I agree with that as well. If you take part in more subjective discussions it's quite important to know other peoples' preferences, using your own as a reference. Objectivity kind of speaks for itself with no regard for listener taste.

How do you subtract preferences from the review?
Simple, by falling back on relative objective measurements. They might be incomplete, but ones made here are comprehensive enough.
After you do that, the rest of the review gets clearer.

The title of this discussion could equally read, ' Meaningless Measurements' if the measurements being taken are not really noticed by listeners. The sine wave measurements used as proof of a good amp in the 70's/80's come to mind here. Also, something I find hard to not take offence to is the quoting of figures at people to prove them wrong and put them down.

It can be such a touchy subject that it can be impossible to discuss on many forums. It seems, not here though!!   

Yes, meaningless measurements do happen. In case of headphones though there are many more things that are measurable and subjectively different.
Ignoring them or making excuses does not help at all.

If someone takes offence at objective reality, they should exit it. Fast. I recommend drugs for this purpose.

Indeed, that's quite a harsh statement. However, it annoys me to no end where people that are completely non-objective (heck, even by choice) write reviews.
Especially when they don't acknowledge their preferences.

This includes all too common shills. People with vested interest, including fans of a brand or people who have no business describing sound due to lack of skill.
Like any skill, comparative review can be learned and trained.

I guess all I'm saying really is that there are some who take measurements as the 'gospel' and are not actually capable of describing what effect the measurements would have on the sound.

That's an important part of what the objectivist should do imo. I've seen one person who is able to communicate really well what a set of graphs show in plain English without resorting to some kind of inverted 'snobbery' or funniness if anyone dare to ask a question.

There's the whole problem in that words are fuzzy. Fuzzy words are better than no words though. However words without any measurement are quite worthless until you "decode" the reviewer, which is often impossible. For people like Joker or purrin, I happen to have some sort of a "decoder ring" because of shared experiences. (Listening to enough IEMs and headphones for comparison on comparable gear.) It's not accurate though, and it can never be - people change.
Measurements made using the same methodology stay extremely consistent assuming the gear is good.

It is very expensive and/or time consuming to "decode" any single review. Even more so if it's full of worthless terms and weasel words.
It's comparatively easy to "decode" a set of measurements. You can even cross-correlate between different measurement gear to a certain extent.

It's that part of objectiveness that subjective listeners take real offence to I think. (Like your suggestion to take drugs   )

Seriously though, many people hate having the flaws of their favorite thing pointed out. Objectivist or subjectivist alike. The "cold shower" approach can sometimes work. This is what piracy is partly about, not mincing words. There's bilion of other, nicer ways to explain the sentiment I felt, but every other would be less genuine.

The real skill is the interpretation of the measurements, not the measurements themselves and that's also where a little subjectiveness might creep in anyway.

You can interpret anything enough so that it becomes meaningless. The less there is to interpret, the better. (as long as what is there is accurate and descriptive enough)

I find the waterfall graphs fascinating actually but if you look at Goldenears, the facts are there in graphs, but not really the whole story. Find a headphone you've not tried and look it up on there. Make up your mind what you think it sounds like from the graphs and then listen to it.

The whole problem is that without measurement equipment or lots of experience there's no way to point out broken measurements.
Purrin here is keeping conservative and scientific here, e.g. by trimming IEMs to 10k (I'd prefer error bars/blur myself), removing lowest end from CSDs (same), providing exact metodology used to create the CSDs from impulse response (included in every CSD). The only thing missing is the complete description of measurement gear (but it makes repeatable measurements) and the exact compensation curve.
Golden Ears are missing the CSD settings in their methodology part. Frequency response though is reasonably close and they do provide "raw" data as well.

Results can be surprising.

People taking measurements should refrain from listening to the headphone. Take the measurements. Predict how it will sound and see how close they were.

That's one way to tune up your skills. Predict and check for errors. You can also do the other way around, listen to the headphone and try to predict the measurements. What matters most is the feedback between the two.

The reality often is, they will listen first, measure and then confirm what they measured is audible.

Or vice versa, measure first, and then describe what they heard in terms of what they measured. It's unavoidable if measurement and/or opinion are mutually accurate.
It's when the two diverge much when one has to be careful - there's an error somewhere or perhaps an inconsistency, or an ignored factor not described by the measurements. (hidden variable)

Nothing is perfect, so "best" has to be qualified. Since it's a human interaction device, ergonomics matter as well.
It's easier to say which are great. It's been done already somewhere. ("state of headphone hobby" I think) I meant "Pirate Booty".
 

Plus, there can be budget considerations.