• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Master Thread: “Objectivism versus Subjectivism” debate and is there a middle ground?

ahofer

Master Contributor
Forum Donor
Joined
Jun 3, 2019
Messages
5,045
Likes
9,148
Location
New York City
As I’ve said before, the question is whether the subjective differences created by the equipment are stable over time. My supposition is they are not, and I submit the “upgrade treadmill” as evidence. But I don’t know that it is an open-and-shut case.

Even the distortion isn’t stationary - varies based on signal, obviously, so might be more euphonic with some things than others.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,762
Likes
37,616
As I’ve said before, the question is whether the subjective differences created by the equipment are stable over time. My supposition is they are not, and I submit the “upgrade treadmill” as evidence. But I don’t know that it is an open-and-shut case.

Even the distortion isn’t stationary - varies based on signal, obviously, so might be more euphonic with some things than others.
Well there is that upward curve during the obligatory 200 hour break in (somehow it is nearly always 200 hrs). Then a period of a lessening of the improvement. Followed by a leveling off which can last a long time. Then one day, it just does not do the trick like it used to or some new gear or tweak lifts a veil, and then........on the treadmill you go for another fix. Subjective differences are much like a drug addiction. Maybe a runner's high if the runner is on a treadmill.
 

Ron Texas

Master Contributor
Forum Donor
Joined
Jun 10, 2018
Messages
6,237
Likes
9,369
The only thing in the middle of the road is a yellow line and dead armadillos.
 

spartaman64

Active Member
Joined
Apr 2, 2021
Messages
137
Likes
140
im an objective subjectivist. I dont believe that measurements are everything but I do believe in blind listening tests.
 

xnor

Active Member
Joined
Jan 12, 2022
Messages
193
Likes
207
What I think has created this whole seemingly false-dichotomy in the audio world is that there's no real, clear method of objectively conveying a subjective experience when listening to a recording through a piece of gear or multiple pieces of gear.
There are scientific methods for evaluating the sound quality of audio components.
Why do you think the audiophile industry largely supports "subjectivist" platforms/formats/outlets/reviews instead? Because there are enough gullible and uninformed consumers willing to give them money. Human hearing is much more limited than the mind. Pay for some high end reviews, get the hype rolling in some forums and you can sell even quite bad hardware for a fortune.

Ironically, audio fidelity is almost irrelevant to those audiophile customers, because they hear what they were told to hear.

And this is that grey, unexplored, ambiguous area where a lot of marketing and snake oil happens. There is, however, an objective way to reference a recording, and see if when you pump that signal through a series of gear, the result is anything like what you started with.
People have shown that some super expensive cables make zero audible difference in null-difference tests years ago. (You can do the same with any piece of digital/electronic gear.)
There's virtually zero grey or ambiguous area.
Do you think that would stop the manufacturer from selling these cables or gullible consumers buying them? It's obvious why the audio(phile) industry doesn't employ these methods and tests or work towards a universally accepted grading scheme.

Again, it's about everything other than actual high fidelity. $$$ first and foremost.

I think that focusing on objectivity gives us a good starting point though, for evaluating something as snake oil or not. For example, if there's a particular piece of gear that doesn't measure well, but a lot of people seem to enjoy listening to it, that gives us an important data point.
Yeah, primarily that human hearing sucks.

Having said all that, I like gear that measures well, because I think if I want to add distortion or effects to what I am listening to, that should happen at the time the recording is made or mastered. Because, mastering and recording techniques vary, but generally, audio reproduction equipment will always impart its own "signature" without prejudice, so it may indeed impart a quality that makes one recording better to my ears, but then it will impart that same quality to another recording that I know I like, but make me not enjoy it. Does that make me an objectivist?
The whole idea to reproduce the same sound as was produced during the recording is a complete joke in most cases. Audio quality varies significantly between albums/recordings.
Unless the artist chose the mics, positions, recording spaces, other recording gear and supervises the mixing and mastering process, you got a plethora of variables that all vary from one recording to the next.
Sure, the artist says "good enough" at some point, but it's a bit like showing a painter a processed image of his painting and selling that instead of the painting.

Just a different mic model or slightly different positioning would cause larger differences than the differences between a good and a perfect DAC, amp.

I acknowledge that there are measurable distortions or inaccuracies that can increase the listener's subjective experience, but again, it's not what I have set out to do when piecing together a listening system. I try to go by what fits in my budget and measures well, but ultimately, I listen to whatever I can before I buy it. So, in my own system, I have gear that didn't measure well here, objectively, but sounded pretty good when I listened to it using recordings I am familiar with and it fit my budget. But, I always use the reviews and information I gather here as a starting point.
It probably doesn't translate/matter because of masking and a host of uncontrolled variables in your listening circumstances.
 

krabapple

Major Contributor
Forum Donor
Joined
Apr 15, 2016
Messages
3,195
Likes
3,763
I'm just a dumb guy, but I think the argument boils down to what your goal is. Do you want gear that accurately presents stored information (in this case an audible waveform) or do you want something that conjures up a feeling when you listen to it?

They aren't mutually exclusive.

Since your premise is so patently dubious, and your post was of Hooperian length, I'm stopping there.
 
Last edited:

AbidingDude

Member
Joined
Dec 30, 2021
Messages
11
Likes
12
They aren't mutually exclusive.

Since your premise is so patently dubious, I'm stopping there.
Point to where I explicitly stated they were. In fact, I spent the next several paragraphs basically saying that they're not.
 

krabapple

Major Contributor
Forum Donor
Joined
Apr 15, 2016
Messages
3,195
Likes
3,763
*Bizarre rant on* Something to think about, on the subject of objectivism vs subjectivism. Almost any decent speaker has been "voiced" by the designer. This usually is done by getting things pretty close to flat by measurement, +/-1dB if possible, then tweaking to taste. I know in my case those little 0.25 and 0.5dB tweaks can make the difference between simply being listenable, and wanting to listen more. *Bizarre rant off*

0.25 dB? really?

That *is* bizarre.
 

krabapple

Major Contributor
Forum Donor
Joined
Apr 15, 2016
Messages
3,195
Likes
3,763
Point to where I explicitly stated they were. In fact, I spent the next several paragraphs basically saying that they're not.

You framed the essay in terms of a dichotomy that is both a fallacy and IME unlikely to exist as actual competing 'goals' in audio hobbyists' minds. I.e., a straw man.

So next time, start better? Or get to the point faster. Or both.
 

fpitas

Master Contributor
Forum Donor
Joined
Jul 7, 2022
Messages
9,885
Likes
14,212
Location
Northern Virginia, USA

krabapple

Major Contributor
Forum Donor
Joined
Apr 15, 2016
Messages
3,195
Likes
3,763
According to whom?

Differences in that range tend to be discernable only in the part of the audio spectrum where the ear is most sensitive. According to science.
 

fpitas

Master Contributor
Forum Donor
Joined
Jul 7, 2022
Messages
9,885
Likes
14,212
Location
Northern Virginia, USA
According to whom?

Differences in that range tend to be discernable only in the part of the audio spectrum where the ear is most sensitive.
Over a broad band? Not so.
 

dlaloum

Major Contributor
Joined
Oct 4, 2021
Messages
3,153
Likes
2,416
*Bizarre rant on* Something to think about, on the subject of objectivism vs subjectivism. Almost any decent speaker has been "voiced" by the designer. This usually is done by getting things pretty close to flat by measurement, +/-1dB if possible, then tweaking to taste. I know in my case those little 0.25 and 0.5dB tweaks can make the difference between simply being listenable, and wanting to listen more. *Bizarre rant off*
My own experience is that even 0.3db variations can (is likely to) cause interesting differences in voicing - they are not obvious as loudness differences, but can highlight specific instruments, give certain instruments a "glow", or improve soundstaging - it is very subtle

But once you get down to 0.2db - differences are not consistently identifiable ... (ie: double blind, chances are you won't be able to statistically differentiate)
 
Last edited:

dlaloum

Major Contributor
Joined
Oct 4, 2021
Messages
3,153
Likes
2,416
0.25 dB? really?

That *is* bizarre.
My own testing showed 0.5db to be differentiable, and 0.2db not to be differentiable (by my ears... - but others also confirmed - not a wide test) - didn't do testing between 0.5db and 0.2db - so not sure where the cut off might be, nor by how much it would vary depending on the individual listener (and perhaps the frequency at which the variation happens).

But I don't find it strange... based on my own experimentation.
 

krabapple

Major Contributor
Forum Donor
Joined
Apr 15, 2016
Messages
3,195
Likes
3,763
Over a broad band? Not so.

0.2dB is a very tiny difference. Just noticeable difference in amplitude is dependent on frequency content of the signal and the starting level of the signal and how close the two levels are presented in time. IOW highly dependent on circumstance. Simply boosting or reducing a musical track on a stereo system by 0.2 dB within normal listening levels is very unlikely to produce a statistically solid difference in a DBT. Even 0.3 dB is a 'maybe'.

If you really want to test your ability to detect a ~0.2dB difference and give yourself the best chance of succeeding, you would use a test signal optimized to the ear's most sensitive range (i.e, the midrange). Not a "broad band"/music signal where masking effects will be strongly in play.

You will, *of course*, have to a use blind listening protocol (DBT) to produce a credible result. Do I even need to point that out on a forum devoted to Audio Science?

If you want a standard reference on the topic, chapter 7 of this book is devoted to it:

Psychoacoustics by Fastl & Zwicker
 

fpitas

Master Contributor
Forum Donor
Joined
Jul 7, 2022
Messages
9,885
Likes
14,212
Location
Northern Virginia, USA
0.2dB is a very tiny difference. Just noticeable difference in amplitude is dependent on frequency content of the signal and the starting level of the signal and how close the two levels are presented in time. IOW highly dependent on circumstance. Simply boosting or reducing a musical track on a stereo system by 0.2 dB within normal listening levels is very unlikely to produce a statistically solid difference in a DBT. Even 0.3 dB is a 'maybe'.

If you really want to test your ability to detect a ~0.2dB difference and give yourself the best chance of succeeding, you would use a test signal optimized to the ear's most sensitive range (i.e, the midrange). Not a "broad band"/music signal where masking effects will be strongly in play.

You will, *of course*, have to a use blind listening protocol (DBT) to produce a credible result. Do I even need to point that out on a forum devoted to Audio Science?

If you want a standard reference on the topic, chapter 7 of this book is devoted to it:

Psychoacoustics by Fastl & Zwicker
You are missing the point. For a single pure tone, it takes about 1dB difference.
 

dlaloum

Major Contributor
Joined
Oct 4, 2021
Messages
3,153
Likes
2,416
0.2dB is a very tiny difference. Just noticeable difference in amplitude is dependent on frequency content of the signal and the starting level of the signal and how close the two levels are presented in time. IOW highly dependent on circumstance. Simply boosting or reducing a musical track on a stereo system by 0.2 dB within normal listening levels is very unlikely to produce a statistically solid difference in a DBT. Even 0.3 dB is a 'maybe'.

If you really want to test your ability to detect a ~0.2dB difference and give yourself the best chance of succeeding, you would use a test signal optimized to the ear's most sensitive range (i.e, the midrange). Not a "broad band"/music signal where masking effects will be strongly in play.

You will, *of course*, have to a use blind listening protocol (DBT) to produce a credible result. Do I even need to point that out on a forum devoted to Audio Science?

If you want a standard reference on the topic, chapter 7 of this book is devoted to it:

Psychoacoustics by Fastl & Zwicker
First you are talking about a discernible difference in amplitude.

Small differences are not discernible as differences in amplitude

(and I wasn't talking about a broad band difference, but a difference in frequency response in a narrow band of frequencies, by 0.5db and less)

They are discernible in differences in "detail", or soundstaging, they are discernible as very subtle differences... and in by BT (not DBT) 0.2db was not discernible - but 0.5db was definitely discernible.... just to reiterate - the 0.5db was in a narrow band - from memory up around 2kHz to 8kHz somewhere in that range - and it manifested subtly and absolutely NOT as an amplitude difference.
 

Axo1989

Major Contributor
Joined
Jan 9, 2022
Messages
2,901
Likes
2,954
Location
Sydney
The dogma is a bit obtuse. On audiocheck I can discern 1 dB up/down (tones) consistently, maybe 0.5 dB on a good day, but we are scolded if we don't level-match to voltage equivalent of 0.1 dB because there's a difference. Contradiction or margin for error?

Tangentially, it wouldn't surprise me if differences visible in the spectrogram (can't remember which German magazine regularly does that) correlate to subjective impressions of (micro?) tonality/timbre and stereo imaging (for speakers). Very small changes in FR (up the chain) I'm not sure about (with actual music playback) but maybe. But that's my idle speculation.
 

dlaloum

Major Contributor
Joined
Oct 4, 2021
Messages
3,153
Likes
2,416
The dogma is a bit obtuse. On audiocheck I can discern 1 dB up/down (tones) consistently, maybe 0.5 dB on a good day, but we are scolded if we don't level-match to voltage equivalent of 0.1 dB because there's a difference. Contradiction or margin for error?

Tangentially, it wouldn't surprise me if differences visible in the spectrogram (can't remember which German magazine regularly does that) correlate to subjective impressions of (micro?) tonality/timbre and stereo imaging (for speakers). Very small changes in FR (up the chain) I'm not sure about (with actual music playback) but maybe. But that's my idle speculation.
I definitely noted difference with 0.5db differences in Frequency Response levels

I definitely could not tell the difference with 0.2db differences in Frequency Response levels.

Your experience may well differ, but if you really want to do a DBT you need to get the differences well below 0.5db

(Edit typo from 0.5 to 0.2)
 
Last edited:

Axo1989

Major Contributor
Joined
Jan 9, 2022
Messages
2,901
Likes
2,954
Location
Sydney
I definitely noted difference with 0.5db differences in Frequency Response levels

I definitely could not tell the difference with 0.5db differences in Frequency Response levels.

Not following, typo?

Your experience may well differ, but if you really want to do a DBT you need to get the differences well below 0.5db

Yes, that's catechism, but if I can't DBT that as a level difference, how does it work?
 
Top Bottom