Big Sound 2015: Getting Some Numbers

A lot of the gear coming if for the Big Sound 2015 project from manufacturers is show gear...you know, stuff that gets shipped from show to show a dozen times a year. For that reason alone I thought it would be a good idea to get measurements of the amps. Unfortunately I don't have a routine for tube amps or DACs in place yet, but I can measure the solid state amps.

All the following measurements were done using the unbalanced (RCA) inputs and the normal 1/4" unbalanced headphone jack on the front panel. I did run into some weird measurements on the Schiit Ragnarok, and after sending measurements and talking with Jason Stoddard we've decided to replace the Ragnarok here. So, Rag measurements are not included in this release, but will be in future when it's replaced.

Clicking on the measurements image will link you to a .pdf file, but you can download an entire booklet of the plots.

JDS Labs O2
The NwAvGuy designed JDS Labs O2 is not being auditioned in this project, but I've included it here because it does indeed measure very well. In my experience this is a very good amp for the money, but ends up sounding a little dry and grainy subjectively relative to really good amps. None the less, it is a very worthy reference when looking at measurements. Let's go over the sheet.

Click on graphs image to download .pdf for closer inspection.
Here are some things to note with these measurements:

  • Frequency response essentially flat from 20Hz to 100kHz; phase response likewise flat with phase angle remaining within 1/2 degree over entire range.
  • THD+Noise is below 0.01% for all loads until clipping is reached. (Clipping is where to the right on the graph where the plots shoot up over 1%.)
  • Noise spectrum baseline is at an astonishing -150dB; noise spectral peaks all below -130dB.
  • DIM/TIM: This is a rather complicated test that looks for transient intermodulation distortion. What you are really looking for is the height of a peak at 750Hz. With the O2 you can see it's at about -118dB.
  • SMPTE IMD (intermodulation distortion): This test excites the DUT )(device under test) with two tones, one at 60Hz and one at 7kHz. What you are looking for with poor performance is sidebands around the 7kHz tone. The only amp that had any problem, and therefor some visible results around 7kHz is the Bakoon amp in voltage mode—which, of course, is not really the way this amp is designed for maximum performance.
  • CCIF IMD: This is a test with two tones at 19.5kHz and 20.5kHz. Poor performance will show up as: a series of peaks around the two tones in 1kHz intervals; a difference peak at 1kHz; and a sum peak at 40kHz. The lower these peaks are the better the performance is.
  • Crosstalk: I do this measurement without a load on the output; I may change this procedure and load it as it will probably be more stressful on the DUT that way. this is a test that puts a tone in one channel, and reads how much of that tone can be heard in the opposite channel. It primarily measures how much audio signal is getting through the power supply from one side to the other, but there are other possible mechanisms to produce crosstalk.

I'd like to preface the following dialog by admitting that I am not an expert in this field. Many InnerFidelity readers are much more familiar with analog electronics measurements and I welcome comments and observations from qualified folks to flesh out my own.

Let's have a look at the other amps. I will just note some of the more remarkable measurements with each amp.

Antelope Audio Zodiac Platinum

Click on graphs image to download .pdf for closer inspection.
Most measurements are fairly good here with the exception of a treble roll-off that's a little over 2dB down at 20Khz, and a bit more sidebands around the peaks at 20kHz in the CCIF IMD measurement. Crosstalk and THD+noise are better than average in the group.

AURALiC Taurus Mk2

Click on graphs image to download .pdf for closer inspection.
While still well below what's considered the audible threshold (-60dB) this amp shows the most noise in the group—as seen by the spiky nature of the Noise Spectrum. Crosstalk flat at -60dB is also the worst of the group. Intermodulation Distortion measurements are among the best of the group.

Bakoon HPA21

Click on graphs image to download .pdf for closer inspection.
Click on graphs image to download .pdf for closer inspection.

WARNING: This is a current mode amplifier. The test sequence I use is designed for voltage mode amplifiers. It is very possible—likely, in fact—that the methods osed to test this amp in current mode are completely wrong for this amp. They should be good for the voltage mode, but, in fairness, this amp is designed to be used in current mode and is likely not optimized at all for voltage mode.

This amp has a very flat frequency response over the entire measured range, and while the noise floor at -130dB is about average, it has very few noise spuria (peaks in the noise plot; possibly due to the fact that this amp has a battery-based power supply). Crosstalk is good in voltage mode, and poor in current mode. I would tend to believe the voltage mode plot. The only thing that concerns me with this amp is somewhat higher than average distortion measurements on all fronts. Again, however, these may not be valid readings. This amp, in many ways, measures better than I expected—even though in some ways it's very weird.

Burson Audio Conductor Virtuoso

Click on graphs image to download .pdf for closer inspection.
Though having a bit more spuria in the noise spectrum and more sidebands around 20kHz in the CCIF DIM measurement, and a bit better crosstalk performance, the Burson measures about average in this group. That's actually pretty darn good considering it's also one of the least expensive amps in the group.

Eddie Current Black Widow

Click on graphs image to download .pdf for closer inspection.
This is the least expensive amp in the test and does perform quite well. Two things to note on the down side are a 1dB channel volume missmatch in the frequency response plot, and a broadband noise floor at -120dB—about 10dB higher than other amps in the test. On the other hand, noise spuria are very low; distortion products of all kinds are low; and crosstalk performance is very good. The long, slow rise in THD+Noise graph to the right side of each plot indicates, it seems to me, minimalist class-A output stage design. I'm very much looking forward to ears-on time with this amp.

Simaudio Moon NEO 430HA

Click on graphs image to download .pdf for closer inspection.
This amp is a beast. Noise floor is about average, and noise spuria maybe slightly better than average, but by all other measures the 430HA kills it. THD+noise simply shows a noise limited slope downward with essentially no distortion measured until voltage exceeds 10Vrms...even with a 16 Ohm load. (My 16 Ohm loads were almost too hot to touch after undergoing tests with this amp.) Other DIM/TIM, IMD, CCIF distortion measures (in the single point measurement box to the lower right on sheets) are very good indeed. Crosstalk measurements are extraordinarily good at high frequencies. This amp looks like it will drive absolutely anything you throw at it.

Violectric V281

Click on graphs image to download .pdf for closer inspection.
Broadband noise floor is about 5dB lower than average for the group at about -135dB, but there appears to be a lot of noise spuria. THD+noise starts at the left well below 0.1%, which is better than average here. Crosstalk is very low, slightly better than the Simaudio 430HA. Evidently there were some problems with the single point measurements using the 16 Ohm load—this is not a problem with the amp, but rather the level setting on those measurements is very unstable. I will try to remeasure this amp. The only measurement worse than average of this group with this amp is the roll-off in the bass to 2dB down at 20Hz.

That's it for today. Lots of stuff coming in, should have some cool stuff to show you Friday!

Equipment List

Front End
NAS - Synology DS414 ($479)
Renderer - Aurender W20 (~$17,600); AURALiC Aries ($1599)
Digital Distribution Amps - Four ATI DMM100 Digital Matchmakers
and one DDA212-XLR digital audio distribution amp ($1450).

Power Conditioning
PS Audio, two P10 power regeneration station ($4999) and four DecTet conditioned plug strips ($499).

DAC/Amps
AURALiC Vega DAC ($3499) and Taurus MkII headphone amp ($1899)
Simaudio MOON Neo 430 HA ($4300 w/DAC).
HeadAmp GS-X Mk2 ($2800)
Schiit Ragnarok ($1699) and Yggdrasil ($2299)
Burson Audio Conductor Virtuoso ($1495 w/PCM1793; $1995 w/ESS1908)
Woo Audio WA-234 ($15,900)
Chord Hugo TT ($5000)
Antelope Audio Zodiac Platinum DSD DAC, Voltikus Power Supply, and 10M Rubidium Atomic Clock. ($13,045)
Apex High Fi Audio (TTVJ) Teton ($5000)
Eddie Current Black Widow ($1248)
Violectric V281 ($2299)
Bakoon HPA-21 ($2995) current output headphone amplifier.
KGSSSRE (Kevin Gilmore Solid State Special Reviewer's Edition E-Stat Amp ($Unobtanium)

Headphones
Sennheiser HD 800 ($1599)
Audeze LCD-3 ($1945) and LCD-X ($1699)
JPS Labs Abyss AB-1266 ($5495)
Stax SR-009 ($4450) and SR-007 ($2350)
HIFIMAN HE-1000 ($3000)
Mr. Speakers Ether ($1499)
Enigmacoustics Dharma (~$1200)
Audio Zenith PMx2 ($1398)

Cables
Digital cables by AudioQuest.
Cable complements for wiring entire systems will be from: Nordost; JPS Labs; WyWires; Cable Pro; AudioQuest, and Cardas.

Accessories
Headphone stands by Klutz Designs

And more to come, no doubt....

COMMENTS
Dreyka's picture

If it sounds 'grainy' then surely you'd be able to tell it apart from better performing amplifiers under blind conditions. I know it can be a pain to volume match the Objective2 due to the potentiometer but what I'd really like to see is you blind ABX testing the Objective2 against the far more expensive amplifiers like the Schiit Ragnarok.

Think about it this way. If an amplifier that is 1x cheaper ($99 vs $1699) than another then you owe it the readers to be able to pass or fail a proper blind ABX test.

You've got some of the best ears out there so setting it up so the results are important.

Tyll Hertsens's picture
I'm not going to do that this time, but I will do it sometime in the future.
zuglufttier's picture

Please, really do this! There are really big differences in headphones to be found, and easily so, but once you aquired a working headphone amp with a very low output impedance, the search is over. There are only headphones left to play with...

tyj518's picture

I heard that for v181 there is an AC/DC jumper inside the amp, and the AC setting will cut off the frequencies below 12Hz. I'm not sure if it is true for v281, but this might be a reason why it has a bass roll-off.

Tyll Hertsens's picture
Nice tip, I'll look into it.
roskodan's picture
Eli's picture

Yes please do, I am interested in this to. I found the 2 db roll-off to be a little much for a top of the line product.

roskodan's picture

also one of the single ended outputs is inverted by 180*

which is odd, since they could have just switched the output socket (+,-) wiring

guess they meant it as an extra feature?

castleofargh's picture

\o/ I was wondering where your amp measurements were going. really appreciate this. thx

tony's picture

How much voltage do headphones require?

Can we conclude that all these Amps can properly drive the HD600 series?

The gist of my interest lays in my wonderment about how much "power/voltage headroom" is important.

I just read about a Preamp that goes "flat" out to 1 Megahertz, which I suspect has no importance for an Audio application and which I suspect "is" designer sizzle typical of the man that builds that kind of stuff. ( silly braggadocio: my thing is bigger than you're little thing )

By comparison, the Bakoon seems far less powerful ( or is it ? ), maybe it just has less headroom than the others.

Hmm, let me see, One Volt @ 300 Ohms = ? , ( probably enough power to do permanent damage to somebody's ears in less than 1 minute of heavy-metal, providing the voice coils don't heat up and melt the adhesive or burn out ).

For Clarity sake, can real power correlations be part of Amp discussions?

Arn't headphones rated in milliwatts per db?

Is 7 or 8 volts output something like " flat out to 1Mhz" in relevance/importance?

-----------

Anyway, this series is thought provoking, thank you.

It's taking my mind off a renovation project I'm in the middle of and the British Open where I have bets riding on that 21 year old hot golfer from the States NOT winning ( fingers crossed ) , not to mention the bet I have on the Froome winning the Tour de France!

I feel you're doing EXACTLY what all of us are trying to do: make sense out-a this wonderful headphone gear. I nearly purchased that Antelope system but held off because of the Schiit Yggy. I nearly bought the MSB Analog DAC ( bout $10k ) but held off for the same reason. How can a person make an impulse decision if the costs are so high and the variables are nearly infinite?

Jason Stoddard's weekly story was fascinating,

Bob Katz's 5 Parts were exciting,

The Tyll Series looks like it too will be a thriller for all of us, I'm betting your clicks go way up as word gets out.

You'll need a week on the Riviera before you commit to doing the 2016 Series Opener

Tony in MIchigan

castleofargh's picture

it's all about how loud you listen to music. as loudness ends up in direct relation to voltage U, and max power is P=U²/R. so how high you will push U is really the main point once you know into what load R(here 300ohm for your hd600).

the trouble is that we're talking max voltage, so when you say you listen to music @80db, depending on the song, you in fact may listen 15 or 20db louder(or maybe less than 5db if it's a justin bieber song). because the average loudness isn't the same as peak loudness.

so you really need to decide on a max loudness you will need/want to feel reassured about the amp(usually people pick between 110 and 120db as a max possible when they have no idea what to use for themselves)
once that's done, you can get the manufacturer's specs @1khz and use the excel stuff here http://www.apexhifi.com/specs.html (click on the picture to download).

the other solution is to take the value measured by someone, don't know if you heard of this website, the guy made a crazy big list of headphones measurements http://www.innerfidelity.com/headphone-data-sheet-downloads ^_^

so for example, I decide that I want 115db max because I know I will never ever pass that loudness but I may come up close:

-from the first link:
the hd600 is given for 97db@1mw and 300ohm. to get 115db the excel file gives me 4.35Vrms so I will want an amp that can provide at least that into 300ohm. (and if you only have power values from the amp specs, just remember P=U²/R. so in this example P=4.35²/300=0.063W so you need an amp that can provide 63mW rms into 300ohm without clipping. not too hard but some DAPs and portable amps will fail to do so.

-from Tyll measurements: he measured http://www.innerfidelity.com/images/SennheiserHD600.pdf 0.230Vrms to reach 90db
so to reach 115db you need to go 25db louder than 90db (I know I'm a genius ^_^), that's about 17.8X voltage gain, so 0.23*17.8=4.09Vrms

so you see both methods come out real close, as from 4.35v to 4.09v there is only a 0.5db difference. more than precise enough to make estimates of our needs.

but again the big unknown is really "how loud will I listen to the music?".
also: we check values @1khz only, it's ok for most usage(IEMs are often an exception but they never really need power anyway), but not an absolute proof for the entire audible range.

if I said something stupid, I hope someone will point out where I'm wrong. thx

tony's picture

Now I realize that headphone amps are very low power devices.

Even the most inefficient headphones have tiny power requirements ( compared to any other type of transducers I've ever owned or used ) .

I never get my Asgard 2 over 11 O'clock on the main Gain control ( I typically listen at 8 to 9 O'clock ) .

The Vast Array of CDs I have coming in every week feature a wide range of listening levels which my Sennheiser/Schiit "coast" thru.

Thank you for helping with your educated reasonings.

Tony in Michigan

m8o's picture

Now more than ever I wish a McIntosh MHA100 found its way to your possesion. I'm really curious what it's putting out and how clean.

thune's picture

I can't make any sense of the phase (green line) in the FR graphs. Even interpreting it as radians, the green phase line still doesn't track the FR in the expected manner.

chik0240's picture

So, by the measurement results the Taurus seemed to sound worst in this group?

Tyll Hertsens's picture
Please remember that even the worst of the measurements above, errant artifacts present are mostly significantly below what is considered the audible threshold. It's probably better to think of the differences seen in the measurements above causing the flavor of the various amps. Sometimes that flavor is tasty and sometimes not. So it's very hard to tell whether what you're looking at when measurements show imperfection as they may be good or bad and there's no way of knowing without listening.
chik0240's picture

Just curious, because most people (including myself being a owner of a Taurus Mk2) the listening experience compared to say the Burson was that the Taurus was actually more transparent and revealing, thus t he measurement of higher cross talk and noise seemed suggesting the other way round.

I do agree that it's all about listening since that's what we are all investing big money on

Alamei's picture

I know you said in an earlier post that you would have to be shuffling the GS-X Mk II around due to time-sharing constraints, but will you get a chance to measure it at some point during this process? Since it's such a well-regarded piece of equipment, I'd really like to see how its measurements stack up.

Tyll Hertsens's picture
Absolutely!
m8o's picture

Tyll, is there any chance you'll get a MHA100? Curious if you requested one and McIntosh did not provide it, or it wasn't requested ; and if the latter if you're interested in testing it, or not at all?

Chiumeister's picture

The headphone output of the NAGRA HD DAC works extremely well with Fostex, Audeze and Sennheiser.

TheAudioGuild's picture

Typo in the first subheading. You have "LDS Labs" instead of "JDS Labs." Dont want to piss off the Mormons. ;-)

Tyll Hertsens's picture
Thanks.
tony's picture

Why not?,

anyone else get a pass?

Tony in Michigan

TheAudioGuild's picture

The Jews get Passover. That close enough?

tony's picture

What'cha think about Floyd Toole's video comments about Wire, Reviewers, Double Blind, Reviewer's tests, using Mono and "Circle of Confusion" ?

I was impressed and felt refreshed by the Candor.

Tony in Michigan

ps. you are probably the most critical evaluator of communicated things that I'm aware of ( maybe Abe Lincoln was once but he's long gone ).

and if that ain't enough bait:

Are you happy with Obama ?

TheAudioGuild's picture

Had t watched it before you wrote this so I just watched it.

Don't have any problems with it at all.

tony's picture

That is quite a compliment,

coming from you of all people.

Thanks , I'd thought so.

Tony in Michigan

TheAudioGuild's picture

Were you expecting me to have problems with it? If so, I don't know what would have given you such an idea. I'm quite familiar with Toole's work and have never had any issues with it.

tony's picture

Not taking the Bait, darn!

Tony in beautiful Michigan

tony's picture

I didn't know what to expect from your thoughts but you'd be the only person I know of that could find a problem with water freezing at 32F. ( not that it's a bad personal quality )

I wasn't aware of F.Toole before I saw this video and I had Sales responsibilities with his Firm ( back in the mid 1980s )!

At that very same time, I was importing the M2 loudspeakers that Toole properly described and admired, I had no clue that US Engineers ( other that us in Michigan and in the UK ) though well of them or took the time to discover them. They didn't sell well ( they were sales duds, if anything ).

About the video: I had the idea that you were critical of Harmon people because of something technical that they were bull-s….. people about. Maybe it was one of the other outfits, oh well, no big deal.

I think I like Toole, maybe there is something important at Harmon, I did like their little T-60 Turntable and their frosted cabinet fronts and beautiful little knobs. Nothing special sonically though but at least they were competitive with the JVC Receivers and Direct Drive turntables in fake Rosewood.

Nice hearing from you,

Tony in Michigan

Augustus's picture

Will you measure the Aune X1s, please?

Seems very interesting for its low price and the advertised specs!

X