The Objective, Subjective & Dejected Thread

Absolutely. I am not sure if I wrote this here or on another forum - but ideally you want a headphone that is individually calibrated to your ears. Also, the frequency charts of audiophile headphones are “meaningless” (they give you a basic idea of what to expect). Meaningless because you don’t know how your ears will interpret this frequency response, you don’t know how they will sound to your ears. This is strongly debatable, because you probably want a basic idea of how the headphone sounds before you buy it.

But with a flat frequency response it’s different. I think it is the most sensible and appropriate one to have a frequency response graph for. The frequency response plays a larger and more important role than it does for colored headphones (audiophile headphones). If the headphone is flat before it is filtered by our ears, then it means that we will all be getting an individual version of this flat frequency response, but the FR on its own is flat. This would be the closest you will get to that flat frequency response without individually calibrating/tuning a headphone to individual’s ears. In an ideal situation you want this flat frequency response headphone to be individually calibrated to your ears so it carries on this flat frequency response considering all the things in our ears that change it - but that would require a costly business service that does this… and let’s just say that the product would not be cheap at all.


Btw, in the studio you don’t necessarily want the listening instrument to sound pleasant. Imagine you have a bright sound and a bright track, a warm headphone would kill this and you wouldn’t hear this quality. This is not something you want to use as a sound engineer.
E.g. my Dekoni Blue’s are particularly warm and never get to produce sparkle or be anywhere close to be fatiguing.

The whole idea of a studio listening instrument is for it to reproduce sound “accurately” and without bias (warmth is a bias, and just like warmth, there are many others).

1 Like

That’s the Nuraphone ~$400 and mixed reviews.

They multi-track and then mix down to the final version. Lots of ProTools processing in between…there’s no firm reality in studio production, just art.

1 Like

I think that we all have a pretty good idea how subjective our hearing is. Just reading people’s experience and preferences on forums, you learn a lot about how hard it is to understand another persons sound perception. This also goes for reviews - one person finds this “bright”, while the other one finds it “pleasant”.

Another large problem in reviews (going off-topic here, will try to make it short) is that people describe sound with words. In no way can we translate words to sound. If you use reference tracks and pin point to what element you are referring to (with the help of minute marks), you will be able to at least be more accurate with letting the reader know what the used sound terminology means in terms of sound. For example, I will use a reference track and minute mark to specify where I hear the headphones ability to produce sparkle, or be bright, or be fatiguing - any sound quality. The reader can play the song at this same minute mark and judge for himself how that sounds to him. Is it something that he likes or not. Of course, this takes more time, and the greater majority of reviewers are not willing to put this extra effort in. I do have to say it is extremely time consuming and I do understand - but describing sound with words is nothing more than entertainment.

1 Like

Yes, but no. Watch the vid I attached. You would need to create molds of your pinnae, and the whole ear structure. Then you would use this data to know where the reflections would be and how the original frequency response (or sound in general) would get altered. I am also pretty sure that ear pressure or stuff like this is also taken into consideration. With all of this and a special method/approach, you would be able to calibrate the headphone to match the way you will actually hear it. So if you want to actually hear and perceive a flat frequency response from a headphone, you would have to do it this way.

Nuraphone doesn’t do this, nor is it using professional measurement mics, etc.

It is a complicated process, that’s why I said it’s an expensive business. You would have to do this for every individual individually. After you have the molds and all the required data from your ears (all the factors that change the sound), then you can place the mics in and start the calibration process. On another forum, somebody said that this altered frequency response that is individual/personal is what you call HRTF. If you can measure this HRTF, then you can know how you will hear the wanted freq. response from the headphones, this lets you take the advantage of it and manipulate the original frequency response and ensure that the one you hear matches the original one. Idk if I got my point across, it’s definitely hard to explain with words.

Might find something of note in there if you want to read through it.

I think it could be an interesting discussion to have in and of itself on a separate thread - measurement targets and what they mean and are they of use…

Personally I think there are some flaws with it and have considered starting a thread on it a few times.

2 Likes

“In order to test this assumption, I have removed the Pinnae on my left side, and have retained it on the right side. I’m now working on a chart of FR impressions using digital trombones playing sweep frequencies duplicated in both channels. Testing with STAX SR-009, ZMF Verite Open, and Sennheiser 800s. I will report on results.” – V. Van Gogh

6 Likes

This is indeed true. I personally do not look or use frequency response graphs for audiophile-grade headphones, i.e. headphones that are aimed for music listening.

However, you cannot really judge a “reference-grade” headphone that is intended to be used as a tool for making music by listening to music. I can’t just play a few testing tracks to subjectively judge how it sounds.

I am with you on the point that you are not listening to frequency response graphs, but with claimed “reference-grade” headphones it is different (imho). However, from my research and findings, the most accurate way to judge the Ollo S4X is not to subjectively base it on my personal interpretation of its sound performance - this would be of no value to somebody who is looking to make music (he could care less about my opinion on how it sounds :confused: )

Whew! All of these points! I was just trying to be funny. I thought digitized trombone frequency sweeps was pretty amusing, although one could use slide whistles for the treble. Yes, I did intend for some stuff to be read into it, but I really enjoy running around with a pin in a roomful of balloons.

5 Likes

The reason why I mentioned that it would be better to exclude pinnae from measurement setups is because then you would be making the headphone be subjectively tuned to that very pinnae… which is not what you want when you are producing a product for the masses. You want to leave the subjective out, and make sure that the headphone itself has a characteristic. I am not a pro, if manufacturers use measurement setups with pinnae, they know what they are doing and are certainly more knowledgeable than me. In my mind it wouldn’t make sense to do that, again… mentioning my attached video, the ear greatly alters sound, so by using it a measurement setup with a pinnae, you are implementing this individual pinnae’s characteristics in the tuning of the headphone itself… just complicated.

Well hey, who said we cannot use something funny to further explain our points :wink:

Always good energy from you David =)

2 Likes

On topic for this thread but off topic for the recent discussion here. I came across this while looking for the oratory1990 links I used above and found several points very interesting.

1 Like

No you got it right, but the specific bass level wouldn’t have existed without listener preference. If you think about it, DF is also meant to match speakers in a room - as opposed to free field, which is in an anechoic chamber. Both versions of Harman are ‘warmer’ than DF and FF, but some ear gain similar to that of DF is there. It’s just that Harman ear gain is more ‘safe’.

3 Likes

For anyone wondering, these are the earlier articles on the subject:

Here’s the difference I’m referencing:

5 Likes

The SBAF thread that @perogie linked to has a few links to presentations Olive gave that summarize that research and are free as in beer. However I do not believe they are free as in speech so I won’t link to them directly here.

3 Likes

I have a very short question (that may require a much longer answer). So at this point we know that HRTF exists and that the original frequency response of a headphone (that the manufacturer got with the help of a measurement rig) will not be perceived as it is, but rather you will have your own frequency response (HRTF).

If you were to calibrate a set of headphones individually to your ears and everything (not intended for mass-production but rather exclusively for you), would you want to tune them to have a flat frequency response if you wanted the headphones to be neutral/uncolored/reference?

I know this is the case with studio monitors that have been tuned in an anechoic chamber, but is this the exact frequency response that you would want to actually hear if you wanted a “neutral” headphone? Or do different principals apply? If so, what are they, and why would a flat frequency response be something that you wouldn’t want?

Neutral should mean that no frequencies stand out, so to me it is sound that the frequency response you want the user to hear would be a flat frequency response - at least if you are aiming for the “no frequencies stand out”.

This is attributing too much to self-awareness. Conscious, verbal reports need not correspond with perception or that which is actually perceived on a semi or non conscious basis.

People will adjust tone profiles and volume based on physical comfort, positioning, and fatigue. We lack stability and precision, so analytical inferences require triangulation through neural scans, group averages, etc. You can either determine individual characteristics or group averages, but they won’t be the same.

Perception science borders on philosophy and can sometimes be clear and sometimes a gigantic fuzzy blob.

1 Like

By perception I mainly meant what we are hearing, this would be the HRTF, and I belive this is something that you can measure (of course, the most accurate measurement would take various data and research of your ears), once you have all this data and your HRTF, you would be able to make the headphones have virtually any frequency response that you tune them to your HRTF

I think you might have gotten the wrong idea by what I referred to as “perception of sound”. Please let me know if I misunderstood your answer

There is no ground truth to perception per external measures. The sensory systems such as eyes, ears, nose, etc. do tend to be quite similar across people per many measures, but the same stimulus can be preferred or disliked at the integrated final stage of awareness.

This is really complex and always a tad vague.

1 Like

That is true in the sense that you do not know the neural stuff and how the person is actually perceiving the sound, the same way we don’t know how we see things.

But. You can accurately take measurements and measure person’s HRTF, also create molds of your ears and place mics in them. Then you tune the headphones according to that data and response. What you would get then is the most accurate frequency response with all the factors that alter the original frequency response (I am pretty sure it has to do with more things than just our ear structure, pressure is probably one of them).

This lets you manipulate the headphones frequency response, this is MUCH more useful and accurate than the original tuning of the headphone that was done on microphones that excluded all these important factors. Again, mentioning the video I attached. Just by watching how pink noise sounds to everybody, imagine if the headphones were calibrated to the individuals ears and they heard the “intended” pink noise sound? This is why I am wondering if you would want the HRTF to be flat if you were trying to achieve a neutral/uncolored headphone.

Did I at least clear it up a bit more? Not sure if it makes more sense what I am asking

So if I understand your question correctly, you would like to have a calibration for your headphones that makes their frequency response as close to neutral as possible. And by neutral you mean that the perceived frequency response of the headphones would be the same as the perceived frequency response of a perfect pair of speakers in an acoustically perfect room. That is, that those speakers in that room reproduce the same overall frequency response that the recording engineer intended at your listening position. Essentially you’d like to compensate for how the frequency response is modified as sound goes around your face, bounces off of your pinnae, and enters your ear hole.

Am I correct?

1 Like