[jaaibananzu]

A flat headphone frequency response should look like at 16:35 for my ear to interpret it as flat?

 

[SilverEars]

A flat headphone frequency response should look like at 16:35 for my ear to interpret it as flat?

It's an old version of the Harman target curve. It was a result of a statistical study done to figure out a sound preference curve, and due to being 'statistical,' it will not apply to everybody or specific to individual. It's more of a 'within the ball park' prediction response that may jive well with most people out there.

A way to figure out your preference is: Hear the headphone long enough to figure it out. Look at the (most accurate) measurements after, and if you do this enough with various headphones, you get a sense of things. Measurement accuracy is quite important. If you are keeping a database in your head, it should be accurate data.

It's best not to do it the other way around. Hearing the headphone first is much better to rid of measurement placebo. Because measurements will influence your thoughts on what you heard.

 

[castleofargh]

A flat headphone frequency response should look like at 16:35 for my ear to interpret it as flat?

We do not know what you will interpret as flat. Because to you, flat is however you've been hearing the world all day long. It's not flat sound at your eardrum, but "flat" sound modified by your body, head, and ear shape. If you have a fairly average head, you might find that some target curves like Harman, sound very flat indeed(or at least pleasing to you if not flat). Otherwise, you'll have to figure out what feels flat yourself. Famous audiophiles can't do anything for you beside convincing you that something sounds neutral even thought to your ear, it doesn't.

IMO such a curve can be a good starting point. I say starting point, first because the Harman target was created on a specific measurement rig, so graphs from other places might not be an exact fit(in general just assume it's not). And also because while very much human, you may still need a few dB here and there to account for your own unique body.

Be careful when EQing high frequencies or subs. Subs because your headphone might not handle a loud sub output without massive distortions. So depending on your listening level, keeping a roll off might simply sound better overall. You can always assess that by yourself and make a decision based on your preference. so keep that in mind, but ultimately do what you want ^_^.
And be careful with high frequency EQ because small physical changes lead to big variations at high frequency(like shifting a huge spike at another frequency). So you should avoid big but very local EQ boosts at high frequency based on measurements. And even by ear, chances are that a change that's great today, will not be so great tomorrow when you don't put your headphone or IEM exactly the way it was the day before. If in doubt, just avoid any big boosts at high frequency. The reward just isn't worth the risk IMO.
beyond that, just fool around with an EQ, give yourself time to get used to it and decide if it's an improvement or not. then move on to another EQ change, rince, repeat until the bother of doing it outmatches the audible improvements


Try to stay with a listening level that's normal to you. It is very easy to start increasing the volume output when we're actively trying to identify stuff, and it certainly makes everything easier, but it also makes you EQ for the wrong loudness. We have a different "flat" at different listening levels, so sticking with your usual volume level is IMO the right move.

 

[gregorio]

I'm not quite sure where this audiophile obsession for "flat frequency response" comes from, flatter than the mountain range freq response of speakers in a typical consumer listening environment sure, but flat? It seems to come from a false assumption such as the following example:

But occasionally when I need to do studio/stage monitoring, I would go for a flat FR, so it can truly convey what the artiste intended in the music ...

EG. The artists heard a flat response in the studio when creating the recording and/or intended the recording to be listened to in an environment with a flat FR. However, neither of these assumptions is true! Firstly, what the artists actually heard is: The "house curve" of whatever studio they were working in and some/most of the time (if they spent much time in the control room to start with) the freq response of relatively mediocre near-field monitors in a far from flat/ideal position (typically on the meter bridge of a large mixing desk). What they would also have heard is the rough mixes and/or final mix on their car system and home (consumer) system. Secondly, it would make no sense to make a commercial music recording (for sale to consumers) that's intended to be played back in an environment with a flat FR, because pretty much no consumers have a listening environment with a flat freq response. This is why commercial studios are not flat and have a "house curve" in the first place.

If you play back a commercial music recording using a system/environment with a flat FR you'll often/usually be experiencing (very roughly) 3dB - 6dB less bass than the artists/engineers intended.

G

 

[SilverEars]

I'm not quite sure where this audiophile obsession for "flat frequency response" comes from, flatter than the mountain range freq response of speakers in a typical consumer listening environment sure, but flat? It seems to come from a false assumption such as the following example:


EG. The artists heard a flat response in the studio when creating the recording and/or intended the recording to be listened to in an environment with a flat FR. However, neither of these assumptions is true! Firstly, what the artists actually heard is: The "house curve" of whatever studio they were working in and some/most of the time (if they spent much time in the control room to start with) the freq response of relatively mediocre near-field monitors in a far from flat/ideal position (typically on the meter bridge of a large mixing desk). What they would also have heard is the rough mixes and/or final mix on their car system and home (consumer) system. Secondly, it would make no sense to make a commercial music recording (for sale to consumers) that's intended to be played back in an environment with a flat FR, because pretty much no consumers have a listening environment with a flat freq response. This is why commercial studios are not flat and have a "house curve" in the first place.

If you play back a commercial music recording using a system/environment with a flat FR you'll often/usually be experiencing (very roughly) 3dB - 6dB less bass than the artists/engineers intended.

G

That's the idea behind modification of Harman curve with the sub-bass lift and I believe something in the high frequency as well.

Recall the slides with the upper-body, head, ear gain response? That's the reason why target response rolls-off around 10k or so, and I recall people would talk about treble extension, like there was something wrong it the treble didn't continuously extend toward the air region and beyond. Which goes against that chart and the target response philosophy. Also, there is the question of accuracy of 8k and beyond.