$600 EMO pre-pro measures better than $10,000 BRYSTON SP3

[Irvrobinson]

Ok, for number fun, I looked at the HTM vs AH on the Emo UMC200:

HTM: Crosstalk -94dB 1kHz 100mV Input (? Output) , SNR -128dBA
AH: Crosstalk -100dB 1kHz 200mV Input (1 Vrms Output), SNR -102dBA (-90dB Unweighted)

So the AH measurement gives better Crosstalk by 6dB, but the HTM gives better SNR by by 26dB!

If HTM measures a SNR of -95dBA on the $5K Arcam, does that mean AH would get a SNR close to -69dBA & close to -57dB Unweighted?

Am I even in the same ballpark ?

Thoughts?

You can't take two measurements taken on [probably] two different samples measured by two different labs, [probably] using different measuring equipment and different methodologies, and draw any conclusions about the relative relationships between the readings. As for that -126db SNR measurement, I have no idea how they measured it when most unweighted SNRs in solid state equipment comes in at -75db to -95db. I actually like the old-fashioned noise specs measured in volts (well, microvolts) better than these new-fangled who-knows-what-they-really-mean SNRs that have to be referenced to some output level and have interesting weightings invented to make the numbers look better.

The real problem is that highly technical data in the hands of non-experts is worse than meaningless. That's not intended as an insult, ADTG, it just happens to be true in every case I can think of. You've pointed that out once or twice in your own field. Perhaps I've just become cynical in my old age, but as soon as I see a figure of merit with complex weightings regarding any measurement in any field I always become suspect of BS.

 

[AcuDefTechGuy]

First, this is not meant to be serious talk. Just purely academic, if even that. Just curiosity.

As you guys have said, most Unweighted SNR are no more than -95dB. Even AH's A-weighted SNR is probably no more than -107dB.

Yet, HTM's SNR are usually all -100dB to -126dB (probably even more). Even a $200 AVR has a SNR of -100dB on HTM.

Of course, we are comparing different lab, personnel, method, technique, etc. The relationship is probably not linear.

Just trying to put things in perspective and talking numbers for fun.

So maybe the only conclusion we can make is that HTM's SNR are falsely high?

 

[PENG]

So maybe the only conclusion we can make is that HTM's SNR are falsely high?

Hard to say, but I noticed enought inconsistencies, even in their power output measurements. One thing I find it laughable is that they would publish the meaninglessly (In invented this word) low ACD number for the 4810 that to me is the last reasonably powerful and heaviest AVR Denon built, aside from the 5000 series and the old 4802. They only casually stating it was the protective circuit doing the job, without consulting Denon about why the made the protective circuit more aggressive than their lower models. That kind of act is not only laughable to some of us subscribers who have read many of their reviews, I but can also be misleading to non subscribers who are not familiar with their pattern.

Again, if the SNR and cross talks are high enough, you have got to give Bryston and others some consideration that perhaps their puriest approach (e.g. minimized use of ICs or not used at all in any amp stages) and other lesser of the evils kind of choices made, might have resulted in worse XT/SNR relative to other lower priced products. At the end of the day, you win where you have to win and lose where you don't care too much about not winning. That would likely not be the case for higher end manufacturers such as Krell, Boulder, Luxman, Moon, Mark Levinson, McIntosh etc., as they are closer to the cost no object category.

 

[AcuDefTechGuy]

Signing OFF on Numbers War Game

Yes, regardless of measurements, I would prefer a Mark Levinson, Krell, McIntosh, Classe, any day, anywhere.

OK, I think this will be the last time I'll bring up numbers (NO! ), but looking through "numbers" one last time, I stumbled upon this AH review of the Yamaha 1010: Yamaha RX-A1010 AVENTAGE A/V Receiver Review Measurements and Analysis | Audioholics

"I measured analog channel to channel crosstalk of the front main channels using the analog inputs and analog multi channel outputs. In “straight mode” crosstalk measured < 100dB at 20kHz which was outstanding. What was even more impressive was in “pure direct” mode, the channel to channel crosstalk was almost below measurable limit (-120dB) of my $40k AP APx585 Audio Analyzer. Yamaha really did their homework with component and trace routing in this baby. Anyone telling you separates provides better channel to channel isolation, have them look at the measurement graphs above."

"Using the HDMI input driven at -20dBFS with 1Vrms out of the preamp analog outputs, I measured 87.5dB SNR (Unweighted)."

I guess the IC Yamaha uses gets pretty good Crosstalk of -120dB!

Okay, I think I've had enough of numbers, which I'm sure 3dB will be relieved to hear.

ADTG signing off on the number war games.

 

[PENG]

Yes, regardless of measurements, I would prefer a Mark Levinson, Krell, McIntosh, Classe, any day, anywhere.

OK, I think this will be the last time I'll bring up numbers (NO! ), but looking through "numbers" one last time, I stumbled upon this AH review of the Yamaha 1010: Yamaha RX-A1010 AVENTAGE A/V Receiver Review Measurements and Analysis | Audioholics

"I measured analog channel to channel crosstalk of the front main channels using the analog inputs and analog multi channel outputs. In “straight mode” crosstalk measured < 100dB at 20kHz which was outstanding. What was even more impressive was in “pure direct” mode, the channel to channel crosstalk was almost below measurable limit (-120dB) of my $40k AP APx585 Audio Analyzer. Yamaha really did their homework with component and trace routing in this baby. Anyone telling you separates provides better channel to channel isolation, have them look at the measurement graphs above."

"Using the HDMI input driven at -20dBFS with 1Vrms out of the preamp analog outputs, I measured 87.5dB SNR (Unweighted)."

I guess the IC Yamaha uses gets pretty good Crosstalk of -120dB!

Okay, I think I've had enough of numbers, which I'm sure 3dB will be relieved to hear.

ADTG signing off on the number war games.

Keep those numbers coming, I love them. Everything else being equal, I would make my decision based on numbers whether I can hear them or not. For example, if Yamaha 3010 has Audyssey XT32, I would have jumped on it but I have to make my irrational decision now on the Marantz because I need XT32 to tame my subs. I don't have high end sub like yours and am too lazy to tune them myself, it used to take me hours if not days to do just the PC Ultra. Like you, I mostly use Pure Direct or occasionally direct + sub in 2 channel listening, but I do like to have the ability to flip between with and without Audyssey on. I know you don't believe in room EQ and still I think it is partially because of your room, equipment combo and partially because Placebo may be working on you in reverse order because of your preconceived idea that pure is best, and that you have hard time believing something you cannot see, I mean the manupilation in the time domain (you can only see the effects in the freq domain).

 

[PENG]

I guess the IC Yamaha uses gets pretty good Crosstalk of -120dB!

Or they put more efforts in shielding and circuit board layout etc., to minimize capacitive coupling effects, while Denon and Marantz put more efforts in other areas that they consider more important to sound quailtiy, efficiency, power output, cosmetics etc.

Just an arbitrary example:

Say you have a $200 budget on the power supply transformer, and you can get one rated 800VA and another 650VA output from two different manufacturers for the same $200. The 650VA one would allow you to achieve your goal of achieving -90 dB XT while the 800 VA could only get you -75 dB, assuming other factors such as PCB trace, cable routing design remains constant.

Which one would you go with? Let's say the decsion maker in Yamaha believe more in dynamic output and that MCH output is not that important for real world HT applications, so they may go for the less powerful transforme, while HK, a proponent of the emphasis on MCH specs, would likely opt for the 800VA transformer. As a user, neither matters to me but since I don't use the amp section I would take the better shielded transformer. If I were to buy it as an AVR in a smaller room then I would go with the latter.

 

[AcuDefTechGuy]

Keep those numbers coming, I love them. Everything else being equal, I would make my decision based on numbers whether I can hear them or not. For example, if Yamaha 3010 has Audyssey XT32, I would have jumped on it but I have to make my irrational decision now on the Marantz because I need XT32 to tame my subs. I don't have high end sub like yours and am too lazy to tune them myself, it used to take me hours if not days to do just the PC Ultra. Like you, I mostly use Pure Direct or occasionally direct + sub in 2 channel listening, but I do like to have the ability to flip between with and without Audyssey on. I know you don't believe in room EQ and still I think it is partially because of your room, equipment combo and partially because Placebo may be working on you in reverse order because of your preconceived idea that pure is best, and that you have hard time believing something you cannot see, I mean the manupilation in the time domain (you can only see the effects in the freq domain).

Well, it's tough arguing with the complexity and sometimes inexplicability of the cerebral interpretation of those sensory stimuli.

I'm at a point where if another member tells me he or she can hear a difference or cannot hear a difference, I will take it at face value. I can't possibly know how other people's brains interpret stimuli.

If it works perfectly, keep on doing it and don't change.

If it sounds perfectly, buy it regardless of measurements and specs.

 

[AcuDefTechGuy]

Or they put more efforts in shielding and circuit board layout etc., to minimize capacitive coupling effects, while Denon and Marantz put more efforts in other areas that they consider more important to sound quailtiy, efficiency, power output, cosmetics etc.

Just an arbitrary example:

Say you have a $200 budget on the power supply transformer, and you can get one rated 800VA and another 650VA output from two different manufacturers for the same $200. The 650VA one would allow you to achieve your goal of achieving -90 dB XT while the 800 VA could only get you -75 dB, assuming other factors such as PCB trace, cable routing design remains constant.

Which one would you go with? Let's say the decsion maker in Yamaha believe more in dynamic output and that MCH output is not that important for real world HT applications, so they may go for the less powerful transforme, while HK, a proponent of the emphasis on MCH specs, would likely opt for the 800VA transformer. As a user, neither matters to me but since I don't use the amp section I would take the better shielded transformer. If I were to buy it as an AVR in a smaller room then I would go with the latter.

I agree implicitly.

But honestly, were you utterly shocked to see a $1,000 Yamaha AVR w/ a XT of -120dB from 20Hz-20kHz? Even a little bit?

In Straight Mode, XT was -100dB. In Pure Direct Mode, XT was -120dB. Of course, no human being could tell the difference, but...

I think you & I may be the ONLY ones even remotely interested.

So unless someone brings it up, I won't initiate the numbers games.

 

[panteragstk]

I agree implicitly.

But honestly, were you utterly shocked to see a $1,000 Yamaha AVR w/ a XT of -120dB from 20Hz-20kHz? Even a little bit?

In Straight Mode, XT was -100dB. In Pure Direct Mode, XT was -120dB. Of course, no human being could tell the difference, but...

I think you & I may be the ONLY ones even remotely interested.

So unless someone brings it up, I won't initiate the numbers games.

I like numbers too so keep em coming. Didn't finish my EE degree, but I know enough to not be confused (most of the time ). Numbers make me happy.

 

[3db]

Okay, I think I've had enough of numbers, which I'm sure 3dB will be relieved to hear.

ADTG signing off on the number war games.

Until next time!!

Seriously though, numbers aside, it always creates side discussions so keep on going..

 

[Steve81]

On a related note, some measurements from S&V of the UMC-200/UPA-700 combo:
Review: Emotiva UMC-200 preamp/processor and UPA-700 7-channel amp | Sound and Vision Magazine

Edit: And a natural competitor for reference:
http://www.soundandvisionmag.com/article/test-bench-web-outlaw-audio-model-970-preamplifier-processor-and-model-7075-power-amplifier

 

[AcuDefTechGuy]

I like numbers too so keep em coming. Didn't finish my EE degree, but I know enough to not be confused (most of the time ). Numbers make me happy.

Until next time!!

Seriously though, numbers aside, it always creates side discussions so keep on going..

Well, I will post some numbers, but only if other members initiate.

Like in this thread: http://forums.audioholics.com/forums/amps-pre-pros-receivers/85208-denon-3805-a.html#.UXAOSMu9KSM

 

[PENG]

I agree implicitly.

But honestly, were you utterly shocked to see a $1,000 Yamaha AVR w/ a XT of -120dB from 20Hz-20kHz? Even a little bit?

In Straight Mode, XT was -100dB. In Pure Direct Mode, XT was -120dB. Of course, no human being could tell the difference, but...

I think you & I may be the ONLY ones even remotely interested.

So unless someone brings it up, I won't initiate the numbers games.

Yes I am very surprised to see those numbers, aside from everything I cited so far, the bottom line must be that they are good at it, and found a way to do it economically. I don't see any harm bringing up numbers, everyone knows you and I can't hear the difference, that's why we buy on numbers.

 

[AcuDefTechGuy]

On a related note, some measurements from S&V of the UMC-200/UPA-700 combo:
Review: Emotiva UMC-200 preamp/processor and UPA-700 7-channel amp | Sound and Vision Magazine

Edit: And a natural competitor for reference:
Test Bench for the Web: Outlaw Audio Model 970 Preamplifier/ Processor and Model 7075 Power Amplifier | Sound and Vision Magazine

So the SNR numbers on S&V are a lot lower than HTM as well.

 

[Irvrobinson]

Yes I am very surprised to see those numbers, aside from everything I cited so far, the bottom line must be that they are good at it, and found a way to do it economically. I don't see any harm bringing up numbers, everyone knows you and I can't hear the difference, that's why we buy on numbers.

The question has never been about whether measurements are relevant, the problem is finding measurements that can be trusted. When someone says they've measured an SNR of -126db in an AVR or a pre-pro, and you know some of the best test equipment around can barely measure noise levels that low, it makes you wonder about the efficacy of the measurements, especially when the tester tells you nothing about how the measurements were made.

 

[Steve81]

So the SNR numbers on S&V are a lot lower than HTM as well.

One thing worth noting there from S&V:

All data were obtained from various test DVDs using 16-bit dithered test signals, which set limits on measured distortion and noise performance.

 

[AcuDefTechGuy]

One thing worth noting there from S&V:

What exactly does it mean when they "set limits on measured noise"?

Like they can't get SNR above -95dB or below -60dB?

 

[Steve81]

What exactly does it mean when they "set limits on measured noise"?

Like they can't get SNR above -95dB or below -60dB?

16 bit audio is limited to a SNR of 96dB; dithering is an intentional injection of noise to reduce the effect of quantization error, which means you won't even get to 96dB.

 

[PENG]

The question has never been about whether measurements are relevant, the problem is finding measurements that can be trusted. When someone says they've measured an SNR of -126db in an AVR or a pre-pro, and you know some of the best test equipment around can barely measure noise levels that low, it makes you wonder about the efficacy of the measurements, especially when the tester tells you nothing about how the measurements were made.

Well that's why we have ISO, UL, CSA, certified professional engineers in US/CA, and Chartered Eng. in UK, Euro Eng. in Europe etc., etc.. I wish there were certified independent labs for audioholics like us who do value numbers, as you said, numbers that can be trusted aside from being relevant.

 

[AcuDefTechGuy]

16 bit audio is limited to a SNR of 96dB; dithering is an intentional injection of noise to reduce the effect of quantization error, which means you won't even get to 96dB.

Ah, so I see why the SNR on Stereophile & S&V will probably not reach -96dB.