Testing audiophile claims and myths

[EthanWiner]

Quote:

Not sure how you would measure timbre and get a completely meaningful value. Instrumental separation is how well the instruments are well not smudged together, but each have their own unique voice. Basically my point is that there is a whole lot going on and one cannot reproduce what one hears with measurements and them make any sense if it all. I'm a professional engineer and we love to measure things, but I think this is one area that can't be fully quantified. Ultimately, you just have to listen.

Timbre - basically frequency response - can definitely be measured. For gear you measure the response the usual way with sine waves etc. For musical instruments you can record them and use an FFT to see the inherent resonances. But you are confusing psychoacoustics, which I have not addressed in this thread, with assessing audio gear. If someone claims that wires or a solid state circuit can "break in" and change the sound over time, that is simple to measure and prove or disprove. And that has nothing to do with human perception. What you describe about instruments being "smudged" together is mostly about EQ choices by the recording engineers, and also the ear's susceptibility to the masking effect.
 
--Ethan

 

[EthanWiner]

Quote:

[too much formatting to quote in and out 10 times]

Earlier you asked:
> As to why people hear a difference with ultra-high sampling rates, you posit comb filtering/room
> acoustics as the most likely reason, but what about headphones?
 
Yes, headphones avoid comb filtering. I never said small positional changes are the only reason people think they hear a difference when none can exist. In fact, most of the time I believe it's simply due to faulty perception! But comb filtering shows that the sound really can change at our ears, even when nothing has actually changed in the signal path.
 
I agree that some SACDs are recorded with more care than other formats. But that's no reason to waste 4x the bandwidth. The obvious solution is to not over-compress at all, then use regular CDs.
 
You also asked:
> "You just need all the data." How do you know you have all the data?
 
I already answered that 5 times now. Null tests. Enough already. Now back to the present:
 
> Here's a story you may find interesting and which you can hear for yourself
 
That's old news, and has been debunked many times. That Michael Fremer and the others do not understand just shows that they in fact know less than me about audio. Here's the scoop on that Neve / Emerick story:
 
I'm sure that console channel sounded different, but not because Rupert Neve or Geoff Emerick were hearing 50 KHz! When a circuit oscillates it creates hiss and "spitty" sounds, and distortion in the audible band. So obviously that's what they heard, not the actual 50 KHz frequency. And what studio loudspeakers reproduce 50 KHz anyway? None. I don't care how many famous acts someone recorded. It doesn't mean they understand the science. In this case clearly they do not.
 
--Ethan
Yes

 

[upstateguy]

 
Quote:

1. I wish you had responded to the post I actually addressed to you, placed directly after the post you responded to. There, I mentioned a few issues I had with statements you made in the video.
 
2. No, because with PCM there is a point where you hit diminishing returns. 24/96 is better than 16/41, but 192kHz is not better than 96kHz with today's technology. 192kHz is apparently plagued by all sorts of problems. See:
 
http://www.soundstage.com/index.php?option=com_content&view=article&id=126:96khz-vs-192khz&catid=57:reader-feedback&Itemid=24
 
However, with DSD, you can theoretically go much, much higher.
 
3. Here's a story you may find interesting and which you can hear for yourself if you follow the youtube link I provided ("Deep Listening: Why Audio Quality Matters"):
 
"Geoff Emerick recorded a lot of Beatles material. He was working in his studio at the Neve desk (named after English electronics engineer Rupert Neve). It's a very famous desk. Emerick had an issue with one of the channels and he kept calling technicians in, and people kept coming in and saying there's nothing there. But Emerick kept saying that he was hearing something, and everybody said no, it's impossible, we've looked at it and don't hear anything. Neve came in and looked at it and they analyzed it with some other piece of equipment that hadn't been brought in. What Emerick was sensing, he wasn't hearing it, he was sensing something in one channel that was up around 58,000 cycles. He didn't hear it, but he felt it, his body felt it. It's a perfect example of how human beings perceive sound. We don't always hear it, but we feel it. [Red Book] CDs stop us from feeling it...SACD, with a much higher sampling rate, is completely different. With SACD, we can feel things." -- Craig Street, Record Producer (produced projects for Norah Jones, K. D. Lang, Cassandra Wilson, Chris Whitley, John Legend, Gypsy Kings, among others).
 
4. Admit vs. explain? The fact is that you agree that not all amps sound alike. If they don't all sound alike, then why do you say things like, "Audio critics can hear differences among amps only as long as they are looking at the label." If there are differences, as you admited -- or explained -- then why wouldn't they be able to hear differences among different amps?
 
5. In response to your call that I defer to people who know more about audio than I do, I can tell you that I have. Not all professionals agree with you. Again, if you watch the youtube video I linked to above, you can hear a panel of experts tell you that you are wrong. The panel includes Greg Calabi, managing partner and mastering engineer at Sterling Sound in NYC. He has mastered the music of John Lennon, David Bowie, Bruce Springsteen, The Ramones, Talking Heads, Patti Smith, Paul Simon, James Taylor, U2, Norah Jones, Bad Brains, The Beastie Boys, John Mayer, Emmylou Harris, et. al. He has mastered more than 6,500 albums. Do you want to claim that he doesn't know anything about audio? Well, watch the video and hear him pontificate against rbcd and praise the virtues of sacd.
 
Another member of the panel who praises the superiority of sacd and high resolution in general over mere cd resolution is recording/mixing engineer Kevin Killen. He has handled the music of Peter Gabriel, Elvis Costello, Kate Bush, Jewel, Bon Jovi, U2, Bryan Ferry, Lorenna McKennit, Duncan Sheik, Shakira, Sugarland.
 
Another panel member is Steve Berkowitz, Senior Vice President of Sony Music's legacy recordings. He's in charge of rereleasing classic recordings at Columbia records. he has worked with Bob Dylan, Tony Bennett, Leonard Cohen, Earth Wind and Fire, Michael Jackson, Miles Davis, Branford Marsalis, Fishbone, John Mellencamp, Jeff Buckley, Ministry, The Cars, et. al. I suggest you watch the video. It's two and a half hours long, but these experts totally contradict what you say about high resolution.
 
In short, just as you ask me to give way to the knowledge of people who know more about audio than I do, let me similarly request that you give way, or at least consider, the knowledge of those who may know more about audio than you. All the professionals in that particular panel work mostly with pop, rock, and jazz, but if you prefer to appeal to professionals working with classical music, I can also appeal to a number of highly knowledgeable audio professionals in that area who would totally contradict your assertions. Many classical music recording engineers vehemently argue the virtues of sacd and high resolution sound in general as opposed to standard cd resolution.
 
"Deep Listening: Why Audio Quality Matters" can be seen both on youtube and the philoctetes center website:
 
http://philoctetes.org/past_programs/deep_listening_why_audio_quality_matters
 
http://www.youtube.com/watch?v=SY5hI98HEi0
 

 
Don't get taken in.  Profit is clearly their motive.
 

 

[Prog Rock Man]

This becoming a bit black and white, so to clarify the position with regards to amps
 
 - all amps do not sound the same as ABX testing finds reliable differences - clicky
 
 - the differences however are no where near as big as audiophiles and makers like to make out - clicky
 
 - all amps can be made to sound the same by equalising and line leveling the volume - clicky
 
Each clicky links to a test from the opening post in this thread.

 

[b0ck3n]

Thanks for helping me make sense of it. Are you aware of a set of headphones that requires +1W outputs? Most small portables seem to range from 50-250mW, the iBasso PB2 2500mW - seems gimmicky to me.

Yes, it is twice the power of the same amplifier in single-ended mode.  But it's not necessarily difficult to build a different amp with the same power, so that's why the power advantages are rather dubious.
 
I do wonder, however, if the portable balanced amplifiers are perhaps the easiest/most efficient way to put out the large 1 W plus output that the most powerful ones are capable of.  I mean, the iBasso PB1 gets more than 20 hours of battery life on high gain (about 24 V peak-peak) driving my HD 600, and it's tiny.
 
The difference in distortion actually comes down to two things - in a speaker amp, you're running at low impedances already, and when you bridge an amplifier you halve the effective impedance of the load that the output devices "see".  That actually causes an increase in distortion (why you'll see amplifiers quote increased distortion when bridged), but in an amplifier designed to be used solely as a bridged/balanced amplifier, that probably isn't so much of a problem.
 
The main thing, however, is very real and quantifiable - even order harmonic distortion products cancel out to the degree that the positive and negative halves of the amplifier are matching.  Usually it is reduced to a very low level, but it does nothing to reduce odd order harmonic distortion.  You may remember that tube amps generally produce quite a bit of even order harmonic distortion, and that it is considered far less offensive (or even desirable) to have even order harmonic distortion (especially the second product) than odd order harmonic distortion.  Anyway, the level of harmonic distortion in modern amplifiers is very low to begin with, more or less inaudible.  So the actual value of the even order harmonic distortion cancellation is rather dubious.
 
Also, to confuse you even more, most headphone cables themselves are already balanced.  If your cable has both lines for both sides running all the way to the plug, it already provides common-mode noise rejection.  There's no benefit to be had to run a balanced amplifier if that's your goal, and even so, such noise rejection is rather meaningless for anything but very low level turntable signals.  Headphone cables, speaker cables, and line-level signals generally have little to no issue with noise unless you're running them right along a power line or perhaps right in the vicinity of a powerful radio transmitter.
 
 
So to me, yes, the claims here for improvements in balanced operation seem to be little more than the typical fare you see in certain sub-forums...  And I say that as an owner of both push-pull and single-ended headphone amplifiers.

 

[BlackbeardBen]

Quote:

Thanks for helping me make sense of it. Are you aware of a set of headphones that requires +1W outputs? Most small portables seem to range from 50-250mW, the iBasso PB2 2500mW - seems gimmicky to me.
 

 

Well, there are some extremely demanding headphones - the AKG K1000, and the Hifiman planar magnetic headphones (among others of the type) - that require far more current than other headphones.  They're low impedance and low sensitivity.
 
High impedance headphones don't require as much power overall - but often, they need as large or larger voltage swings to avoid clipping.
 
Having a high impedance means that less current is needed to reach a given potential (voltage), so high impedance headphones don't generally require more power - but they can utilize the high voltage of such amps.
 
Take, for example, a 50 ohm headphone versus a 600 ohm headphone.  For argument's purpose, if they have the same sensitivity (in dB/mV, not dB/mW), you'll get the same volume from the same voltage swing.  But, it takes 12 times as much current (and power) to create that voltage swing on the 50 ohm headphones!
 
However, exactly how much power you need to avoid clipping at loud listening levels isn't exactly as easy to calculate (thanks to the dynamic nature of music).  For many people it's easier (and more comforting) to know that they have far more power than they'll ever need.  Look at the Schiit Lyr, for example.

 

[xnor]

But the Lyr is ridiculously overpowered for normal headphones. With many headphones a few milliwatts are enough to produce high sound pressure levels.
I'd take a small amp with just enough (clean) power which also allows me to use most of the volume control's range over something bigger with higher gain and probably noise etc. any day.

 

[MacedonianHero]

Timbre - basically frequency response - can definitely be measured. For gear you measure the response the usual way with sine waves etc. For musical instruments you can record them and use an FFT to see the inherent resonances. But you are confusing psychoacoustics, which I have not addressed in this thread, with assessing audio gear. If someone claims that wires or a solid state circuit can "break in" and change the sound over time, that is simple to measure and prove or disprove. And that has nothing to do with human perception. What you describe about instruments being "smudged" together is mostly about EQ choices by the recording engineers, and also the ear's susceptibility to the masking effect.
 
--Ethan

Let me try it this way then. I agree that timbre of a particular note can be "measured", but listening to music, there's a whole lot of notes being played at the same time....so making sense of the data would be next to impossible. Instrumental separation is a function of not only the recording, but source, amp, headphones too. Ultimately it is how the particular gear sounds...measurements be damned so to speak. Sometimes we get too hung up on how gear measures and we forget that we listen with our ears.

 

[TakashiMiike]

Quote:

Sometimes we get too hung up on how gear measures and we forget that we listen with our ears.

 

If we listen with our ears why has not one person passed a blind test on cables, once their eyes are removed from the equation? We listen with our brain, our fallible brain, step up and take some blind tests.

 

[CEE TEE]

[size=medium]

Hi Ethan,
 
Thank you for having a discussion with us!!    I watched your link on YouTube and it was really interesting.  
 
I know that during my previous photo career I had tools around to sometimes help me with the color cast of my images...either changing ambient conditions or fatigue or habituation could lead me to check myself.  I had a GATF/RHEM light indicator on my monitor to let me know when ambient light around my monitor wasn't 5000K degrees Kelvin and would use the color tool in Photoshop to see what the actual "numbers" for color were since our brains adjust and our conditions are not consistent.  (CMYK or RGB values- is there a color cast?  Find a grey/neutral area and check the numbers- is R or G a larger number or are RGB all equal?  Remove the cast and all of a sudden you see another "version" of your image that seems "correct" too!)  Trying to control an image from capture to display as intended is difficult.  The actual color of that J.Crew sweater you are shopping for won't look the same on your monitor.
 
I know I'm using audiophile jargon but could you speculate on what might cause a "warm" or in the extreme- a "smeared/bloomy" sound in a DAC/Amp?  Can almost sound like a "reverb" or "fast echo" to me if that makes any sense whatsoever...I know my description isn't related to a good technical understanding!  How do we talk about this?
 
 
 
I am also trying to understand what "neutral/flat" or transparent sounds like.   I understand from your video that it can be defined as "flat from 20hz to 20khz with little decibel deviation" and low distortion with good timing.  If I have iems that bypass my pinnae, and they have a flat  20-20 FR graph (no third-octave averaging/large decibel divisions in graph) is that the only way to know for sure that the gear is "transparent"?
 
What is the best way to hear/understand "transparent" in a headphone system or to learn the Frequency Range to be able to communicate?  Frequency sweeps?  Test tone CD?  Sheer memorization of instruments and ranges?  Should I go through the "equalizing my phones" process or play with an EQ to understand what/where things are in the frequency range (by frequency and by decibel amount)?  My mastering/musician friends then start to talk about "q" and parametric vs. graphic EQs...

 
Thanks for any help.  I'm trying to test my understanding, learn, and communicate better.
 
CEE TEE
​

[/size]

 

[Prog Rock Man]

Quote:

....... Sometimes we get too hung up on how gear measures and we forget that we listen with our ears.

 

The irony is that those who do pay attention to the measurements and the information we get from them are more likely to be listening to the music with their ears alone. We are also far more likely to concentrate on what really makes a difference to sound quality, room acoustics, speakers etc.

 

[nick_charles]

Quote:

but listening to music, there's a whole lot of notes being played at the same time....so making sense of the data would be next to impossible

Not really, the process is exactly the same as a simple sine wave ! - the complex wave addition creates a complex wave but its characteristics are easily decomposed into individual frequencies and amplitudes, if you could not do this then psychoacoustic coding would be impossible, literally impossible, a FFT works just as well on both simple and complex.

 

 

 

. Instrumental separation is a function of not only the recording, but source, amp, headphones too.

How exactly does an amp create more or less instrument separation? How does it know where the instruments are in space and how to separate them? Does it filter the bands and then selectively attenuate some? If so it is seriously flawed ! - what is it doing to create this illusion ? -

As I understand it crudely, the key to the illusion is in the difference in levels and timings and channel placement of individual wave components, if you have two cellos playing the same note and one is 3' behind and to the side of the other then there is a small but definite difference in both timing and level, the speed of sound is finite and distance decreases level, we use these cues to recreate the illusion of separation, that works even for mono. Then we have the stereo effect of using multiple mics which have slighly different inputs, then one cello appears more to the left than the other...the only way an amp can impact this last bit is if it has terrible channel separation , but that would decrease the illusion of space !

 

 

 

 

 

 

[maverickronin]

As far as I can tell instrument separation appears to be a function of distortion.  Low distortion (or possibly just very even levels of distortion without large peaks) allows you to hear those subtle differences in levels and timing.  I don't have any real data to back that up other than comparing what I hear from a headphone with measurements of that same model of headphone.
 
I'd assume the same would hold true for amps as well.  Besides that, some people say that tube amps somehow enhance the soundstage but I haven't really noticed that with my DIY Bottlehead Crack and my HD650s as compared to my transistor amps.  I'm not really sure how that would actually be possible though.  The closest thing I have to a hypothesis is that the extra 2nd order harmonics are somehow similar to a reverb effect but it doesn't seem too likely to me and I don't have any idea where to start investigating it in the first place.
 
Feel free to shoot all this down if you want since its really just idle speculation.

 

[EthanWiner]

Quote:

Let me try it this way then. I agree that timbre of a particular note can be "measured", but listening to music, there's a whole lot of notes being played at the same time....so making sense of the data would be next to impossible. Instrumental separation is a function of not only the recording, but source, amp, headphones too. Ultimately it is how the particular gear sounds...measurements be damned so to speak. Sometimes we get too hung up on how gear measures and we forget that we listen with our ears.

Again, you are confusing psychoacoustics and perception with the specs and tests used to assess the fidelity of audio gear. Further, it is mostly untrue that "instrument separation" (whatever that really means) is a function of the gear. A piece of audio gear will affect clarity only if its noise and distortion are high enough to mask the music, or if its frequency response is skewed enough to reduce important frequencies enough to notice. So an amplifier that's 10 dB down at 5 KHz and above will surely sound muffled. To say "measurements be damned" only shows ignorance and maybe even arrogance. I'm sorry, but I don't know a nicer way to put it. Thankfully, the people who design the gear you enjoy don't have that attitude!

 
--Ethan

 

[EthanWiner]

Quote:

[size=medium]

I know I'm using audiophile jargon but could you speculate on what might cause a "warm" or in the extreme- a "smeared/bloomy" sound in a DAC/Amp?  Can almost sound like a "reverb" or "fast echo" to me if that makes any sense whatsoever...I know my description isn't related to a good technical understanding!  How do we talk about this?
​

[/size]

The way to talk about this is for both of us to use the same terms. These terms must have a universal meaning that everyone understands. So that leaves out "warm" and "smeared" and "bloomy." If you are hearing reverb or echoes that tells me either 1) you have a receiver set to some "surround enhancement" type mode, or 2) the problem is with your room. I can't help you resolve these terms because it requires being there in person as you play examples. I'm sure I could tell you the correct terms for what you hear if I was there.
 
Quote:

I am also trying to understand what "neutral/flat" or transparent sounds like.

 
For that you'd have to visit me, or find someone local with a good system that's set up properly in an excellent room. Are there any hi-fi clubs in your area?
 
--Ethan