NADAC vs. MSB vs. Playback Designs blind listening test

[Keith_W]

Preamble

Since I bought my NADAC MC-8, I had suspected that it does not sound as good as my Playback Designs MPS-5. I had conducted some listening tests in my own system, however these tests were done sighted as I did not have someone else to switch the inputs. My system has now been reconfigured such that an 8 channel DAC is absolutely required, because the crossover is done in the PC. As a result, I am no longer able to compare the 2 channel Playback vs. the 8 channel NADAC.

 

A friend of mine has been curious about the NADAC from the beginning. I arranged to bring both my DAC's to his place for a shootout.

 

There were four participants, all self-confessed audiophiles. Regrettably, a fifth could not make it. Also regrettably, I was unable to borrow a DCS DAC to participate.

 

 

System

The DAC's under test were:

 

- Merging NADAC MC-8, (AES/EBU input) AUD$17,000

- Playback Designs MPS-5 with the latest firmware upgrades, AES/EBU input. AUD$17,500

- MSB Diamond DAC V, with the optional Femto 33 clock, power base, analog volume control, I2S input. AUD$49,000

 

The rest of the system:

 

- MSB UMTV transport

- Tenor 350M monoblock amplifiers

- Vivid Giya G1 speakers

 

As you can guess, this is a highly resolving system built comprised of some of the best components that money can buy. The Giya's are Vivid's flagship speaker. They have a really clean, crisp sound and are highly resolving. Tonality was a little on the bright side, but was never harsh. The bass was adequate but not "in your face" like Wilsons.

 

 

 

 

Method

The testing was done in three parts.

 

In the first part, we played test tones and used the stepped analog volume control on the MSB preamp to volume match the output of the DAC's using a calibrated SPL meter. Because the volume control had 1dB increments, it was not possible to precisely match the volume, but we managed to do well. These were the settings:

 

- MSB DAC - 84.5dB (-20dB on the preamp)

- Playback Designs - 84.4dB (-18dB on the preamp)

- NADAC (in 8 channel mode) - 84.1dB (-16dB on the preamp)

- NADAC (in 2 channel summed mode) - 84.2dB (-16dB on the preamp)

 

In the second part, we did sighted listening to form our subjective opinion on the DAC's under test. We openly discussed our subjective findings and agreed on the sonic character of each DAC.

 

In the third part, the DAC's were covered with a towel, and one of us switched the inputs on the DAC and preamp while the others turned their backs. One track would be played for a minute, followed by a brief interval where the DAC was swapped. This would continue until we had listened to all the DAC's. We were then asked our subjective opinions of the DAC under test (DAC 1, 2, 3, 4). After each of us had our say, the identity of each DAC was revealed. We did this five times with different types of music.

 

 

 

 

IMPRESSIONS

It was quite easy to tell the difference in character between these DAC's. On some pieces of music, it was more difficult. One of the participants found it difficult to hear any difference between any of the DAC's and decided to abstain from voting or making comments.

 

Iteration 1: Bach's Goldberg Variations / Gustav Leonhardt (harpsichord)

This CD was mostly about the midrange and high frequency. It is an excellent recording with the timbre of the harpsichord well recorded and presented. All 3 of us correctly guessed the identity of the NADAC. Two of us guessed correctly the identity of the Playback/MSB, and the third got the identity of the Playback/MSB reversed.

 

Iteration 2: The Tragedy of the Cathars / Jordi Savall / Alia Vox

This CD was recorded in a large acoustic space. It demonstrates decay and soundstage very well. All 3 of us correctly guessed the identity of the NADAC. Two of us guessed correctly the identity of the Playback/MSB, and the third got the identity of the Playback/MSB reversed.

Iteration 3: Copland's Fanfare for the Common Man / Eiji Oue / Reference Recordings

The opening of the CD contains a massive bass drum followed by a wind accompaniment. It tests ability to produce low bass, dynamics, and tonality of the wind instruments. All of us correctly identified the DAC's.

 

Iteration 4: Schubert Kennst du das Land? / Zomer and Schoonderwoerd / Alpha

Zomer is a powerful soprano with a big voice, probably too big for lieder. She has a tendency to overpower the piano and can sound relentless with the wrong DAC. The piano used is a period piano which has a different tonality to modern piano. All of us had difficulty telling the DAC's apart with this track.

 

Iteration 5: Prokofiev Symphony No. 2 Allegro / Valery Gergiev / Decca

A really complex symphony with violins, horns, drums, and gongs. This disc tests the ability to separate the instruments into something intelligible. Two of us correctly picked the NADAC, but all of us had difficulty picking between the MSB and the Playback.

 

Iteration 6: Beethoven Late String Quartets / Takacs Quartet / Decca

A superb recording demonstrating the ability of the DAC to resolve the various timbres of the string instruments. All 3 of us correctly identified each DAC.

 

 

COMMENT

From worst to best, in order:

4. Merging NADAC MC-8, in 8 channel mode

We consistently felt that without channel summation, the NADAC sounded thin with a pronounced glare in the treble that made it very easy to identify. Indeed, it was always the NADAC in 8 channel mode that was identified first.

 

3. Merging NADAC MC-8, in channel summation mode

This is where it gets difficult. With the exception of (4 - i.e. the NADAC without channel summation), which was clearly the inferior sound of the evening, the next three were very close. What a difference channel summation makes. All of us thought that it sounded richer and more engaging with better resolution of detail. It was so close to the Playback that we often confused the sound of the two under blind conditions.

 

2. MSB Diamond DAC V

We felt that the MSB had a tad bit less timbre, a tad bit less attack, and a more laid back sound. I would describe MSB as the most polite sounding in this company.

1. Playback Designs MPS-5

This was the surprise of the evening. The Playback had the best resolution, the best dynamics (sometimes beaten by the NADAC in summation mode on some tracks), the widest soundstage, the best timbre. I am in fact thinking of selling the Playback and deliberately did not invite the potential buyer, because I expected the Playback to be trounced by the MSB and I did not want him to hear it! Now I wish I had. The margin of victory over the NADAC (in channel summation mode) and the MSB was very slim, however and our preference varied from track to track. But in the end, 2 of us felt the Playback was better, and one of us felt the MSB was the best (not the MSB owner).

 

At the end of the listening, we decided to compare the sound of the MPS-5 using its own built-in transport against the MSB with its own transport. Listening was unblinded. The margin grew even greater - if anything, the Playback gained even more transparency and even better dynamics.

 

Sum-up. The NADAC MC-8 with channel summation can clearly hold a candle to some of the best DAC's money can buy, but without it - it sounded so dramatically inferior that all of us could easily hear it. The Playback is the oldest DAC in this group, but still sounded the best. I felt that the MSB was a tiny bit disappointing given the sheer expense of this DAC. If it was priced the same as the Playback and NADAC, it would be very good.

[semente]

Thanks for taking the time to describe it in such detail.

 

Could you briefly explain what channel summation consists of?

 

R

[esldude]

Thank you for doing such a test. Results are interesting. Few have access to such pieces of gear to do such a test.

 

Now please forgive me for being critical of one thing you did. I have put up test files and know how bothersome everyone second guessing you can be.

 

Using a sound level meter to set levels is not an excellent method. The better method is to measure voltages of test tones at the loudspeaker terminals. A simple multi-meter would suffice for that. I would be interested in what tones you used with the SLM. Also did you use A weighting?

 

Couldn't find measures for all the DACs, but looking at specs the MSB should have been over 2 db louder, and you had it 2 db reduced. So that was as close as you could get. However, the other two DACs should have been the same, and you had the NADAC 2 db louder. Meaning if specs are mostly correct, and I haven't misunderstood your volume methodology the MSB and Playback Designs were likely less than a half decibel different in loudness, and the NADAC was about 2 db louder than those.

 

The NADAC being 2 db louder would explain why it was usually discernible as different. The other two being closely matched in loudness would explain why they were less reliably discerned. If I read between the lines of your blind results properly, you collectively picked the correct one between the MSB and Playback Designs 10 out of 18 times.

[matthias]

Thank you for reporting this results. It is interesting that according to David W. Robinson from Positive Feedback the new Merlot DAC seems to outperform the MPS-5, so the Merlot would be the least expensive of the three competitors.

 

Matt

[Priaptor]

Interesting thread. I don't have experience with any but the MSB V which I own. I can say that the preamp stage of the Diamond IV +, pretty much the same as the Diamond V, which I chose to ignore in my purchase, didn't WOW me. My understanding is that the preamp stage of the Select is a whole different level, which I will be trying after the new year.

 

One other point I would make. I tested the UMT and was never crazy about it, but since you tested all the DACs listed, that variable is pretty much fixed in your comparisons. I would add, that not just me but many others, including owners of the Select II who have posted here, have stated that the QUAD USB port moves the MSB to a totally different level.

 

Just my two cents about the MSB which I own and love and have had for over 4 years.

[Keith_W]

Could you briefly explain what channel summation consists of?

 

Hi R, channel summation consists of summing the output of four channels into two. Merging's webpage touts this as an advantage of buying the 8 channel version for 2 channel use.

Now please forgive me for being critical of one thing you did. I have put up test files and know how bothersome everyone second guessing you can be.

Using a sound level meter to set levels is not an excellent method. The better method is to measure voltages of test tones at the loudspeaker terminals. A simple multi-meter would suffice for that. I would be interested in what tones you used with the SLM. Also did you use A weighting?

 

Feedback is welcome! We used test tones from a CD that contained test tones and measured at two frequencies - 200Hz and 1000Hz. Obviously the SPL's at 1000Hz was different to the SPL at 200Hz, but it was similar enough that I did not bother recording it. An oversight, I suppose. We used C-weighting. Unfortunately, we didn't have a multimeter at hand. To be honest I didn't even think of that.

Couldn't find measures for all the DACs, but looking at specs the MSB should have been over 2 db louder, and you had it 2 db reduced. So that was as close as you could get. However, the other two DACs should have been the same, and you had the NADAC 2 db louder. Meaning if specs are mostly correct, and I haven't misunderstood your volume methodology the MSB and Playback Designs were likely less than a half decibel different in loudness, and the NADAC was about 2 db louder than those.

 

Perhaps you are right. The SPL meter was placed on the chair at the listening position and it wasn't moved when we were switching the DAC's to determine volume levels, so minor variation from the position of the SPL meter was eliminated as a variable. As far as we were concerned, they sounded the same volume subjectively and the SPL meter reported only very minor differences.

The NADAC being 2 db louder would explain why it was usually discernible as different. The other two being closely matched in loudness would explain why they were less reliably discerned. If I read between the lines of your blind results properly, you collectively picked the correct one between the MSB and Playback Designs 10 out of 18 times.

 

Yes, to be honest it was very difficult to tell the difference between the Playback, the MSB, and the NADAC in summed mode under blind conditions. Especially since we didn't use an ABX method. There was a 1 minute interval between each track. My audio memory is pretty good, but speaking personally I had a lot of trouble remembering which was which after listening to so many DAC's. It would have been much easier if we used an ABX method AND didn't try comparing so many DAC's.

 

The subjective impressions I posted at the end was what we thought was the difference between the DAC's from unblinded listening.

Thank you for reporting this results. It is interesting that according to David W. Robinson from Positive Feedback the new Merlot DAC seems to outperform the MPS-5, so the Merlot would be the least expensive of the three competitors.

 

That is exciting to hear! I have heard a Merlot, but not in comparison with the MPS-5 back to back in the same system. I might get in touch with the Playback dealer and ask if I could borrow a Merlot for evaluation.

 

There are two very important drawback of the Merlot when used in my application (multichannel), however. Firstly, no clock input to synchronize the output of all the DAC's. And secondly, I believe the output voltage is about the same as the MPS-5 (the specs are not published in Playback's website). If it's 2V, it will not be enough to drive power amps directly without a gain stage in between.

One other point I would make. I tested the UMT and was never crazy about it, but since you tested all the DACs listed, that variable is pretty much fixed in your comparisons. I would add, that not just me but many others, including owners of the Select II who have posted here, have stated that the QUAD USB port moves the MSB to a totally different level.

 

Thank you! I will send that feedback through to the MSB owner. As you said, we deliberately chose to use the same transport to eliminate that as a variable. If anything, this should have given the MSB an advantage because it was talking to its own DAC via an I2S connection, as opposed to the Mogami XLR cable we used to connect the transport to both external DAC's. This was plain old audio interconnect and was not designed for digital use. That is, if you believe that cables make a difference. We had other cables on hand but didn't use them because we wanted to listen to the DAC's and not cables.

 

By the way, I should emphasize again that the sonic difference between the NADAC (in summed mode), the Playback, and the MSB was miniscule. The cost of purchasing four Merlots for use in multichannel is AUD$48k - a substantial amount of money! If I had unlimited funds, I would probably do it. But for now the NADAC stays.

[One and a half]

 

Using a sound level meter to set levels is not an excellent method. The better method is to measure voltages of test tones at the loudspeaker terminals. A simple multi-meter would suffice for that. I would be interested in what tones you used with the SLM. Also did you use A weighting?

 

A quick check of Fluke hand held DMMs reveals very different frequency ranges of voltage measurements. Some can read 100-200kHz, others only to 10KHz. A brief look on Amazon didn't reveal any frequency specs for the less than $10 meters, so it would be wise to check the capabilities of the meter using the voltage input method at the speaker.

 

In any case, I would have thought the calibrated spl meter would have been fine, since everything in the room remained the same, except the DAC under test. Does the spl meter have too much tolerance?

[One and a half]

Thanks Keith for the great test and well written, with a surprise conclusion at the end, totally unexpected.

[bibo01]

I wonder if NADAC would perform better with Ravenna input using a computer transport...

[esldude]

A quick check of Fluke hand held DMMs reveals very different frequency ranges of voltage measurements. Some can read 100-200kHz, others only to 10KHz. A brief look on Amazon didn't reveal any frequency specs for the less than $10 meters, so it would be wise to check the capabilities of the meter using the voltage input method at the speaker.

 

In any case, I would have thought the calibrated spl meter would have been fine, since everything in the room remained the same, except the DAC under test. Does the spl meter have too much tolerance?

 

Measuring speaker voltages is simply the gold standard for this purpose. Every other method has more variables and sources of contamination and less accuracy. For instance the reason I asked about weighting. A-weighting rolls the bottom end off. I have seen freight trains when using C weighted measures move the total results around more than a decibel from 5 miles away. At frequencies you won't even notice or pay attention to at all. So in this case A-weighting would have probably been better.

 

If the meter responds much at all, the absolute accuracy isn't much of an issue. You might wish to use 440 instead of 1 khz or higher. Still, if it is wrong by a fair percentage, even cheap meters will be consistent in that error. So you will get consistent, reliable, accurate level matching. Which in this case is the important point more so than absolute accuracy.

 

Most even cheap multi-meters are good to 400 hz AC because some aircraft, industrial gear, and military generators around the world run on 400 hz AC. Most would still be fine for 1 khz even if the readings are a bit low. Beyond that unless you can confirm it or get assurances from the maker you can't be sure.

 

Of course nothing wrong with having a pretty nice meter for this.

[Keith_W]

Thank you esldude. I thought I would post the full methodology for people to look at and criticize. I will take your suggestion into account next time.

[Priaptor]

I think I mentioned this before, The MSB gear seems to tip the upper mid and tweeter level frequency to a state where it just sounds unnatural and fatiguing.

 

So dragon_vibe finds the MSB too "analytical" and in today's he post claims the above in his liking the Play Back Designs over the MSB whereas you guys find it the most "polite and laid back" just the opposite.

 

It's an interesting world out there

[Keith_W]

I can't explain why he thinks that. I can only report on what we, as a group, heard. I don't think the MSB sounded analytical at all.

[Priaptor]

I can't explain why he thinks that. I can only report on what we, as a group, heard. I don't think the MSB sounded analytical at all.

 

Ask 100 people about their opinions regarding audio equipment and you may just get 1000 different responses.

 

Having been into this stuff for 35 years, I know what I like, I have built an amazing system that few if any can come close to, at least IMHO. I have owned and lived with all different brands of DACs, etc and now I just listening to the music.

 

The maker of my speaker has demoed with Playback Designs with his wares on occasions at shows and really liked it (although he is not a real digital person)and I trust Carl that it must be a great DAC if he likes it. I have to admit, the new Merlot does interest me. I just finished my headphone system which I REALLY enjoy and sonically it is quite different than my big rig. This system consists of the T&A DSD DAC at 512DSD--Cavalli Liquid Gold-->Senn HD800S. Sonically it is pretty much the polar opposite of my main system BUT almost equally as enjoyable. After listening to all the hyperbole about the T&A I tried it on my big system and I knew it wasn't for me and was no comparison to the MSB. On my headphone rig I tried the "more polite" sound and really didn't love it as I do my current very neutral but analytical headphone system; from the amp, headphone and digital perspective. Yet on my big system I have had setups in the past where you could hear a mouse fart in the hall and I did find it fatiguing which I didn't like I can't explain my or other people's preferences or wants. However, my perspective is I never get attached to anything. I am willing to walk away if I find something I like better; life is too short. Having said that my big system hasn't changed much for the last 4+ years.

[astrotoy]

Keith, thanks for the great reviews. I had Dom of NADAC create a special configuration for me. It allows me to use channels 1 and 2 for stereo (and thus use the combination of all 8 DAC s chips in stereo mode, while using channels 3 thru 8 for my 5.1 set up. I have two different preamps, one for stereo and the other for mch, so I don't have to change cabling. The configuration is available on their latest revision of Emotion.

 

Larry

[Keith_W]

Where did you get Merging Emotion from? It can't seem to be able to find it for download in Merging's website, and I had a pretty good look.

[astrotoy]

Where did you get Merging Emotion from? It can't seem to be able to find it for download in Merging's website, and I had a pretty good look.

 

Dom set up my NADAC remotely from Switzerland. He downloaded Emotion while doing the setup. You can email him at [email protected]. Larry

[One and a half]

Where did you get Merging Emotion from? It can't seem to be able to find it for download in Merging's website, and I had a pretty good look.

 

This thread discusses Emotion. Goes back a long wayyyy.

[astrotoy]

This thread discusses Emotion. Goes back a long wayyyy.

 

I wrote the last post on that thread. During my visit to Geneva in the Summer of 2015, I saw the new version of Emotion (not a Beta) and got it when I bought my mch NADAC last fall. Dom Brulhart, the technical director of Merging, is the Emotion and NADAC man there.

 

Larry

[twist222]

Keith,

 

Great to see people still willing to share their impressions on an a comparision between 3 different dacs.

Just as a general point of interest could I ask if you were you using the JS2 Uptone analogue power supply for the NADAC or its own internal SMS.

 

Regards

Mark

[Keith_W]

Hi Mark, I was using the internal SMPS of the NADAC in this shootout. Good point though, the NADAC does sound better with the Uptone JS-2 powering it. Enough to beat the Playback or MSB, I am not sure.

[sixman]

Is there a big difference when using the Ravenna interface on the Nadac?

[Keith_W]

I don't know. I am not able to test that in my system, nor in his.

[bmoura]

Where did you get Merging Emotion from? It can't seem to be able to find it for download in Merging's website, and I had a pretty good look.

You should be able to get the Emotion software from your NADAC dealer or from Merging directly.

I tried it here but ended up staying with JRiver Media Center 22.

[wisnon]

I wrote the last post on that thread. During my visit to Geneva in the Summer of 2015, I saw the new version of Emotion (not a Beta) and got it when I bought my mch NADAC last fall. Dom Brulhart, the technical director of Merging, is the Emotion and NADAC man there.

 

Larry

Shoulda told me! I live just outside Geneva.