[Gurra1980]

The frequency responce kurve show you the frequency response, it's pretty easy. If it's properly meassured you can tell if a headphone has more or less treble, bass or midrange (in most cases) I think headrooms meassurements are pretty accurate. But it want tell you how big the soundstage is, how detailed it is, which amplifier that would make them sound the best, how fast transient response they have for example. But it takes practice and experience to read those graphs (and I'm not very good at it but I have started to learn more and more)

If you have problem with exaggerated treble don't bother with these headphones for example:

 

[Fitz]

Quote:

Originally Posted by ericj /img/forum/go_quote.gif The FR graphs at headroom are extremely suspect, even when used strictly as a comparison between models. Even when used to compare similar kinds of headphones. Compare the K501 to the HD580 on their site to see what i mean -- the headroom graph of the K501 gives it almost precisely the same bass curve as the hd580, but the hd580 is regarded by basically everyone who has heard both of them to be obviously far bassier than the k501.

Selectively reading just one portion of the graphs is a bit misleading, since they also show a big difference (5-10dB) over part of the treble. This doesn't mean one could objectively conclude the HD580 sounds bassier from the graphs alone, but the graphs do not necessarily contradict the observation that the HD580 sounds bassier than the K501, if you read them in their entirety.

 

[ericj]

Quote:

Originally Posted by Gurra1980 /img/forum/go_quote.gif If it's properly meassured

Thing is we can't actually say that for sure about the headroom graphs.

They've also admitted that they're not publishing the raw graphs - they apply a normalizing curve to the raw data that they haven't shared with us.

Quote:

Originally Posted by Fitz /img/forum/go_quote.gif Selectively reading just one portion of the graphs is a bit misleading, since they also show a big difference (5-10dB) over part of the treble. This doesn't mean one could objectively conclude the HD580 sounds bassier from the graphs alone, but the graphs do not necessarily contradict the observation that the HD580 sounds bassier than the K501, if you read them in their entirety.

The hd580 has a bigger dip at about 3.8khz than the k501, and the k501 recovers at 4.2khz or so while the hd580 doesn't recover until 5khz, but they're otherwise similar according to the graph.

 

[Fitz]

Those ticks after 2kHz are measuring the treble in 2kHz increments, not 1kHz increments like graphs used by... everybody else in the world.

So according to the HR graphs, the HD580's treble starts dropping below the K501's at around 5kHz, and doesn't come back up to close enough to the K501's level until around 8kHz. There's also what looks to be a nearly 9-10dB difference at around 6kHz (both are nearly vertical lines, so it's hard to tell exactly how much). That's a very significant difference in treble response, not one that can be casually brushed aside to say the graphs make them look more similar than they actually sound.

 

[applaudio]

Ahem.

 

[applaudio]

I found this article, which I find rather interesting. It helps clarify how one should look at a frequency response graph.