What objective measurements qualify as "High Fidelity"?

[terryforsythe]

Whilst the above for electronics may be the SOTA , they are ludicrously excessive as a minimum standard for Hifi. THD of -60dB is perfectly adequate for transparency, and a S/N ratio of 80dB more than acceptable. Stereo separation of 20 -25dB is good enough for stereo, and a 20-20kHz response of 1dB sufficient, even 3dB wouldn't be audible at the frequency extremes.

It might be useful to do some blind listening tests to see just how tolerant we are.

S

In the video posted in this thread JJ Johnston stated the following:

I before had never heard of Mr. Johnston, but considering his video was posted in this thread, it appears that at least somebody considers him to be knowledgeable on this subject.

With differing opinions as to what qualifies as "high fidelity", I remain of the opinion that this is subjective. What I may consider to be high fidelity may be different than what someone else considers to be high fidelity. With no standards to which to compare objective measurements, neither of us is wrong.

 

[Jim Taylor]

In previous posts, e.g. this, I said that I think that the idea of fidelity as a vague expression of aspirations was probably useful in the past but that it is now obsolete.

Yes, I read that post. I disagreed with it then, and I disagree with it now.

You are conflating two totally different things. One is the degree to which the audio industry can deliver a certain product to us, and the other is the degree to which we like that product.
"Fidelity" is not vague. The iteration may be non-standardized, but the concept is the same for everyone the world over, namely, "How accurately does this product over here resemble that source over there?"
That's why we're having this discussion right now ... to try and illuminate which standards may apply, if any, and what ramifications for the consumers that may have.

Let's look at your view logically. In your quoted post, you said regarding the priority of equipment, ".. It's a question of what gives you satisfaction ...". I can agree with that; as @j_j said in his presentation, "there is no arguing with preference." But you have slightly missed the point: preference or satisfaction compared to what? You have to have some thing, some object, some process, some effect, some standard first, in order to have a preference about it. You can't have a preference about nothing.

The industry knows this. They make money by providing is with something and taking our money in exchange. For people who concentrate on the principles that you've mentioned (personal satisfaction) the audio equipment industry offers all sorts of EQ components and software, tube gear, funny-looking speakers, tweaks up the yin-yang, different cartridges for LP playback, et al.
But all those modifications have to have a central target to modify, and that is the recording. So people who believe as you do will play around with their gear to come up with the effect that they desire, and that is to their satisfaction.

What about the rest of us? Let's say I want to listen to a recording of Debussy's La Mer by the Indianapolis Symphony Orchestra. I was there for the performance, and I thought it was magnificent. I want to listen to it again, and revel in the performance as I did the first time.
If there is no effort on the part of the recording industry to aspire to some sort of fidelity, what will I hear? Will I just hear some junk that someone else recorded to their "satisfaction"?
No, I want a recognizable illusion of what I heard that night. I know it won't be perfect ... I've known that for well over 40 years. But if I can hear the second violin's chair squeak, the percussion section cough and the unique tonal balance of that orchestra giving new life to Debussy's little child, I can rest easy that the recording has FIDELITY.

If someone else wants to modify that recording to their "satisfaction", I see absolutely nothing wrong with that. Go ahead ... play with it to your heart's content. That way, you can attain your goal.
But it the recording isn't created with some sort of "fidelity" in mind, then I cannot achieve my goal ... not ever.

And I would not be happy.

Jim

 

[ahofer]

It feels as if the important part of this concept has been replaced with "transparency" for electronics, meaning the old straight-wire with gain, or the most accurate recreation of waveform from digital.

High Fidelity/HiFi is too loose a term, therefore in the ears and eyes of the beholder, although I'd say transparency in the electronics is a prerequisite.

 

[terryforsythe]

Yes. And we have ways to express and to measure that ability. We don't need to involve fidelity to do it. Trying to relate fidelity to the objective measures that have proven their value confuses matters. That's why I think the concept isn't useful.

I actually agree.

However, what is considered to be "high fidelity" cannot be objective absent some standards to which objective measurements can be compared. Without standards we are left with subjective opinions based on interpreting measurement results. I'm fine with that, but I am not fine stating that those subjective opinions are objective. E.g., one person states 80 dB SNR is high fidelity while another person states that 90 dB SNR "could be just fine, and it could be really, really bad".

 

[terryforsythe]

High Fidelity/HiFi is to loose a term, therefore in the ears and eyes of the beholder

Dare I say "subjective"?

I'd say transparency in the electronics is a prerequisite.

Is transparency objective or subjective?

If objective, what objective measurements, and standard values to which to compare those measurements, do you propose as an indicator of transparency?

 

[ahofer]

Dare I say "subjective"?


Is transparency objective or subjective?

If objective, what objective measurements, and standard values to which to compare those measurements, do you propose as an indicator of transparency?

The ultimate measure would be to compare and null signal measurements through each component (which has been done many times), but the SINAD composite (say >95) , frequency response measurements (0.1db), reasonably low output impedance, and sufficient power (as necessary) are reasonable proxies. I think this is well defined in Amir's reviews and the audibility thresholds thread. I'm not sure why we need to repeat it here.

 

[terryforsythe]

Can't imagine high fidelity judgement being solely subjective. Probably best approached by objective measurement then tweaking to preference.

There are some equipment that many will agree are "high fidelity". On the other hand, there are some equipment that some consider to be high fidelity and others consider to not be high fidelity, especially with speakers.

I think the best approach is for people to review the objective measurements, and perhaps review subjective opinions. Based on their review, they should make their own subjective opinion as to whether the equipment in question is "high fidelity" to them.

 

[terryforsythe]

I'm not sure why we need to repeat it here.

The reason for this thread is to discuss whether what is considered to be high fidelity is subjective, as I have stated, or objective, as others have stated.

We have all sorts of objective data, but the issue comes down to interpretation of that data, which I contend is subjective.

As an example, when reviewing the objective data for the Sonus Faber Lumina II speaker, I do not consider it to be "high fidelity". But, many other people do consider that speaker to be "high fidelity". Who is correct? Absent standards to which to compare the objective data, neither contention is wrong.

As another example, the Schitt Vidar 2 amplifier has a SNR of 96 dB at 5 W. Some may consider this "high fidelity", but some may not (e.g., see the video posted in this thread). Indeed, Amir classified that amplifier as being "Very Good/Good" but not "Excellent". Is any equipment that is "Very Good/Good" and above considered to be high fidelity, or only equipment categorized as "Excellent"? The answer to that question is subjective, and likely will not be the same for all people who may answer it.

 

[kemmler3D]

As an example, when reviewing the objective data for the Sonus Faber Lumina II speaker, I do not consider it to be "high fidelity". But, many other people do consider that speaker to be "high fidelity". Who is correct?

For this you need an objective definition of "fidelity", and as you've attempted, what we should call "high". From Merriam-Webster:

Definition #2 gets us closer. It's synonymous with accuracy, which we know means lack of distortion in electronics, but once we begin to consider off-axis and in-room performance of speakers, this attempt to nail down "fidelity" or "accuracy" or any similar notion falls apart.

There's virtually no way to know what the "original" in-room sound was, and even if we know what they heard in the studio, it's hard to argue convincingly that that should be considered "the original artifact" and all others must be considered less correct. And even among the most reliable authorities in the audio world, there is no final consensus on a single "most correct" dispersion or in-room response, just good sense on what's generally preferred.

So I think fidelity is a great concept for electronics, and an okay one for speakers (on-axis response ought to be flat, we can at least get that far) but beyond that I think we lack references to tie a specific objective measure to.

I will add that if you want to, you can say that fidelity SHOULD mean fidelity to Harman-type dispersion characteristics, as the in-room response that has the most public data behind it. This is an appeal to break the circle of confusion and come to an agreement on in-room response. But you can find a lot of people who disagree with this idea, too (maybe just in the details) so I don't know that you can have an objective snout-to-tail measure of fidelity, you always need to inject opinion at some point.

 

[MattHooper]

I generally take "fidelity" to mean the same thing as "minimal distortion", but I appreciate that you put a deeper or more philosophical concept behind the word. I don't anymore because it involves too many opinions. More specifically, aiming for "fidelity" to a certain experience is a faulty premise. "You can't step into the same river twice," etc.

I’ve argued previously why I prefer to leave the term “high Fidelity“ out of the definition of “audiophile.“ And that I see a audiophile as:

Someone who is enthusiastic about gear.

Or perhaps:

Someone who is enthusiastic about audio gear and high-quality sound.

(because not everyone in this hobby is chasing technical High Fidelity per se)

 

[kemmler3D]

I’ve argued previously why I prefer to leave the term “high Fidelity“ out of the definition of “audiophile.“ And that I see a audiophile as:

Someone who is enthusiastic about gear.

Or perhaps:

Someone who is enthusiastic about audio gear and high-quality sound.

(because not everyone in this hobby is chasing technical High Fidelity per se)

I agree. An audiophile's interest in fidelity is a secondary characteristic. I think if you want to gatekeep "audiophile" as some kind of earned status, you might want to pin it on fidelity, but for everyday usage of the word it wouldn't make sense.

 

[terryforsythe]

Definition #2 gets us closer. It's synonymous with accuracy, which we know means lack of distortion in electronics

Some audio electronics have extremely low distortion, but none are completely distortion free. So the question becomes "what level level of distortion qualifies as 'high fidelity'?" As far I can tell, there is no consensus on the answer to that question, which makes the answer subjective absent some standard to be applied.

 

[MattHooper]

I agree. An audiophile's interest in fidelity is a secondary characteristic. I think if you want to gatekeep "audiophile" as some kind of earned status, you might want to pin it on fidelity, but for everyday usage of the word it wouldn't make sense.

Yeah, and then you'd get in to "what then is the objective criteria for a high fidelity system, and whose system therefore merits the claim to high fidelity?" Given all of us are working with some level of compromise that would be a lot of picayune gatekeeping work. And it would be odd as hell to not see the people in audio forums who have spent years putting together tubes and horn systems or whatever as not being an audiophile.

 

[kemmler3D]

Some audio electronics have extremely low distortion, but none are completely distortion free. So the question becomes "what level level of distortion qualifies as 'high fidelity'?" As far I can tell, there is no consensus on the answer to that question, which makes the answer subjective absent some standard to be applied.

Well, as I said before "high" is more a marketing term than anything in this situation.

So a better question is: what do you want to accomplish by pursuing more fidelity?

If you know the objective, then you can simply say a system does or doesn't have enough fidelity for that. And you would probably decide that's "high fidelity" also. However, I sincerely doubt you can get everyone to agree on it, because not everyone has the same objective for their systems.

 

[MattHooper]

As an example, when reviewing the objective data for the Sonus Faber Lumina II speaker, I do not consider it to be "high fidelity". But, many other people do consider that speaker to be "high fidelity". Who is correct? Absent standards to which to compare the objective data, neither contention is wrong.

I would say that disagreement does not automatically entail "subjective." Or at least it doesn't rule out objectivity.

To the question "who is correct" above, I'd ask why you consider the SF not to be high fidelity, and I'd ask the other person what standard she has for "high fidelity" that the SF meets. One case may not be as coherent as the other.

I think for sources and amplification, we can consider "high fidelity" any equipment that reproduces the recorded signal without distortion rising to the audible threshold. That's objective.

Do we throw this out the window with speakers? No reason to. We should be careful about throwing the baby out with the bathwater, or making The Perfect the enemy of The Good. Loudspeakers may not achieve the level of High Fidelity (as described above) as electronics, but they can get closer or further from that goal, and we can know what types of distortion are audible or less audible, and grade speakers on scales being 'more high fidelity' and 'less high fidelity.' For instance a high end Genelec is certainly going to be measurably more High Fidelity than, say, a Zu loudspeaker. Likewise we could say the Genelec is higher fidelity than the SF speaker. Having an objective goal doesn't necessarily require getting absolutely everyone on board - heck, there are objective facts about the earth but Flat Earthers just aren't interested in playing that game. Doesn't make the facts any less objective.

I think the subjectivity is more about which goals one has in the hobby. But one can choose a goal, like High Fidelity To The Signal, that allows for objectivity.

 

[Nopiano]

Is transparency objective or subjective?

I think it is probably both. Here on the forum, it seem to be used quite often to indicate achievement of low noise and distortion in electronics. That’s the objective side, and your efforts suggest minimum criteria for that (though I think you are setting the bar too high).

The subjective side is when I try a new device and get to hear things I’d not previously heard in a favourite recording. Or sense the hall reverb better matches my recollection of that venue. Those are what I’d describe as transparent. But of course I cannot be sure without some measurements in case the ‘detail’ I suddenly think I’ve heard is simply an exaggeration of a particular frequency range, for instance.

My suggestion many posts ago that some vintage kit can still delight with its sound is why I think that we can overlook (aurally) some surprising aberrations yet get persuasive results. And I suspect as individuals we tolerate different flaws, hence the disagreement we sometimes see between experienced listeners.

 

[terryforsythe]

For instance a high end Genelec is certainly going to be measurably more High Fidelity than, say, a Zu loudspeaker.

Good examples of two ends of the spectrum.

 

[Salt]

At the end of the day we may resume
1. There's fidelity, but no objective number to it
2. High fidelity is not definable, see 1.

 

[terryforsythe]

At the end of the day we may resume
1. There's fidelity, but no objective number to it
2. High fidelity is not definable, see 1.

Well said!

 

[MattHooper]

At the end of the day we may resume
1. There's fidelity, but no objective number to it

Only if you don't bother defining a term.

2. High fidelity is not definable, see 1.

High Fidelity in audio: Reproducing an audio signal without audible levels of distortion.

How is that not definable?