How do i read this graph? *headphones*

[Couch Potato]

 
what does frequency response mean?
 
what does those line mean?
 
how do i know which one is better than the other one?

 

[Radio_head]

So there's this
 
 
 
 
 
And I stole this from a post of ClieOS to give you a general idea.  Top is rolled off highs, middle is recessed mids, bottom is weak bass.  
 

 

[RPGWiZaRD]

Well most of the time a more flat response is what's looked for but it depends so ultimately the curves doesn't actually tell you which is best, it's just different sounds. People want to hear a little different amounts of different frequency ranges but if strictly looking at mixing music then as flat as possible curve is wanted. But if you're just listening to music for your own enjoyment this necessarily doesn't have to be true.
 
But yea that's what frequency response is about, how loud the different frequencies appear or how much they stick out compared to each other. A higher stretching curve means the louder it'll appear to be percieved to our ears and vice versa. A perfectly straight line would mean the frequencies if they'd all be played at the same time at same volume levels would appear to be equally loud.
 
Bassheads like me prefer the lows at say 0 - 200Hz range to be elevated higher up than mids (most instruments and vocals) and highs (cymbals and alike) for example. Unfortunately only by testing different headphones or playing around with an EQ you can determine what kind of frequency response would be the best for you.

 

[NinjaSquirt]

-Frequency response in ridiculously layman's terms is basically "How does a system (in this case headphones) respond when fed X frequency of sound?
 
-The lines MAY, I repeat MAY indicate how a headphone will sound at certain frequencies (i.e. rolled off highs, recessed mids, lacking in bass (Radio_head's chart))
 
-How can you tell which one is better? You probably can't. Frequency response charts are only one dimension of a headphone's sound characteristics. Sure a flat frequency response is desired for studio work and monitoring but that's the charted-based exception not the chart-based rule.
 
If we all went by frequency charts:
A. This forum wouldn't exist
B. We'd all have a lot more spare dime
C. Everyone would own a pair of Grado SR60i (WOOT! It's gotta be better than Beats which 99% of the population owns...)

 

[Head Injury]

Quote:

C. Everyone would own a pair of Grado SR60i (WOOT! It's gotta be better than Beats which 99% of the population owns...)

No we wouldn't. Rolled off bass, upper mid-range hump and big treble peak. The tamest Grado, but not tame really.
 
The LCD-2's frequency response is mostly ideal, with some tweaks to make it universal. Fill in the upper mid-range and 10 kHz troughs, as they appear on these graphs at least, and you're good. The shelved treble and upper mid-range is a good thing for headphones, if you're aiming for speakers as a reference, but it could be bumped up a little bit (2 dB at most) to make the trebleheads happier.

 

[jasonb]

As long as the measurements are taken correctly then a graph will tell you how a headphone will sound in terms of frequency response. A graph wont tell you everything about a headphone but it will show you what the sound signature will be like. It definitely is a helpful tool if you understand what you are looking at.


Sent from my iPhone using Tapatalk

 

[Loquah]

Agree that they're helpful, but it's really hard to know because what looks like a small difference on the chart (e.g. 3dB) is actually twice as loud. Notice the contents of the circled area here:

The HD800 and LCD-2 are at the top of the range while the PS1000 is declining sharply and the SR60 is at the bottom. They don't look massively different, but the Grados will sound very different to the HD800 and LCD-2 at this frequency. Carry that same kind of 3dB analysis across the entire chart and you'll see how different these headphones all are.
 
The other thing to consider is that all of our ears are very different. We all have some degree of hearing loss and it will be at different frequencies for different people. I believe that we will also all hear higher frequencies slightly differently too as a result of the varying shapes and sizes of the folds and chambers in our outer ears (i.e. the parts you can see).
 
I've personally given up on frequency response charts in most situations because it's easier and more accurate to just listen. The one thing the charts are helpful for (as mentioned by Radio_Head) is identifying the overall shape of the curve and therefore the general sound style of the headphones. Beyond that - it's got to be personal auditioning or chatting to people whose taste in sound you know and trust.

 

[Head Injury]

Quote:

Agree that they're helpful, but it's really hard to know because what looks like a small difference on the chart (e.g. 3dB) is actually twice as loud. Notice the contents of the circled area here:

10 dB is twice as loud, not 3 dB. 3 dB is twice as much power, which isn't twice the perceived volume.

 

[RexAeterna]

start from ''0'' mark. on the side will be decibels(volume of frequency range) and bottom of the graph is the listed frequency range. for headphone or speaker to be consider ruler flat it has to stay at ''0'' across the spectrum but usually that's difficult and meaningless is some ways since everyone hears different. also a flat/accurate sounding headphone won't technically have a ruler flat response due to human hearing and how close to the drivers are to our ears. it's lot to explain but if you get curious just look up free-field hearing and diffused-field on google and you'll get a better understanding.speakers is another story since they're usually flat but majorly influenced by room acoustics.

 

[Blue Boat]

Quote:

10 dB is twice as loud, not 3 dB. 3 dB is twice as much power, which isn't twice the perceived volume.

I never understood this..
 
Does it mean that the difference between an 80db tone is twice as loud as a 70db tone?
Or does it mean that an 80db tone is 16 times as loud as 0db?
 
 

 

[Head Injury]

Quote:

I never understood this..
 
Does it mean that the difference between an 80db tone is twice as loud as a 70db tone?
Or does it mean that an 80db tone is 16 times as loud as 0db?

It's a logarithmic scale (or exponential or something). Every 10 dB is double the volume of what comes before it. So

1. 10 dB is twice as loud as 0 dB (assuming there's a signal at 0 dB if we're talking digital)
2. 20 dB is twice as loud as 10 dB and four times as loud as 0 dB
3. 30 dB is twice as loud as 20 dB and eight times as loud as 0 dB
4. 40 dB is twice as loud as 30 dB and 16 times as loud as 0 dB
5. 50 dB is twice as loud as 40 dB and 32 times as loud as 0 dB
6. 60 dB is twice as loud as 50 dB and 64 times as loud as 0 dB
7. 70 dB is 128 times as loud as 0 dB
8. 80 dB is 256 times as loud as 0 dB
9. 90 dB is 512 times as loud as 0 dB
10. 150 dB is 32768 times as loud as 0 dB

 
If you want to calculate it yourself, take the dB you want, subtract the dB you're comparing it to, divide the result by 10, and take 2 to the power of whatever you get. So 55 dB is 2^((55 - 40)/10) = 2.83 times louder than 40 dB.
 
Somehow I've done more math in the past few days than I did in all the rest of college 

 

[Blue Boat]

Quote:

Originally Posted by Head Injury /img/forum/go_quote.gif
 
Somehow I've done more math in the past few days than I did in all the rest of college 

Thanks for taking the trouble. 
 
I'll have to do a bit more reading to understand sensitivity ratings on headphones.. 
 
Also... will a 130db 500Hz tone be more painful/damaging than a 130db 30Hz tone?
 

 

[Head Injury]

I don't know about the damaging effects of various frequencies. The answer's probably in here somewhere.
 
Sensitivity ratings are usually per milliwatt, which is a measure of power. To increase decibels by 10, you need to multiply power by 10. It's another exponential thing. So if a headphone has a 100 dB/mW sensitivity, it needs 10 mW to reach 110 dB and 0.1 mW to reach 90 dB. If you want to find the power needed at a specific volume, take your target volume minus sensitivity, divide it by 10, and raise 10 to that power. So
 
100 dB/mW sensitivity needs 10^((110 - 100)/10) = 10 mW to reach 110 dB, and 10^((90 - 100)/10) = 0.1 mW to reach 90 dB.
 
It's the same formula as the last one, just use powers of 10 instead of powers of 2.
 
Then if you want to find the difference in voltage, do the same thing but raise the square root of 10 to that power. Hey, I think I just figured out the relationship between dB/mW and dB/V sensitivity ratings. I was trying to do that yesterday 

 
We've sort of gone off-topic.

 

[Blue Boat]

Both my Senn and Grado are rated at 32 ohm impedance, but the Senn has a higher sensitivity. 
But I need to turn up the volume whenever I use my Senns to get the perceived loudness as I would with my SR60.
Sennheiser doesn't specify their exact units though, just a vague "db". 
 
 

 

I don't know about the damaging effects of various frequencies. The answer's probably in here somewhere.
 
We've sort of gone off-topic.

 
Yes, we have. lol. 


 

 

[jasonb]

You need 2x the power everytime you increase the volume by 3db. Its not a 10 for 10 deal.


Sent from my iPhone using Tapatalk