I’ve been doing A LOT of thinking lately. In particular, about the importance of audio measurments for source components like DACs and CD players.
Let us first assume that we have 2 identical DACs or 2 identical CD players. You wouldn’t dare suggest that the same models sound inherently different, now would you? Well we can prove that the output of each device in this scenario is identical by doing a null test. We capature the output of the DACs and CD players and learn that their waveforms (let’s say a 30 second clip) are identical. The only time we might see a difference is in an engineering/manufacturing hiccup...and that is RARE considering we have globalization in the modern world today followed by quality control standards that are not necessarily difficult to get right.
And so, if put to practice, any 2 digital audio components that have similar enough measurements should sound identical. For example, a DAC with a SINAD or SNR or 120 dB vs one with a SINAD or SNR of 123. Tiny differences in linarity and frequency response above 20 KHz are not audible to us humans anyway.
Because most of our listening dare not go up to 110 dB, which is the threshold of discomfort. You could only listen for up to about 30 minutes at this level without risking hearing loss! For this reason, the ideal listening level is below that!
Should we forget about what companies try to sell us as high-end and focus purely on measurements with respect to accurately reproducing digital audio?
Here’s what’s really funny. The Chord DAVE performed worse with respect to measurments than the Chord Hugo TT2! Just see audio science review.
Lastly, I consider ASR the best objective website on the internet, bar none. Because if Amir really had a business relationship with any of these audio companies, their flagship or most expensive products would always perform at the very top; we see that is not the case and measured performance is all over the place!
Looking forward to hearing from you guys. Let’s not turn this discussion into a flame war. If you disagree with what I’ve written, just tell me why. I will investigate.
The one thing I haven't seen mentioned is that most, if not all measurements are static. A sine wave (or two in the case of IM distortion) is fed into the device and deviations from the input are noted, or a power output recorded, etc.
However, music is complex and dynamic. There is no guarantee that a static measurement accurately represents the device's behavior when handling a complex signal. One of the best examples of this were high-feedback solid state amps from the 1970s that measured well but sounded shrill and harsh. Things are better these days, but there is still no perfect correlation between static measurements and musical performance.
Then, there is also the interaction between multiple components. An amp that works well with a speaker with a benign impedance curve may not sound near as good with a difficult load.
That said, measurements are still important, particularly when designing and manufacturing products. For manufacturing, measurements are the only practical way to maintain consistent quality control as there is no way to listen to every unit as it comes down the production line.
Finally, don't think I'm a pure subjectivist. We are humans, and subject to both conscious and unconscious bias that have little or nothing to do with sound but which still affect a person's perception. Just because the subject is hi-fi doesn't mean those issues don't exist.
I really do not see the point of these posts. ASR is not taken seriously by those who are really interested in music. It is best to ignore them and let them argue amongst themselves.
There are so many technical factors. Crown International in the late 1970s, 80s made some of the best built and technically measured stereo equipment in the world. Dollar for dollar the best. .00025 intermodulation .0025 THD at full rated output in their preamps. Power amps .0025 IM and .025 THD at full rated output at any of the Amplifers from their 45WPC amp to their 500WPC Stereo reference. Did they sound good? You bet! However, if you wanted to pay the money for a less powerful amp with worse specifications that sounded remarkably better you would buy a Mark Levinson Mono Block pair with their preamp. I owned a Crown PL 1 SL1 and enjoyed it for 40yrs. Still sounds acceptable as long as the switch buttons work for power and function. So outstanding quality. After 15 years I sent it back to Crown. They replaced one small capacitor in the power protection circuit. It all tested like new. So, testing is great. It gives you an idea on how accurately quality control throughout is designed into equipment. But it does nothing to tell you about sound. The system might spec great but be overly bright or too robust on the same speakers. So room matching, speaker matching preamp, power amp, source are all important factors just as important as technical measurements. My old Crown equipment was so much better technically than my new tube integrated but in no way sounds as good. No contest.
I saw people making unfair comments here discrediting Stereophile in a similar fashion on ASR. Stereophile supplements their extensive listening with measurements, while ASR occasionally adds de minimis listening test in the end of measurements. Although I look at their measurements as one of the decision factors, I understand there are other psychoacoustic components not being measurable at least based on today's technology. That is why identical measurements (available today) can not be equated with identical performance. I found Stereophile reviews are as such much more trustworthy.
I understand both sides of the issue. Objectivists posit that if it can't be measured, then it doesn't exist. There is a certain logic to the notion that what we hear is a function of the measurements of a piece of equipment, or that one cable sounds different because it measures differently in some way. Subjectivists posit that measurements are irrelevant, and that all that matters is your subjective impression of the equipment. Subjectivist opinions are therefore . . . subjective. Some will have one impression of something, others will have a different impression.
I think that there is merit to measurements but it is clearly not the be-all and end-all. There are so many factors, such as other equipment, speakers, room acoustics, cables, power, sources, etc. that it seems that subjective analysis is the only way to decide whether a piece of equipment is good for you. How does one measure depth and width of soundstage, for example? How does one measure the fact that one piece of equipment has an almost 3D presentation in space while another is more 2D? Logically, if it sounds different, it should measure different. But I don't think that there is a true correlation between how a piece of equipment sounds and how it measures. so, bottom line, identical measurements do not mean identical performance.
Any competently designed DAC should nail SINAD and frequency response measurements. That's why, on cursory listening, DACs sound the same. Only on a sufficiently resolving system/room with extended sessions will the subtle differences between DACs or similarly measuring amplifiers become apparent. You have to know the recording and you have to spend some time. There is a lot that can go awry when reconstructing the analog waveform from 1s and 0s and if you listen, those differences will reveal themselves. All IME and IMHO of course.
Who listens to asr? Measurements don’t tell you how it sounds. What is the sq difference of a dac with .005 distortion vs .0003? Good luck. Stereophile is the worst on producing bad measurements and reviewers love the sound. I never look at measurements of a dac before purchasing, I use my ears
I have mulled over this very argument for over 50 years and Once again I will use a speaker as an example or a comparison as to why things that are unmeasurable or even hearable by the human ear can change the overall makeup of a sound. Just in case you noticed I already explained the whole story already. It is a matter of what one minute aspect that can't be heard affects the overall.
We should all know that the different materials used in making a speaker cone act differently at different frequencies and power levels. Also if a speaker is playing only one frequency say well within its primary intended range, it has one sound but if another frequency that might be outside of its normal range is present the first primary note reacts differently. You may not even hear the second out-of-range note, but it changes the stiffness or flexure of the cone which in turn modifies how the primary note comes out. Any portion of what makes a sound output, be it an amp, a speaker, a transistor, the room treatments, or whatever, all make up the combined sound and thus its perceived sound. We may not be able to receive a particular sound or tone, but we can recognize the effort that such a sound causes on the rest of what we do hear. A different way of putting it might be, A Note that is far beyond the human hearing range affects the overall quality of frequencies/sound which we do hear.
Can you measure willingness or enthusiasm? Are you in the mood for music and it sounds better? Are you in the wrong frame of mind and it just doesn't gain your interest? Can you measure that?
Just about any decent amp measures better than almost any tube amp of almost any price in most all of the typical parameters. Look at the square wave reproduction for one good example. Does that mean the solid state amps sound better / more like live music? We all know better.
That said, I’m not sure John Atkinson of Stereophile does as rarely has an overall good thing to say about tube equipment & if you believe him & his “bias”, Benchmark products should sound better than anything else at any price as they supposedly test that way.
The conventional performance measurements of hi fi are summative and the wood does not necessarily describe the trees. Walking through the wood is a different experience to looking at if from an helicopter. Neither experience invalidates the other but the two are valid in different ways.
@erik_squires+1. It’s called Occam’s Razor. Been around for a while. Was the basis of the scientific reasoning that created the tools that take the measurements. Now that we have the tools, we forget about the principals. It would be funny in any other age.
I think the complexity of active listening is such that the technology used to measure just can’t catch up. How many million tests would be needed to capture this? Add to that, the relationships between tests (i.e. poor results at a rate against good results from a different test at a rate) and I think if any two devises test exactly the same, they weren’t tested enough. Combine this with the fact that, as others have said, the tests are being performed on the wrong side of the eardrum, I consider measurements to be nothing more than a baseline start to inform an evaluation. Will the technology catch up in a few hundred years - maybe — but It is the curse of every age to consider themselves sufficiently advanced and ‘done’. 50 years from now someone will be laughing on this forum about the crudeness of our testing processes.
i have tested high voltage current at hospital facilities. 13.3k volts. Everything tested fine. Dozens of tests using specialized equipment costing more than the most expensive audio equipment. Then, through an analog waveform capture on the feeders, and the B phase is flat coming in from the utility. My point being that every test is a snapshot of a piece of a photo. The relationships between the tests are what get you, and an analog experience (like listening) is the easiest way to get that.
Is there agreement and proof of each measurement’s value and the range of human sensitivity?
Is there any proof otherwise? The answer is no.
It doesn’t work that way in science. We do not make someone prove that pixies do not exist, in the same way that we do not make people prove that they’re innocent.
The proof is upon the claim maker, or prosecution, to prove guilt.
What are the relative merits off each measurement in terms of a broad range of listeners as well as you , specifically?
Is there any proof otherwise? The answer is no.
It doesn’t work that way in science. We do not make someone prove that pixies do not exist, in the same way that we do not make people prove that they’re innocent.
The proof is upon the claim maker, or prosecution, to prove guilt.
But the merits of the measurements are documented in many places.
For instance the high distortion of tube equipment being more pleasing to the ear than much lower distortion of some SS gear.
So yes the relative measurements are not all created equal… and some measurement are harder to do than others.
We lack measurements which can take into account the ear/brain mechanism as well as self-training of the neural pathways.
Irrelevant as it pertains to preference, not to audibility, which was covered by your first two points, and I will point out again, none proven.
In some sense, if we know the sound field, like a measurement of it, then we could assume that a reproduction of the field should look the same in the time and frequency domain. And the more “sameness” would be higher fidelity than less “sameness.”
Take a set of published tests from 1972. Now compare them to published tests from say ASR in 2022. It would be dishonest to claim that the 1972 tests are nearly as comprehensive as what is and can be done in 2022. I think the thing mainly stuck in 1972 is audiophiles, not the measurements
Ok lets take 1982, then.
We had many more measurements happening in the 80s.
And so, if put to practice, any 2 digital audio components that have similar enough measurements should sound identical. For example, a DAC with a SINAD or SNR or 120 dB vs one with a SINAD or SNR of 123. Tiny differences in linarity and frequency response above 20 KHz are not audible to us humans anyway.
I am not sure SINAD is all the most useful measurement.
If we are playing tones, then it might be, but we are playing impulsive thinks like drums and other percussion instruments as well.
Secondly if the SINAD is all 2nd harmonic, or all 5th a harmonic, will we hear a difference?
But at some point do the speakers matter? And their contribution to SINAD, or SNR, or distortion products?
asvjerry, I doubt your proposition could be proven right or wrong. If I knew something was 0.0000000003% THD does that mean I can hear it, does it mean I believe I can hear it, or does it mean my expectation is I can't hear it. As probably the amplifier and certainly the speakers have more than 0.003% THD then what distortion can you expect to hear 0.3%. 0.03%. Investigation seems to agree that we could hear 0.5% but not 0.05%, so as we cannot possibly hear 0.003% why would we ever have such an expectation?
ASR recently gave a DAC a junk rating because of 1 reading, distortion at 0.003%. Can anyone hear distortion at 0.003%, no. Tube amps with 0.5% distortion are considered good, turntables would be consider at least OK, speakers are often stated as less than 1% distortion. As a measurement predicting poor performance quoting 0.003% is totally useless , to use as the basis for allocating junk status is ridiculous. No one can hear it, so an average CD or other DACS measure 'better", this is analysis paralysis.
At a point in time not all that long ago, various items...amps, pre's, and various 'outboard' items like eq's all began to 'measure' relatively the same in the 'heard' or sonically perceived range of human hearing.
Beyond that range, one could argue that 'felt' or physical stimulation of ones' presence in a sound field could be added to that perception of 'being There...subsonics for the gut grab, supersonics for 'air', the 'larger' quality of a larger space than actually exists for the listener.
This mix, delivered by equipment designed in various combinations by various groups/companies in wildly varying combinations in even more variable spaces.
MHO?
All that's left is our personal psychoacoustical response to whatever we've amassed in our sound caves...
The only measurement left is the rulers in your head.
Now, not that I'm against anyone anywhere at any time spending whatever they find apropos to scratch their particular itch. Far from that...
Overall, the lucre spent on that pursuit generally 'improves the breeds' for the masses to enjoy in the long run. But, as pointed out frequently, try to enjoy the music even when deep into 'critical listening' of the perceived shortfalls, huh...?
"And now you just described the science of human perception and psychoacoustics, which seeks to understand how dissimilar two items can be from a definable standpoint (measurements), and still sound the same."
All measurements are science. The problem is whether the measurements done are complete. The Stereophile measurements are perfectly good. But they are not sufficient to describe the sound produced. I've seen measurements that sometimes give good clues to the sound. They just aren't the classic ones used in ads or in Stereophile. As an example I once had a friend select a cartridge based on a frequency sweep, a separation curve and a 1 kHz square wave with no clue to which cartridge it was and it turned out just as he predicted. But this was one case and he couldn't do it in most cases. He got his clues because he had modified cheap Grado pickups and done extensive measurements on what he was changing.
I’m beginning to think jack isn’t really a hifi guy. I’m beginning to think he’s paid by ASR. Who are you trying to convince? You’re not. You’re just trying to wind people up.
If only we knew everything to measure, and the subjective-weighted impact of each, and we measured each and every one.
if oyu think that’s extreme, the subjective-weighted part was the Bell Labs approach int he 1930s and resulted in A and C message weighting.
Those who purport ot know all are either charlatans or just a little naive.
I sure wish i could make easy measurements and know all i need to! would sooooooo simplify my life.
As a researcher in real electronics i face very similar problems every day and need to fight the "We know it all and know better" of the rank and file. And more importantly "what i know is my security - don;t threaten it". Dont move my cheese. I’m paid very well to shake the tree.
While measurements are very helpful, they are not 100% sufficient. And they certainly do not take into account individual tastes. Make no mistake, many highly respected components have significant distortion - but almost always euphonic distortion of one flavor or another. Before you revolt at that, example A is a Steinway piano, who's rich sound comes largely from the harmonic distortions of its sounding board.
When I test drive a car, specs take a back seat... (get what I did there?). Headroom is a spec that matters, I’m a big guy and don’t fit in lots of cars. 10ths of a second in 0 to 60 time, not a decision maker. It’s not a real world spec that would determine the affect of my daily driving or my enjoyment at the track. Cars are a perfect example of what I’m trying to portray because when I test drive a car the range, MPG amount of people it seats has been determined so I’m looking for cars within a specific class. Sure some specs matter but there is no spec for feel. How does car A feel as opposed to car B? That feel spec is what matters but it doesn’t exist. Feel is engineered in by a team working with measurements to create that unique flavor there is no outside group reporting those measurements.
They’re testing the wrong stuff, Class of vehicle or component sure 3 watt amp. 84db speakers are are important measurements but the voicing of equipment isn’t being portrayed with the mostly inaudible specification ASR is making their recommendations on.
What I’ve written has to do with common sense. What you’ve written has to do with exactly the same drivel that every person who believes in blind testing says. That, and some of my words in there for good measure to make yourself sound cohesive.
You know I’m right. Just like test driving a car, trying on a new suit, or visiting an open house.
Once you see/hear/experience for yourself, any preconceived beliefs you had about whatever it is will be crushed. You will only be left with how to piece together the reality or real world experience you just had.
So here is an example:
Dealer tells you about a system. You do research for a few days. You are super excited.you now have some preconceptions about this system sounding great with say, rock music and complex instrumentals because you were told so.
But then you go and listen. To your disappointment, the dealers description does not align with what you just heard.
You don’t need a blind test to confirm that your experience is true.
Same thing with any product or service you can imagine.
Sound itself is intangible but it can either impress us or leave us disinterested.
There is no middle ground or basis to your arguments.
Wudabout to things that sound alike, but measure differently?
And now you just described the science of human perception and psychoacoustics, which seeks to understand how dissimilar two items can be from a definable standpoint (measurements), and still sound the same.
I never suggested test driving a car while using a blind method of testing.
Of course that would be extremely dangerous!
When you keep a process for testing simple - like test driving a car for example, you eliminate variables that would otherwise reduce the accuracy of the test itself.
So while one person may like the way a Mercedes looks, after test driving they make an informed decision about the handling, acceleration, interior quality, etc.
On the other hand, that same person could then go to a BMW dealership and pick out their new car at that dealership instead.
The simple fact that we know what is being tested does not reduce our hearing facilities or auditory memory. We are psychologically okay providing that we don’t allow the opinions of others or the spec sheet of the car get in the way of our real world experience - driving it.
There is bias in everything. Unless of course we are intellectually honest with ourselves. My method of listening allows a person to be absolutely certain about their listening impressions.
You listen, have a reference point, then revist your findings later and confirm them.
My method is repeatable and requires no further reductionism - because it is simple and straightforward, unlike blind testing which can become complex.
That makes perfect sense actually, just like test driving cars.
Blind testing is flawed and you know it - especially for audio.
Like wine tasting, some people can correctly identify which wine they tried.
Even at a later date. Australian red wine vs Italian red wine for example.
Like I said, way too many variables in blind testing for audio. It’s idiotic.
The only way I could know for sure which is better for me would be to listen to both side by side but by asking some Dave owners why they spent the big bucks over the Hugo would surely sway my opinion over strictly numbers.
I just bought a used PSAudio DirectStream from a person that replaced it with a Dave. I didn’t get to hear it but he believed it sounded much better that the DS. He didn’t buy a Hugo either time.
ASR measured the DirectStream and it didn’t do well at all... (failed). Just my opinion and mine alone but I personally only care about how it sounds and to me sometimes late in the night, I’m just stunned by it.
To me the only "measurement" that matters is will it fit on my rack and IT DOES! The rest is just symbols on a piece of paper that is trying to reduce art to "Dynamic range, multi-tone, thd, headphone output power, output power of the analog/balanced stages, frequency response, and linearity", Important maybe but not what music reproduction actually is.
Seems like you’ve got more "reputation" than Taylor Swift’s "are you ready for it?" Album. When you are cornered like this on a forum you are effectively "dancing with your hands tied," and may need a "getaway car" but don’t blame me!
Joking aside, that is an exceptionally weak answer.
Just because we have elicited one our senses (sight)
That does not mean our results are valid. At some point, you would have no choice but to resort to guessing. If you get 9/10 right, that is like scoring the same on a multiple choice test when you guessed.
Bias is also found in blind listening test because:
1) we are aware a blind test is being conducted and we are the subject
2) we want to be right, which clouds our judgements
3) we actively believe that something infallible is being conducted
4) we may easily confuse ourselves without other reliable senses (sight)
5) our conclusions will end up being s bunch of half truths and guesses.
For this reason, any listening session should be conducted like this:
1) listen to the gear
2) take notes
3) listen to another piece of gear
4) take notes
5) wait a day or two (as long as you need to forget)
6) compare the notes
7) audition the systems again to confirm that your notes align with your listening impressions
So I the same way that some people have an excellent visual memory,
We audiophiles have an auditory memory. If we use it wisely, we hear differences.
Freak, what is your obsession? Did you dog die and you are fixated on me now? Get a life loser.
No. I just can’t understand your need to come back here over and over after being banned for at least 15 times. Maybe more. Maybe, just maybe, YOU have an obsession issue?
My advice: refrain from engaging this "audiomaniac" dude. And if you feel that you have dealt with the dude before, like a deja vu, despite the fact he just "joined" Audiogon last week, you are not wrong. You have. With his previous 15 usernames, now all banned. Here is the full list:
But beyond these measurments, how can we conclusively prove that the DAVE is better than the Chord Hugo TT2?
Controlled blind listening test, with obvious level matching which will show they are different, and if done as a preference test, then better for that single person doing the test.
The ASR test of the DAVE shows problems on one channel. Is that endemic or single unit.
I think your analogy would be more suited when discussing the longevity or lifespan of something. For instance, do power conditioners help components live longer without intermittent failures?
Back to your ending statement - ASR is not measuring the right stuff.
I’m no advocate or fan per say, just want to know the truth.
Dynamic range, multi-tone, thd, headphone output power, output power of the analog/balanced stages, frequency response, and linearity.
Please excuse me if I missed one or two. But beyond these measurments, how can we conclusively prove that the DAVE is better than the Chord Hugo TT2?
Perhaps you can shine some light on this. Are there measurments beyond a null test that can find out if one DAC has greater instrument separation and detail?
You must have a verified phone number and physical address in order to post in the Audiogon Forums. Please return to Audiogon.com and complete this step. If you have any questions please contact Support.
