Darko Audio podcast - Beware the measurements

[pozz]

Thankfully most of this is absent on the pro audio side where most customers are more pragmatic and won’t spend dollars on snake oil.

To some extent. Audio is so messed up that even pros fall for that same kind of stuff.

 

[Archsam]

Given that many use digital image processing to create horribly unnatural "high impact" images that look like a motivational poster on meth... there is some similarity to audio.

There really is a lot of parallel between modern consumption of music and images, now that I think about it.

Instagram = Spotify

Both are very popular content driven apps intended to be used on mobile devices. Both uses compressed lossy media format (OGG and Jpeg), both are not meant to represent SOTA level of quality but are more about popularity of the content and connection of like-minded people on social media.

Most people are now more than happy with the image quality of their phone's selfie cam. You can apply all sorts of filters to create 'cool looking' and 'bold' images which will look good on FB and Instagram, but will not likely to look impressive once you open it in Photoshop and zoom in past 100%.

Meanwhile I'm holding on to my DSLR and I'm on ASR moaning about the state of hifi today. I think my middle age angst is kicking in

 

[Xulonn]

Definition of "fidelity":
3. the degree of exactness with which something is copied or reproduced.

Definition of "coloration":
3. subtle variation of intensity or quality of tone

Definition of "audiophile":

1. a person who is enthusiastic about high-fidelity sound reproduction

"Fidelity" or "exactness of reproduction" can be accurately measured. We can measure variations in sound better than we can hear them, but predicting the subjective perceptions and reactions to those variations, in many cases, is very difficult to predict. "Coloration" in audio is, by definition, a type of "distortion" of an audio signal and can also be measured.

All audiophiles enjoy "high-fidelity", but those in a certain subset of audiophiles are ridiculed because they obviously listen more to subjective "sound quality" and not so much to the music. Musicians tend to do the opposite, and professional musicians are not a large segment of the world of audiophiles. Many professional musicians do not even own expensive "high-fidelity" audio systems for home listening.

We seem to argue endlessly here at ASR about "fixed" vs "variable coloration". The subject PS Audio phono amplifier is from a company known for its b.s. audiophile snake-oil products. Since it was apparently not designed to measure well, it must then be adding "color" (a type of "distortion") to the source signal. If I want to "color" my music, I want that option to be able to be turned on and off at will - and variable in its parameters.

With respect to distortion in music recordings, it begins at the front of the audio-recording chain. "Raw" audio signals are created by an electro-mechanical transducer and amplifier, and then run through various recording/mixing/mastering stages, which are also not distortion free. "Fidelity" varies greatly from recording to recording, and even supposedly excellent "digitally remastered" recordings can be worse than original analog releases. Any coponent in the signal amplification chain that attempts to use a "fixed" set of parameters to correct distortions contained in one recording is likely to "amplify" the distortions that are present in other recordings.

Philosophically, I am firmly in the camp that the ideal solution is to amplify the audio source signal as cleanly as possible, and use variable digital signal processing to "color" or "flavor" to taste.

I prefer a balanced, neutral amplification chain where well-recorded music sounds great in a room that has been corrected at least minimally via DSP. Indeed, I can usually tolerate listening to truly good music that is poorly recorded rather than wasting time and energy trying to correct flaws in bad recordings. (I simply enjoy the music and whine later on audio forums about the sound quality.)

However, as others have said at ASR and elsewhere, if one is secure in knowledge of their preferences, for example, preferring a "warm coloration", then choosing components that make a warmer-sounding audio system makes sense.

 

[MattHooper]

Colored sound isn't hifi

If only it were that easy ;-)

First, note that indicated even in the link you posted that the road to hi-fidelity had the impetus of producing more natural, realistic sound - "faithful to the original sound" - e.g. a singer, violin, orchestra etc. In the Wikipeda link it tells us how someone like Avery Fisher drove the movement towards Hi-Fi:: "He wanted to make a radio that would sound like he was listening to a live orchestra—that would achieve high fidelity to the original sound." This is one reason why live vs reproduced demos were part of claimed advancements in sound quality. But the notion of Hi-Fi still carries with it the association with "the QUALITY OF THE SOUND" (subjective impression of Sound Quality, what it "sounds like") - as the Wikipedia link suggests: "high-quality reproduction of sound.[1] This is in contrast to the lower quality sound produced by inexpensive audio equipment, AM radio, or the inferior quality of sound reproduction that can be heard in recordings made until the late 1940s."

"Sound Quality" "Quality Sound Reproduction" has, from early on, been associated with comparison to known sounds like what a human voice really sounds like a violin, drums, orchestra, whatever. Here for instance is an old RCA film introducing and explaining the concept of Hi-Fi (and hi-fi equipment):

Quotes from the film: (Hi-Fi is) "The truest, most life-like reproduction of the original music possible. Or "perfect fidelity. Or..."Hi-Fi"

"Hi-Fi is sound reproduction from an instrument (amp pictured) that has been especially designed and constructed to give you sound that will be as nearly as possible the same sound you would hear in the original playing. What you get in a Hi-Fi set is best described as 'Presence,' a feeling of life and nearness in the music. Once you've heard hi-fidelity reproduction you'll be hard to satisfy with anything less!"

So the point is that "accuracy" in the technical sense we tend to use the term here, and which is being used almost interchangeably with "Hi-Fi," arose as a means to an end, not as the end itself: the end being the Sound Quality - of natural, realistic sound...or that at least produced certain desirable aspects of live music. And you can see from the way Hi-Fi was sold that it was generally done so in terms of the subjective impression of the sound, rather than appeal to measurements and numbers.

Now, we can say "I am going to make this less messy and define "accuracy" as equipment that meets certain measurable parameters, and I'm going to define Hi-Fidelity Sound as a signal played on such equipment" which is fine and useful. But that is different than thinking that Hi-Fi is always understood this way, or that it arose to mean that goal, or has always been sold on this meaning. If for instance a certain coloration actually makes a voice or instrument sound more natural, more like hearing the "real thing" that fits pretty well with the original goal of "Hi-Fi."

And as instrumental and production palettes expanded, notions of Sound Quality spread to accommodate, so while it would maintain that a vocal with excellent Sound Quality would usually mean one that sounded natural and more like the real thing, I'd say other common aspects are things like "richness" (as opposed to "thin" sound), "smooth" (as opposed to gritty and unpleasant - artificial artifacts one doesn't hear from most real voices or instruments), "clear" (as opposed to muddy sound), "detailed" (as in revealing life-like or rewarding complexity in the sound), "dynamic" (able to reproduce the sense of impact, energy, presence or excitement of either real musicians or electronic sound), etc.

That "Hi-Fidelity" is not simply and only about pure electronic accuracy is shown by the fact that demos of "Hi-Fidelity" audio equipment typically wish to use tracks with High Sound Quality (the subjective aspects we value). Given the choice between demoing a Hi Fidelity system with a track that is absolutely terrible in Sound Quality (old, thin, harsh, muddy, no dynamics etc) and one that has great Sound Quality (very naturalistic and/or rich, clear, dynamic etc), the obvious choice is the latter. If it we only cared about pure technical accuracy, then putting on the crappy recording would be just as suitable (or...even just a test signal).

Finally, even if you want to stick to "coloration = Not Hi-Fidelity" you aren't out of the weeds yet.

How exactly will you draw the hard line between a "Hi Fidelity" system and one that isn't Hi Fidelity?"

Let's say you have a nice accurate solid state amp driving bookshelf speakers with a digital source. Amp distortion is below audible thresholds, the speaker is about as good as it gets in terms of technical specs, linearity etc.

Vs a system using full range speakers but it has an element of "coloration" in the system - e.g. mine where I use tube amps that may introduce a subtle coloration or deviation somewhere in the signal.

Would you say "Well, mine is the Hi Fidelity System, because it is reproducing the signal with less distortion; greater Fidelity."

But, if your bookshelf speakers - say Revel M16 or whatever- only plays a linear signal down to 55 or 50 Hz - then it is missing, throwing away, the bottom portion of the source signal in many recordings. It can't even reproduce them, much less with "fidelity." Where my larger floor-standing speakers can reproduce more of the full signal. So which is "higher fidelity" to the signal? The system that is linear within it's limited range but which can not reproduce the sub bass signal? Or the system that has a very slight deviation in linearity, but which can reproduce more of the source signal, if not all of it?

Or what if you have a speaker that has more "fidelity" to the frequency response in terms of being neutral vs another system that is less frequency neutral but can reproduce the dynamics of the signal/performance with more fidelity?

Etc.

So, yes to a degree we can make these questions easier by chopping away troublesome nuances, questions and details. But there is always a certain level of arbitrariness in doing so.

Cheers.

 

[Nango]

Darko worships Brooklyn mytec Dac and amp, I wonder what these measures... I am sure he is financed heavily by Brooklyn.

Darko is a good sales person.

 

[Vini darko]

Comprehensive well laid out retorts. Your not wrong, no reproduction can perfect so choosing your preferd poison of imperfection is just fine. I was just having a tongue in cheek tease.

 

[AudioJester]

This is just another attempt to justify poorly measuring gear and convince the unsuspecting consumer they are buying something worthwhile....
People out to make a buck will say anything.

On the flip side its quite clever, these type of controversial statements draw traffic to his site....

 

[tmtomh]

Oh dear.....

Edit: Darren Myers compared the design of an audio component with digital image processing (tweaking exposure / satiation / contrast to arrive at a 'pleasant' looking image, rather than a raw and unprocessed 'true' image).

I am not sure if this is a fair comparison? I'm very familiar with digital photography but not an engineer, so I am intrigue to what ASR members may have to say about this.

https://darko.audio/2020/06/podcast-beware-the-measurements/

Is this not the perfect example of what Floyd Toole calls "the circle of confusion"?

Darren Myers' analogy of image processing to create a "pleasant" image that is better than the raw, unprocessed, "true" image is a totally valid analogy - for the production of music (mic and recording techniques, effects, mixing, mastering, etc).

It's got nothing to do with the playback chain. There you want the "raw, true" product precisely because that product already has been "processed" to sound "pleasant" in the way the artist and producer wanted it to sound.

Any system coloration that compensates in a positive way for sonic shortcomings in a recording is a coincidence that is valid for at best a small subset of recordings that might happen to share the same production characteristics. It's got nothing to do with fidelity.

So yes, Beware - Beware the Darko.

 

[Wes]

Is this not the perfect example of what Floyd Toole calls "the circle of confusion"?
.

Ha! Another optical analogy!

 

[MakeMineVinyl]

Ah, but a video of a dedicated engineer making endless measurements on an AP would have about the same interest as watching paint dry.

 

[Xulonn]

To extract yet another worm from the can, if high fidelity is "hearing" the same music as a live, the "gold standard" would be the subjective experience. The historic standards are based on listening to acoustic music, especially classical music played in an orchestra hall. The threshold of experience to quality as "high-fidelity would itself be subjective, with different threshold ranges for different people.

I don't consider 78RPM records to be high-fidelity, but IMHO, with their introduction in 1948 - when I was six years old, "microgroove" 33 RPM vinyl monaural records, introduced the concept of very high fidelity to consumers. There were R-R tape-based systems earlier, but they were not widely-available consumer products. I experienced good high fidelity audio in the early 1950s and acquired my first full-range mono hi-fi component system in 1958 when I was a junior in high school.

My adoptive father was a Presbyterian minister, and that denomination seemed to attract white collar professionals in those days, and some of them had the money and interest to purchase good "HIFi's". We moved to Chicago in 1953, a little more than a year after I was adopted at age ten. We lived in a mixed white and blue-collar neighborhood near Marquette Park on the southwest side of the city. Protestant clergymen at the time made "house calls" on parishioners which was a kind of PR thing to establish and maintain a personal rapport. My father usually took my mother along, and occasionally I was invited as well for these events which were typically right after dinner, and included coffee and dessert. Although I do not remember the specific incident, It was one of these family-oriented visits that introduced me to the concept of high fidelity audio.

I immediately became fascinated with this concept and the technology behind it. My dad mentioned my interest to a parishioner who was an engineer at the local Jensen Speaker Company, and we ended up getting a design for a pretty big bass-reflex speaker cabinet for free, and Jensen drivers and crossovers at cost. Dad built the speaker in his basement workshop. We added a used Bell Model 2300 6L6 monaural amplifier and a used Garrard "record changer" with a ceramic cartridge. It was definitely high-fidelity for the time, sounded pretty good (great for its day). The first two records I purchased were George Shearing's "Latin Escapade" and Harry Belafonte's "Calypso" album. That hifi system would probably sound pretty decent even today. It certainly sounded far better than the Stromberg Carlson console radio/record player in our dining room which was really not high-fidelity, even in those days.

So I have an audio history that pretty much parallels the introduction and development of consumer-oriented high-fidelity music reproduction. Although classical music was not my favorite at the time, I worked as an usher for the Chicago Symphony Orchestra in 1958 and 1959 when Fritz Reiner was the conductor. Those years were the beginning of the magnificent series of RCA stereophonic recordings that became classics and have been reissued by Sony, and I have a few that were reissued in digital format.

So for me, "high-fidelity" is a relative term and not a mark of audio perfection. Every component system I have owned since 1958 was a high-fidelity system. If you want to be truly pedantic, you could say that it is impossible for R-R tape with it's background hiss, or vinyl records with their pops and clicks to be truly high-fidelity, because those artifacts do not exist in live performances - the standard for comparison. OTOH, one could also say that audience noise, including the occasional cough, is the gold standard, because recording without an audience is analogous to a rehearsal rather than a performance.

So lighten up, you pedantics. High fidelity music covers a range of listening pleasure and standards that far exceed the fidelity of wax cylinders and 78RPM shellac records, and for me it is definitely not an obsession with electro-mechanical and acoustic perfection.

 

[RayDunzl]

Darko worships Brooklyn mytec Dac and amp, I wonder what these measures...

https://www.audiosciencereview.com/...-and-measurements-of-mytek-brooklyn-dac.1828/

 

[RayDunzl]

My adoptive father was a Presbyterian minister, and that denomination seemed to attract white collar professionals in those days, and some of them had the money and interest to purchase good "HIFi's".

My Grandfather was a Presbyterian minister in Wilmington NC.

https://www.google.com/maps/@34.233...PMboDTiEHFLGx1qSzdTQ!2e0!7i13312!8i6656?hl=en

He had an ancient house a few blocks down third street.

https://www.google.com/maps/@34.228...RVPqauac4Xpxv35GWtqQ!2e0!7i13312!8i6656?hl=en

And owned some shacks he rented (since, obviously highly removated) to po' folks

https://www.google.com/maps/@34.229...MvdtdR4hbXVNBwLimfoQ!2e0!7i13312!8i6656?hl=en

It was a cool church, dark, woody, granite block, stained glass, bells, and pipe organ.

https://www.facebook.com/video.php?v=2232656283683218

Dad let me "help" him build a nice Eico integrated amp kit in 1962 or so.

He was white-collar with Nabisco - eventually head of Regional Accounting Office, any step up from there would be "move to corporate in New Jersey" which he didn't want to do.

 

[MattHooper]

So for me, "high-fidelity" is a relative term and not a mark of audio perfection. Every component system I have owned since 1958 was a high-fidelity system. If you want to be truly pedantic, you could say that it is impossible for R-R tape with it's background hiss, or vinyl records with their pops and clicks to be truly high-fidelity, because those artifacts do not exist in live performances - the standard for comparison. OTOH, one could also say that audience noise, including the occasional cough, is the gold standard, because recording without an audience is analogous to a rehearsal rather than a performance.

As I've mentioned before, I think "Believable" is probably the term that most closely describes what I'm thinking when using live music as my reference. "Believeble" in the sense watching a movie is "believable." And depending on what type of content I'm consuming, I care more...or less...about this believability.

So for instance in a movie that is generally meant to be fairly naturalistic, some suspense or drama or whatever, then there are elements of "believability" that are drawn from real life - that what is happening in the story/script makes sense, or how a character is acting or reacting etc.
If I find myself consiously thinking "Oh, as IF that would ever happen" or "Geeze, nobody would do THAT!" then these are the things that yank me out of the movie - out of the illusion I had willingly engaged up to that point. Objectively speaking, images on a big flat screen diverge from real life in many ways, but if the right elements of "realism" are there - (color isn't off so skin is purple etc), the actors seem like real people, the situation flows with some level of causal realism etc, then it is "believable" even if not strictly purely "realistic/accurate" to what it would be like seeing it happen in reality. If I'm watching a war movie and there is a Tank moving past on screen firing at the enemy, I don't want it "perfectly accurate" in it's realism because I'd go deaf watching the movie. But so long as it's making a sound that is recognizably that of a "tank" (and not, say, a vacuum ceaner or kazoo), then this is the elements of "believability" that I want.

Similarly, when it comes to certain recordings of acoustic instruments or voices, I like a similar "Believability" where some of the important characteristics seem "right" as drawn from experience of real life, while not demanding some perfect recreation of reality. So, a generally recognizable sense of vocal or instrumental timbre, that kind of thing. A real drum set in my living room would drive me crazy; but I want certain aspects of "real drum" sound represented in how drums sound through my system.

 

[stevenswall]

Oh dear.....

Edit: Darren Myers compared the design of an audio component with digital image processing (tweaking exposure / satiation / contrast to arrive at a 'pleasant' looking image, rather than a raw and unprocessed 'true' image).

I am not sure if this is a fair comparison? I'm very familiar with digital photography but not an engineer, so I am intrigue to what ASR members may have to say about this.

https://darko.audio/2020/06/podcast-beware-the-measurements/

Let the photographer or videographer do the tweaking. I'll enjoy it on my quite accurate OLED TV. If I add some extra color or luminance to some things to make one movie look better subjectively (often this means different and noticeable, and gets tiring), it's probably going to look terrible for some scenes and other movies,

 

[mikeburns]

What are peoples thoughts on his explanation of global feedback? Can anyone explain why global feedback, which improves the distortion figures on measurement gear, changes the sound in some way that is different to the source and worse (he mentions sound stage specifically)? His idea on making a better circuit initially so as to reduce the amount of feedback kind of makes sense to me but I wonder if I am being led down the bs rabbit hole? Anyone want to share some illumination on this? Surely if the sound is the same as the source it shouldn't matter much how it got there?

 

[amirm]

What are peoples thoughts on his explanation of global feedback? Can anyone explain why global feedback, which improves the distortion figures on measurement gear, changes the sound in some way that is different to the source and worse (he mentions sound stage specifically)?

There is no such effect on soundstage. People put a new knob on their amp and say soundstage is improved.

Feedback has no such side-effect. Improperly designed amplifiers with feedback can oscillate, but anything you use that is not destroying itself already doesn't have that problem.

Amplification components are inherently non-linear and generate distortion. You can try to hand pick them to reduce that but the beast is the beast. Feedback allows us to sample the output, compare it to input, and take corrective action to match the two. This is no different than you paying attention to your speed in your car and adjusting the gas pedal to keep it at a certain speed.

 

[MattHooper]

Ugh...

https://darko.audio/2020/08/kih-78-rack-em-up/

Putting new high-end audio racks under your gear produces "mega" results and "Unless tried, you simply won’t believe what the elimination of micro resonances will do for the sound of your existing hardware."

With absolutely nothing but this reviewer's say-so, zero evidence or even a specific technical theory to back it up. May as well be peddling astrology.

It's frustrating when the hi-fi industry ends up often mirroring a stroll through a Psychics Fair.

 

[Sgt. Ear Ache]

LOL, what a shill....

 

[misteracng]

Ugh...

https://darko.audio/2020/08/kih-78-rack-em-up/

Putting new high-end audio racks under your gear produces "mega" results and "Unless tried, you simply won’t believe what the elimination of micro resonances will do for the sound of your existing hardware."

With absolutely nothing but this reviewer's say-so, zero evidence or even a specific technical theory to back it up. May as well be peddling astrology.

It's frustrating when the hi-fi industry ends up often mirroring a stroll through a Psychics Fair.

This makes absolutely zero sense. I can't even believe someone would actually put this crap down on paper