[leeperry]

Quote:

Originally Posted by PiccoloNamek /img/forum/go_quote.gif Theoretically, any headphone can be brought to a mostly flat response using EQ. Also, I don't use music test tracks, I use SineGen and/or pink noise to see exactly what needs to be done. Test tracks are for confirmation after you've found the correct response. I sent a message to a mod about getting this stickied. Be sure to voice your support.

ok ok, I'm doing the work backwards I guess

well I've found the spikes in SineGen but they don't "fix" my test tracks...and in the treble sine.wav goes very high pitched and resonating....I'm not sure how to know what is normal and what is not.

like learning colors to someone blind, or sumthing

well, we'll bounce this topic until everyone sees it

 

[PiccoloNamek]

You should never equalize by song and especially not on a per-song basis. That is the perfect way to get a very inaccurate response. With the sine wave generator, it shouldn't be too hard to hear approximately where the loudest frequencies are in the peaks and remove them.

Keep in mind that some songs are just mixed and mastered poorly, and nothing will save them. IMO, a flat response sounds the best with most music most of the time, not all music all of the time. That is impossible regardless of what your headphone's FR is.

 

[leeperry]

well yeah, I'm using several tracks to check EQ, and it sounds pretty good! except that nick cave tune where in the end his voice sounds metallic to death..

but it's a far cry from no EQ

do you have that song handy ? maybe they simply used a very crappy microphone...like an SM58

so all we care for in the sine.wav is volume? not resonance?

if I make the volume linear in that sine.wav, I'll be saved?

 

[EnOYiN]

Say, isn't it a lot easier to stick a few very small mics in my ears and record the a signal going from 20 Hz to 20 kHz and then figure out which parts are louder?

That would work right? I just thought of this and figured it might be a good idea.

 

[PiccoloNamek]

The microphone would not have the same acoustic impedance or frequency response as your eardrum.

Quote:

Originally Posted by Lee so all we care for in the sine.wav is volume? not resonance?

I am not sure what you are listening for when you are listening for resonance, but resonance itself isn't something you can actually hear per se, only its effects, that is, an increase in volume at certain frequencies.

 

[EnOYiN]

Quote:

Originally Posted by PiccoloNamek /img/forum/go_quote.gif The microphone would not have the same acoustic impedance or frequency response as your eardrum.

Ah, yes. I should've thought about it a little longer I guess. That tone is extremely nasty to listen to for extended periods of time though.

 

[leeperry]

Quote:

Originally Posted by PiccoloNamek /img/forum/go_quote.gif I am not sure what you are listening for when you are listening for resonance, but resonance itself isn't something you can actually hear per se, only its effects, that is, an increase in volume at certain frequencies.

yeah, what I'm calling resonance is the same that you would get from a resonant low pass filter(like the Waldorf Dpole), the frequency resonates...so I don't only have a volume problem in the trebles, I also have resonant frequencies...like a resonant TB303 you know

as you said I can't fix those, but I can lower them so they would "appear" flat....so except for that Nick Cave song that's prolly using some cheapo microphone, I think my settings are as good as it's gonna get.

should I try add dampening in the cups again? it did wonders on my 770Pro, but its cups are made of cheapo plastic...the 770 Premium is made of ABS and sounds a lot "tighter" to me

 

[PiccoloNamek]

When I say "resonance" I'm thinking of something more along the lines of "a resonance due to the mismatched acoustic termination of the ear canal".

 

[leeperry]

well, I'm sure you know how a resonant low pass filter sounds, that's exactly what I hear when I turn off my EQ...except I can't get it off Nick Cave's voice in that song at 4:00, but well the rest sounds fine....I'll stop focusing on this song altogether

why it happens is not really what I'm concerned about, I just want it to stop...if I wanted my music to sound resonant, I'd be using a Waldorf VST plugin on all my FLAC's

honestly, I don't think the trebles resonances can be "hidden" on my 770...

 

[K3cT]

Quote:

Originally Posted by leeperry /img/forum/go_quote.gif not sure, this EQ works by ear...not figures anyway I've got a bunch of tracks I use to check that my EQ is fine...and I think I need a third band like this : this fixes the gap between the HRTF graphs Beyer got from their dummy simulator and MY ears...and I like it A LOT

I see. I'm trying to do this with my modded D2000 and it seems that by directly applying your dips in the PMEQ, it loses too much sparkle. I suppose this is because the Denon is naturally a dark can and pretty much differs with your DT770. If it's the 80 ohm version, I had heard it before and it really had a V frequency response to my ears.

Right now, I'm using Cakewalk SonitusFXEqualizer because I can input the Q, gain and frequency values directly instead of only relying on the graph.

 

[PiccoloNamek]

Quote:

Right now, I'm using Cakewalk SonitusFXEqualizer because I can input the Q, gain and frequency values directly instead of only relying on the graph.

You can do that with Electri-Q.

 

[PiccoloNamek]

I've been performing more detailed tests on my Etymotic ER4Ss using SineGen, and I have to say: these IEMs have one of the most jagged and uneven frequency responses I have ever seen. And Etymotic claims they are neutral! After more than an hour of tweaking, I cannot seem to find an EQ curve that sounds quite right, despite SineGen clearly revealing where the peaks in the response are. Here is the response that I have right now:

As you can see, this is terrible. No wonder nobody can stand the ER4s! There is a fairly broad peak centered at around 3400 to 3800hz. I cannot tell exactly where because the tone is so fatiguing to listen to. It is interesting that the official literature claims the peak is centered around 2500Hz. Does the insertion depth and type of tip used affect the perception of this peak? The amplitude is about 5db up. Testing is ongoing. There is a very large peak centered around 9700-10100hz at 9db and another at around 12800 to 13200hz at 6.5 or 7db.

Turning the EQ off results in a wretched response that is not fit to be listened to by any human being. I can hardly even believe I used to think these were good! Take a look at Etymotic's graph showing the listener's perceived response:

It looks relatively flat, but the reality according to my ears is far, far from flat. It is downright awful. Going back to my equalized Lambdas after listening to this is like being comforted by a dear friend after a dreadful and emotionally scarring ordeal.

 

[vvanrij]

Quote:

Originally Posted by gregorio /img/forum/go_quote.gif Sorry but this post was so inaccurate I had to respond. Sine waves do get reproduced perfectly, this is a basic tenet of digital audio theory. True they are not usually quite like the output of analogue gear because analogue gear can actually distort the sine waves whereas with digital it is perfect. However, some listeners prefer the analogue distortion to the perfect digital reproduction. What gets output from a DAC is definitely not stepped notch patterns but smooth perfect waveforms. G

And to that I just had to respond

. You are not into sound synthesis apparantly, because saying that a digital sine wave is perfect is just wrong. Is it almost perfect yes? Is it so perfect that it becomes almost in-audible (and some say inaudible), yes? Is it perfect? no! This is atleast what I learned from synthesis and working with synthesizers, I'm offcourse open to be proved wrong.

vs

 

[zantetsuken]

how come my headphones dont play the pink noise file?. i cant hear anything. when i put it on the speakers it works. whats going on.

P.S im using the KSC75

 

[Dr. Strangelove]

Quote:

Originally Posted by PiccoloNamek /img/forum/go_quote.gif You can do that with Electri-Q.

Generate a 20-15k sweep, turn on Electri-Q, and play in foobar.
When I played it, I heard quite a bit of distortion.