Head-Fi.org › Forums › Equipment Forums › Sound Science › Terminology and understanding subjective reviews
New Posts  All Forums:Forum Nav:

Terminology and understanding subjective reviews

post #1 of 20
Thread Starter 

Subjective reviews are supposed to have an advantage over objective reviews in that they relate better to how we perceive something (versus numerical data).

How true is this in practice? Subjective reviews describe sound using a common set of terminology (bright, dull, detailed, transparent, etc). How often do these make sense, and how do we know the reader understands them in the same way as the writer?

 

I find that statements that describe frequency response make the most 'sense' (e.g. v-shaped, flat, bright, etc).

On the other hand, some are completely incomprehensible. For example, "creamy luscious mids, dynamically vibrant" or from the same author, "smooth, enveloping warm gooey liquidy flowing mids". This is completely incomprehensible - while music can sound amazing, it sounds like music, it doesn't feel like a viscous fluid. I can't relate to this description, can you?

 

Other terms like 'texture' are not as easy to interpret but do directly relate to how music is perceived.

 

Terms like 'PRaT' are often thrown around and on one hand it is obvious what it means yet I have never heard this issue with any equipment and can't see how an earphone would have timing off by more than a few milliseconds, a difference that isn't really consciously perceptible.

 

Leaving aside the validity of subjective reviewing (and subjective versus objective), is the terminology we use adequate?


Edited by higbvuyb - 6/11/13 at 5:32pm
post #2 of 20

Here's a list of commonly used terms that you'll often find in subjective reviews. http://www.head-fi.org/t/504503/is-it-really-worth-upgrading-sennheiser-hd-580-600-650-cables/210#post_9509833

post #3 of 20

http://www.head-fi.org/a/describing-sound-a-glossary

 

This is guide I generally refer to when people mention things like "lush, warm, bright, luxurious yet some what smooth treble with good PRaT" 

post #4 of 20
Quote:
Originally Posted by higbvuyb View Post
...
Leaving aside the validity of subjective reviewing (and subjective versus objective), is the terminology we use adequate?

No, it is not. Because the terms are subjective they necessarily require personal interpretation, which will vary from person to person. Even with the guide on subjective terms that Parallel linked to, there is no universal way to quantify such things (otherwise it's objective reviewing) and the descriptions of the various terms are open to the variability of individual interpretation.

post #5 of 20

Creamy luscious mids????  Yeah I know what they mean.  Tubes mostly do this, mostly triodes of one sort or another.  Once you hear it, and hear it described it makes perfect sense. 

 

A very interesting experience.  Huge electrostatic panels driven by huge OTL's.   Jazz club recordings were almost religious.  A smoky, palpable 3D experience.  Some other recordings well maybe okay, maybe not.  Highly colored, sometimes beautifully so, yet nothing like straight wire with gain.  Strange to be using some of the most low distortion, flat response speakers to reproduce the most colored signals you will get to hear.  Subjectively pretty interesting.  In terms of high fidelity rather laughable. 

post #6 of 20
Thread Starter 
Quote:
Originally Posted by ab initio View Post

No, it is not. Because the terms are subjective they necessarily require personal interpretation, which will vary from person to person. Even with the guide on subjective terms that Parallel linked to, there is no universal way to quantify such things (otherwise it's objective reviewing) and the descriptions of the various terms are open to the variability of individual interpretation.

Being open to interpretation is not necessarily a huge flaw in all subjective observations. For example, a trained listener could conclude that a source sounds like it has a high-Q peak in the 5000Hz region in the frequency response. This is of course one extreme and you would never see such a comment in a subjective review. However we can see that some terms are less open to interpretation than others. 'bass-heavy' is an example of a fairly non-controversial term, wheras on the other hand we have "gooey mids" which to me is about as useful as saying these headphones sound 'purple'.

 

On the third hand we have terms that are easily interpreted (e.g. PRaT) yet their meaning has no basis in reality.

 

The question is, which terms lean closer to validity and which ones indicate that the writer has an excessively creative approach to reality? How much of a subjective review do people really understand?

 

 

 


Quote:
Originally Posted by esldude View Post

Creamy luscious mids????  Yeah I know what they mean.  Tubes mostly do this, mostly triodes of one sort or another.  Once you hear it, and hear it described it makes perfect sense. 

 

A very interesting experience.  Huge electrostatic panels driven by huge OTL's.   Jazz club recordings were almost religious.  A smoky, palpable 3D experience.  Some other recordings well maybe okay, maybe not.  Highly colored, sometimes beautifully so, yet nothing like straight wire with gain.  Strange to be using some of the most low distortion, flat response speakers to reproduce the most colored signals you will get to hear.  Subjectively pretty interesting.  In terms of high fidelity rather laughable. 

 

How can you describe sound with terms that are completely unrelated to sound? Such an analogy is so open to interpretation that you might as well say the mids sound brown.

post #7 of 20
Quote:
Originally Posted by higbvuyb View Post

Subjective reviews are supposed to have an advantage over objective reviews in that they relate better to how we perceive something (versus numerical data).

How true is this in practice? Subjective reviews describe sound using a common set of terminology (bright, dull, detailed, transparent, etc). How often do these make sense, and how do we know the reader understands them in the same way as the writer?

They are attempts of the listener to describe what he/she hears. How this relates to your hearing is hard to predict with headphones (due to different ear canal length, shape, pinna size, shape etc.) and maybe even harder with speakers (due to room acoustics).

 

Some subjective terms do make sense, but are very fuzzy in their meaning because they are used that way. For example, bright can mean anything from flat but boosted highs to a completely erratic treble response with big peaks. Some think a boost at 6 kHz causes brightness, others would say 8 or 10 kHz, but the boost could also be across that entire range, or just a single but really nasty peak at 9 kHz.

 

I'm not sure if readers understand these terms the same way as the writer. Even if they use the same definition there will still be huge interpretative differences.


Edited by xnor - 6/12/13 at 6:07am
post #8 of 20

Supposedly some studio types can use terms like these (though maybe not specifically PRaT etc.) and effectively communicate some rather specific ideas, probably better than you'd imagine. Repetition of usage and experience are key. It's inevitable that we loan words outside of audio, and I don't think it's much of a problem so long as in this context the meaning is relatively clear and consistent between most people. And it is, to some extent, but definitely not completely. Sometimes the relationships to other senses may be helpful; sometimes they may not be.

 

When assessing headphones—and note the differences in response in ear canal between subjects—the Olive-Welti study had some data on listener descriptors for headphone sound quality. There's some significant consensus but quite a few differences of opinion.

click to see image (Click to show)

 

(click for larger version)

 

from these slides (here).

post #9 of 20
Quote:
Originally Posted by higbvuyb View Post

 

 


 

How can you describe sound with terms that are completely unrelated to sound? Such an analogy is so open to interpretation that you might as well say the mids sound brown.

As mikeaj says, repetitive usage and experience among people can give a commonality of meaning.  Is it super precise? No.  Is it possible to give wrong impressions?  Yes.

 

But for instance without much experience most people having heard smooth and rough sound find the words appropriately descriptive.  Some triode tube amps are very smooth in the mids so not too big a leap to understand creamy smooth as a descriptor.  And odd you mention brown.  Actually a word I have used in regard to very old triode amps at least among some friends.  Don't know I would expect others to know innately.  It refers to a very smooth sounding amp with fairly rolled off highs.  Not rolled off rapidly, but rolled off.  Old tubes that are very weak along with a design having limited high frequency response seem to get this sound in time.

 

Certainly audio writers have gone above and beyond in the descriptive language.  But there is some validity to some of it.

 

Analogies are used in thinking all the time.  So audio is no different, you can describe sound with terms not related to it by analogy. 

post #10 of 20

Sure, there is some validity to it.

 

The question is why sugarcoat or paraphrase it with flowery words?

 

Autophile: "My car gives me a highly detailed impression of the texture of the street, especially in right-hand bends."

Mechanic: confused.gif, checks, responds: "Oh, your front left shock absorber is broken."

 

 

I admit, that's a bit exaggerated.

post #11 of 20
Quote:
Originally Posted by higbvuyb View Post

Being open to interpretation is not necessarily a huge flaw in all subjective observations. For example, a trained listener could conclude that a source sounds like it has a high-Q peak in the 5000Hz region in the frequency response. This is of course one extreme and you would never see such a comment in a subjective review. However we can see that some terms are less open to interpretation than others. 'bass-heavy' is an example of a fairly non-controversial term, wheras on the other hand we have "gooey mids" which to me is about as useful as saying these headphones sound 'purple'.

 

On the third hand we have terms that are easily interpreted (e.g. PRaT) yet their meaning has no basis in reality.

 

The question is, which terms lean closer to validity and which ones indicate that the writer has an excessively creative approach to reality? How much of a subjective review do people really understand?

 

 

 


 

How can you describe sound with terms that are completely unrelated to sound? Such an analogy is so open to interpretation that you might as well say the mids sound brown.

 

The question you asked was "Leaving aside the validity of subjective reviewing (and subjective versus objective), is the terminology we use adequate?" And the answer, I suppose, is how you define adequate. I am arguing that subjective descriptions are imprecise and change meaning from one person to the next. I agree with you that some descriptions may be more consistent on average across the population, but that doesn't make them precise. Your example about a 5kHz peak is a quantitative measurement by a system whose measurement uncertainty is unknown; hence, I would call it an unreliable objective measurement.

 

Your comment on "bass-heavy" is exactly the reason why I gave my first response---it all depends on your reference point, and not everybody has the same reference point. I listen to Sennheiser HD 280 pros, and my girlfriends AKG K240s sound "bass heavy" to me. On the other hand, somebody with a pair of Sony MDR-XB500s will probably disagree with me and say the AKG K240s have anemic low end. The difference isn't that the bass response of the AKGs drastically increased on my head, but that I have a much different point of reference that the other guy.

 

Here are the measurements of these headphones: http://graphs.headphone.com/graphCompare.php?graphType=0&graphID[]=533&graphID[]=2611&graphID[]=1153&scale=30

 

The reason why subjective descriptions/reviewing are in adequate is that the point of view of the reader and the writer are not the same. Often a subjective reviewer makes an effort to communicate their tastes and point of view to orient their readers---which is commendable---however, doing so in a subjective manner just adds layers of interpretable descriptions to interpretable descriptions.

 

Fortunately, in the observable universe, 1+1=2 and 0=1+(-1), so actual quantifiable measurements provide a universal reference point which is consistent for everybody. I think the subjective descriptions are a nice supplement to quantitative information. We don't hear with our eyes, so when somebody provides their personal interpretation of how a component sounds in conjunction with the relevant quantitative measurements, it can help readers understand how the graphs of data relate to the sound, and perhaps equally important, why the component sounds the way it does.

 

Cheers!

post #12 of 20

My headphones sound like hot coffee brewing and bacon and eggs frying smell on a cold crisp April morning in the High Sierras.

post #13 of 20
Quote:
Originally Posted by ab initio View Post
[...]

The reason why subjective descriptions/reviewing are in adequate is that the point of view of the reader and the writer are not the same. Often a subjective reviewer makes an effort to communicate their tastes and point of view to orient their readers---which is commendable---however, doing so in a subjective manner just adds layers of interpretable descriptions to interpretable descriptions.

 

Fortunately, in the observable universe, 1+1=2 and 0=1+(-1), so actual quantifiable measurements provide a universal reference point which is consistent for everybody. I think the subjective descriptions are a nice supplement to quantitative information. We don't hear with our eyes, so when somebody provides their personal interpretation of how a component sounds in conjunction with the relevant quantitative measurements, it can help readers understand how the graphs of data relate to the sound, and perhaps equally important, why the component sounds the way it does.

 

Well, the argument goes to follow those reviewers whose terminology, experiences, and preferences seem to follow your own, but it's that's hardly perfect.

 

Unfortunately with most available quantifiable evidence in the form of headphone measurements, the response at the microphone in a dummy head setup is not going to be quite the same as the response in your ear canal, and there can be significant unit-to-unit headphone variation as well. Regardless, it provides a lot of good information and in a much more compact fashion.

 

 

Quote:
Originally Posted by bigshot View Post

My headphones sound like hot coffee brewing and bacon and eggs frying smell on a cold crisp April morning in the High Sierras.

 

Seems awfully sibilant! Not to mention cold and strident. basshead.gif

 

But not overly sterile.

post #14 of 20
Thread Starter 
Quote:
Originally Posted by ab initio View Post

 

The question you asked was "Leaving aside the validity of subjective reviewing (and subjective versus objective), is the terminology we use adequate?" And the answer, I suppose, is how you define adequate. I am arguing that subjective descriptions are imprecise and change meaning from one person to the next. I agree with you that some descriptions may be more consistent on average across the population, but that doesn't make them precise. Your example about a 5kHz peak is a quantitative measurement by a system whose measurement uncertainty is unknown; hence, I would call it an unreliable objective measurement.

'Objective' itself isn't necessarily defined well. I would say that an objective measurement is one that reliably gives a similar result when the stated methodology is followed, a common source of variation (but not the only one) being human perception and interpretation. There is no such thing as a fully objective measurement because at some point a human has to observe a result. Hence being subject to interpretation isn't an immediate failing. The example I gave would therefore be more objective than 'PRaT' but it would not 'be objective'.

 

The actual relevance of a measurement to what someone hears is independent of objectivity, though. Because we all have human ears and brains and similar perception of sound (perception unfortunately includes the placebo effect and other extraneous effects), what these subjective descriptions aim to measure are more relevant than, say, a frequency response chart. It's all and good if you have a reliable measurement but it has to be useful. 

We do have an understanding of how the hearing system works and so we know what to look for in measurements and how they correlate to perception but this understanding isn't complete. Thus subjective descriptions have a potential 'advantage' in relevance, which is balanced out by the variation in definitions. (Now, if only all subjective reviews were done as blind tests done in comparison to a known standard).

 

 


Quote:
Originally Posted by esldude View Post

As mikeaj says, repetitive usage and experience among people can give a commonality of meaning.  Is it super precise? No.  Is it possible to give wrong impressions?  Yes.

That sort of common meaning arises because these people are able to collectively hear the same sound and come up with a word to describe it. As long as this common understanding is there, any word could be used to describe it. The problem is this is less useful for someone reading a review for equipment that they have never heard before.

 

Quote:
But for instance without much experience most people having heard smooth and rough sound find the words appropriately descriptive.  Some triode tube amps are very smooth in the mids so not too big a leap to understand creamy smooth as a descriptor.  And odd you mention brown.  Actually a word I have used in regard to very old triode amps at least among some friends.  Don't know I would expect others to know innately.  It refers to a very smooth sounding amp with fairly rolled off highs.  Not rolled off rapidly, but rolled off.  Old tubes that are very weak along with a design having limited high frequency response seem to get this sound in time.

It's easy to hear a word and then assume that the X you are hearing corresponds to that word. The problem is your X is not necessarily the same as what the reviewer is describing. I'm not always sure what 'smooth' means in a review and I doubt the reviewers are all that sure either.

 

The difference between 'creamy smooth' and 'smooth' is the added creamy (obviously). But what does this add that 'smooth' doesn't already have? Could using a less figurative term be more useful to most people reading reviews?

 


Edited by higbvuyb - 6/12/13 at 9:06pm
post #15 of 20

^^ Not to mention some ways are easier for the 'layperson' to describe and use.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Terminology and understanding subjective reviews