No? And why would we care if another uninformed group did the same thing?Indeed music producers use them all the time, no?
No? And why would we care if another uninformed group did the same thing?Indeed music producers use them all the time, no?
My sense is that music producers HAVE TO use these words and think in this way. They want to engineer subjective impressions in their audience, and the objective manipulations are just means to an end.No? And why would we care if another uninformed group did the same thing?
I feel Harman has focused on preference rather than what words people would use to describe a subjective impression. And so there really isn't a body of scientific literature for subjective reviews to contradict.And the science absolutely cares about subjective impressions. Where do you think Harman research came from? It was controlled subjective experimentation.
1. crazy is an intensifier, like "very", "significantly", "considerably", which are all used in science. It is just an informal version of those.I am no authoritative person in this, but I can still easily translate most if not all of them into measurement features.
1. crazy is an intensifier, like "very", "significantly", "considerably", which are all used in science. It is just an informal version of those.
2. two FR ranges both being boosted
3. one FR range slightly boosted above the rest
4. this FR range is masked by another boosted FR range
5. need context for pummel, but again likely associated with boosted FR range that is heavily implicated in repeating rhythmic patterns
6. a certain instrument's frequencies are masked by boosted FR range elsewhere
7. too much bass, too little mids/treble
8. too much upper treble
9. FR boosted in the mids and treble, leading to sharpened transients, combined with bass boost
10. not one FR range is boosted to mask others
11. no sharp change in FR boosting or cutting
12. conflicting tuning goals, e.g., warm mids and bass + overly boosted upper treble
13. slight boost to the mids and treble
14. slight boost to a combination of FR ranges
15. peaks or valleys in the treble
16. smaller peaks or valleys in the treble
17. slightly excessive bass
18. boosted treble
19. boosted treble
20. no FR range sounds boosted or cut
21. unpleasant FR deviations from a certain target
22. the frequencies of a certain instrument cut or masked by boosted frequencies elsewhere
23. these pertain to spatial effects, and we know for a fact that even Harman hasn't researched into this, so there are no scientific things to say at all.
24. too much bass
25. ??? you sure this is related to sound?
There is no way you can do that. Go ahead and draw the graph, and provide back up for this statement in Listener's review:The associations are truly arbitrary. Yet we can reconstruct what the graph looks like from many subjective descriptions.
Your feeling has no place in such a discussion. Harman training tools allow humans to describe colorations so well that they correlate quite nicely with measured results.I feel Harman has focused on preference rather than what words people would use to describe a subjective impression.
So does that mean if I use my wife's Airpods, which she has EQed using Personalized Spatial Audio for her own head and ear geometry, I would get that same/similar quality of spatial audio effects as her? If not, does that conflict with the conclusion that each person's unique HpTF has very little influence on preference?The logical conclusion of "this means that each person's unique HpTF is influencing both in the same way" is that, for IEMs at least, measurements on a standard GRAS fixture tell most of the story. The "fear" that these measurements does not apply and aren't valid to you (generic you), and therefore must be viewed with a heavy dose of skepticism, is vastly overblown.
Spatial hearing is a separate issue. Cup your ears or deform your pinnae will significantly degrade your hearing location accuracy. They are highly sensitive to very narrow band features in the frequency response curve (results from the frequency response alterations by your outer ear).So does that mean if I use my wife's Airpods, which she has EQed using Personalized Spatial Audio for her own head and ear geometry, I would get that same/similar quality of spatial audio effects as her? If not, does that conflict with the conclusion that each person's unique HpTF has very little influence on preference?
Subjective sighted reviews are very problematic, the science taught us this. Unless the listening tests are done according to proper procedures, they are of very limited value.The actual peer review process in science actually does not care about your past credentials. Instead, new research is evaluated solely on its own merits. The peer review process is double-blind, after all. And so I think critiquing the preference bounds or whatever DF+tilt target that HP.com have been doing in terms of the conceptual and empirical support for them is totally valid and should be encouraged. But repeatedly calling into question their credentials is not helpful, nor is it the standard practice in actual science.
And you can't equate subjective reviews with spreading flat earth theory. Flat earth theory contradicts established science. Subjective reviews don't, they are in a domain that science doesn't even care about, but the layperson does. The layperson wants to know "how the headphones sound like, subjectively" and doesn't have the skills to translate measurements into subjective expectations. And so the reviewer helps them do it. And the way HP.com does it is still relatively well-grounded in the measurement: I can easily identify which features in the graph correspond to which subjective descriptors they are using. Listener has also already illustrated it earlier.
These are for objectively describing where colorations are, not about describing the subjective impressions about what impact those colorations have on music.Your feeling has no place in such a discussion. Harman training tools allow humans to describe colorations so well that they correlate quite nicely with measured results.
From AES Presentation:
View attachment 568977
Listeners are trained for discrimination of tonality errors in different audible bands:
View attachment 568978
Using this, human subjects were able to describe with high accuracy impressions that very closely matched measurements:
View attachment 568979
This is the power of knowing what you are talking about when it comes to describing fidelity of a speaker/headphone/IEM.
We have science. You just have to learn to use it.
And yet this year, just a couple of months ago, in his presentation entitled "How Much Headphone Personalization Do Listeners Need?", Sean said this:We have seen what he said in 2025, which I repeat here:View attachment 568135
He did his own research with Dan Clark on these variations and came to such conclusions. I think the 2021 comments should be superseded by these 2025 comments.
In the context of their reviews, we know what the target is. It is the tilted version of their own HRTFs or the population average HRTF.1. crazy is an intensifier, like "very", "significantly", "considerably", which are all used in science. It is just an informal version of those.
To what specific quantifiable degree, using what specific metrics, in what direction, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
2. two FR ranges both being boosted
To what specific quantifiable degree, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
3. one FR range slightly boosted above the rest
To what specific quantifiable degree, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
4. this FR range is masked by another boosted FR range
Define “masked”, as well as the definition in totality as far as what specific quantifiable degree, using what specific metrics, in what direction, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
5. need context for pummel, but again likely associated with boosted FR range that is heavily implicated in repeating rhythmic patterns
Define “rhythmic patterns” objectively, and the definition in totality as to what specific quantifiable degree, using what specific metrics, in what direction, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
6. a certain instrument's frequencies are masked by boosted FR range elsewhere
Define “masked”, as well as the definition in totality as far as what specific quantifiable degree, what instrument, where else would the boost be, using what specific metrics, in what direction, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
7. too much bass, too little mids/treble
To what specific quantifiable degree, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
8. too much upper treble
To what specific quantifiable degree, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
9. FR boosted in the mids and treble, leading to sharpened transients, combined with bass boost
Define “sharpened transients”, along with the definition in totality what specific quantifiable degree, using what specific metrics, in what direction, compared to what standard?
10. not one FR range is boosted to mask others
Define “masked”, and to what specific quantifiable degree, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
11. no sharp change in FR boosting or cutting
To what specific quantifiable degree, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
12. conflicting tuning goals, e.g., warm mids and bass + overly boosted upper treble
To what specific quantifiable degree, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
13. slight boost to the mids and treble
To what specific quantifiable degree, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
14. slight boost to a combination of FR ranges
To what specific quantifiable degree, using what specific metrics, compared to what standard, to what specific FR ranges? What would the compliant and non-compliant descriptor be in relation to that standard?
15. peaks or valleys in the treble
To what specific quantifiable degree, where in the treble, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
16. smaller peaks or valleys in the treble
To what specific quantifiable degree are they smaller, where in the treble, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
17. slightly excessive bass
To what specific quantifiable degree, where in the bass, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
18. boosted treble
To what specific quantifiable degree, where in the treble, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
19. boosted treble
To what specific quantifiable degree, where in the treble, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
20. no FR range sounds boosted or cut
To what specific quantifiable degree, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
21. unpleasant FR deviations from a certain target
Define “unpleasant”, then in totality to what specific quantifiable degree, where in the treble, using what specific metrics, compared to what standard, what certain target? What would the compliant and non-compliant descriptor be in relation to that standard?
22. the frequencies of a certain instrument cut or masked by boosted frequencies elsewhere
Defined “masked”, to what specific quantifiable degree, where in response, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
23. these pertain to spatial effects, and we know for a fact that even Harman hasn't researched into this, so there are no scientific things to say at all.
If all of these terms had are easily correlated to objective measurements as you claimed they were, define spatial effects, then what specific quantifiable degree, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
24. too much bass
To what specific quantifiable degree, where in the bass, using what specific metrics, compared to what standard? What would the compliant and non-compliant descriptor be in relation to that standard?
25. ??? you sure this is related to sound?
I’m not sure any of them are related to sound, which is why I’d imagine they’re subjective terms and not objective science.
This isn’t purposefully obtuse, it’s demonstrating the difficulty in establishing a secondary language based in non-scientific metrics and personal impression without a floor, target or ceiling and applying them to scientific objective evaluation, or even a subjective one based in correlative reality.
We have some loose attempts at assigning subjective terms and instrument representation to areas of the frequency response, but there is no formal consensus or standard for most terms used or how they’re quantified and most “technicalities” have little to no basis in anything measurable or substantiated as legitimate:
View attachment 568988
View attachment 568989
It’s enormously difficult to find a sweet spot for communicating a device evaluation in a way that’s engaging and universally understood to a large consumer audience, more difficult to do this in ways that satisfy objectivists and subjectivists. To his credit, I don’t know of anyone in audio review that does this better than Andrew Park - The objective presented with explanation, the subjective presented with qualification. Mark Ryan also does a very good job trying to bridge measurements into subjective interpretation layered with consumer education.
Could be true, but still incomparable with spreading flat earth theory. of limited value =/= false/refuted.Subjective sighted reviews are very problematic, the science taught us this. Unless the listening tests are done according to proper procedures, they are of very limited value.
I think the chest has an acoustic resonance between 100-200 hz, and a boost there could be where the chesty impression comes from. Sense of physicality makes intuitive sense to me as whether an acoustic instrument sound as if it is physically played rather than synthesized. Say for example if the sustain of a drum hit is unrealistically loud compared to the attack due to excessive bass in the tuning, this will then "lack physicality".Chesty impact, a phrase that makes no sense for a headphone, could be attributed to anything in bass. Sense of physicality is just nonsense audio poetry. Bass that stayed firm into lower midrange? Firm means what?
Back to palm reading?I think the chest has an acoustic resonance between 100-200 hz, and a boost there could be where the chesty impression comes from.
You too have a future in audio fiction writing....Say for example if the sustain of a drum hit is unrealistically loud compared to the attack due to excessive bass in the tuning, this will then "lack physicality".
Nope. They were the words they used as trained listeners and they correlated with listening preference:These are for objectively describing where colorations are, not about describing the subjective impressions about what impact those colorations have on music.
No one said those things. Actually, HP.com folks keep saying their made up impressions are due to stuff you write. The peak from measurements to justify some random stuff they want to say about a headphone.The important thing is, the layperson has no way to interpret what "boosted 3k, cut 200hz" means, or what a headphone with those features will sound like.
And so, chesty = boosted 100-200hz. Physicality = the colorations (if any) are not making acoustic instruments sound unrealistic. I have de-fictionalized these terms, no?Back to palm reading?
You too have a future in audio fiction writing....
top end = entire treble range, air = upper treble; take this mm-500 example, you see that lower treble is relaxed yet upper treble is boosted."Audeze is known for favoring of a more relaxed top end compared to other planar manufacturers... except in the air where they're often quite forward, for me."
So it is simultaneously relaxed and angry at the same time?
No. You are so far into it, you don't know what is or is not fiction.And so, chesty = boosted 100-200hz. Physicality = the colorations (if any) are not making acoustic instruments sound unrealistic. I have de-fictionalized these terms, no?
That is all extrapolated stuff with no listening tests to verify validity. Actual work on this front as I reported was peformed by Dr. Olive and Dan Clark in the study that came out last year. As I explained, they arrived at different targets based on how many headphones were averaged and what headphone was being measured. Results were not conclusive.Anyone who uses a GRAS rig with KB500x pinnae should theoretically be using the green target.