Looks like purrin and other users are right after all.

Also find it funny that the filter's code is pretty broken in a way, makes it harder to do proper subjective evaluations unless you keep to the same sample rate throughout the auditioning process.

Title: Re: PS Audio PerfectWave DAC MKI Measurements (FW 2.0.2 vs 2.4.4)
Post by: Anaxilus on May 23, 2014, 06:57:14 PM
Thanks for the measurements and all the work!  This certainly confirms a few thing many of us have maintained for awhile.  The PWD via USB always sucked.  Though we don't have XMOS data here, as far as subjective impressions it's better but not best, still behind the other inputs  quite noticeably.

The other observation is that these set of measurements are not sufficient to pinpoint the effects of widely reported subjective firmware differences many of us and many across the country have reported independently from each other.  The potential for placebo and expectation bias is also lower, though never can be eliminated, since all the boxes look the same, they have no glowing tubes or art/woodwork and concern DACs the users have owned for awhile with the newer firmwares often (not always) sounding subjectively worse to owners, both of which work against new toy syndrome as a possibility.
Title: Re: PS Audio PerfectWave DAC MKI Measurements (FW 2.0.2 vs 2.4.4)
Post by: Marvey on May 27, 2014, 01:07:27 AM
FYI, I've always preferred the Filter 1 and Filter 4 on PWD1/2. Any correlation to measurements?
Title: Re: PS Audio PerfectWave DAC MKI Measurements (FW 2.0.2 vs 2.4.4)
Post by: Hands on May 27, 2014, 03:52:10 AM
You know, I should post a zip of all my PWD test files, in case people want to look at specific comparisons. I just had way too much stuff to post it all on here. Even then, trying to get DAC measurements for 5 filters and 2 different FWs, not to mention native/upsampling modes, can be pretty mind numbing! There are some tests I missed, but I should at least have RMAA measurements for the different filters. I'll sort through my files in the next few days...hard part is using a simple naming convention that is easy for others to follow without explanation.

I know the filters won't measure too differently in 16/44 or 48, but 24/96 might better show some differences (keeping in mind measurements indicate the "HD" filters don't always match their "SD" counterpart). Then you have to account for whether or not you were using "HD" or software upsampled music files and/or whether or not you were running native mode or using the PWD's upsampling. While I felt the filters themselves were pretty homogeneous sounding and measuring with "SD" audio files (I don't have much "HD" stuff to test subjectively), there's still a staggering number of combinations you can set up with this DAC.

It's kind of like the measured differences with the two firmwares. Some of the slight differences my measurements show are probably just variance/margin of error. Some differences seem to be more consistent and often take place outside the audible band, but even those are slight enough and of the sort where it's hard to look at them and say, "Ah, this shows why I prefer X over Y." Though, maybe someone more of an expert in the area might be able to look at these slight differences and start to guess what might be the underlying cause.

I think PS Audio was going for fairly consistent performance across the digital filters, compared to something like the Audio-GD Wolfson DACs or, slightly, the Gamma2. But, let me dig through my stuff and see if I can't find some identifiable and consistent differences besides the obvious square waves. :)