[Degru]

This is an eq setting I just made. Also be sure to insert deep enough to ensure the similar response as measurements. Too much taking. Just try it out. Apply both of the files to er4b. eq.zip

Will do later this evening after work

 

[Degru]

This is an eq setting I just made. Also be sure to insert deep enough to ensure the similar response as measurements. Too much talking. Just try it out. Apply both of the files to er4b. eq.zip

Alright, listened to the EQ. I can't really pick out any difference from my own, but I heard the same difference between the two here. ER4B is more dynamic, ER4S is more resolving and cleaner sounding, but more compressed. The difference is even more obvious when converting S to B via EQ vs the real thing, which I wouldn't mind also having accurate EQ settings for.

My EQ is merely a two point graphic EQ slope from 1k to 10k of 5dB, and a Q 0.5 ~1dB peaking filter around 3.5khz to reproduce the slight hump in the EQ there. Yours is actually reducing the upper mids and low treble slightly more than mine, but I didn't hear a noticeable difference in tonality between the two.

 

[JohnYang1997]

Alright, listened to the EQ. I can't really pick out any difference from my own, but I heard the same difference between the two here. ER4B is more dynamic, ER4S is more resolving and cleaner sounding, but more compressed. The difference is even more obvious when converting S to B via EQ, which I wouldn't mind also having accurate EQ settings for.

If you are going to convert S to B you need to rist of all make the eq subtractive instead of additive, meaning decrease the overall gain to not have red in the fft. Hence, you also need to match the level, which is impossible to do without measurements. (or you can lower the same amount for both) That's why you hear a difference. You can only eq from er4b to er4s. Not the other way around.

Apart from the above, the output impedance of the amplifier will lower the effectiveness of the RC filter. If you substract the RC filter assuming it's with low output impedance may leading extra lowered high frequency response. It depends on your amplifier.

Also the components all have tolerance, the unit you have may not be the same as the one that's measured. Even, there may be small damages to the BA driver without you knowing. Less high frequency is one of the outcome. That's all why I said you really need to get your unit measured instead of the one from online.

 

[Degru]

If you are going to convert S to B you need to rist of all make the eq subtractive instead of additive, meaning decrease the overall gain to not have red in the fft. Hence, you also need to match the level, which is impossible to do without measurements. (or you can lower the same amount for both) That's why you hear a difference. You can only eq from er4b to er4s. Not the other way around.

Apart from the above, the output impedance of the amplifier will lower the effectiveness of the RC filter. If you substract the RC filter assuming it's with low output impedance may leading extra lowered high frequency response. It depends on your amplifier.

Also the components all have tolerance, the unit you have may not be the same as the one that's measured. Even, there may be small damages to the BA driver without you knowing. Less high frequency is one of the outcome. That's all why I said you really need to get your unit measured instead of the one from online.

I still notice much more impactful bass response from *stock* er4b compared to *stock* er4s, which if it was only FR should be the other way around, since er4b is brighter. I am using both out of an Apple dongle right now, which has sub 1 ohm impedance.

 

[JohnYang1997]

One last thing is the absence of blind test will almost always lead to the expected outcome. The only way to perform is to record both er4s and er4b and perform eq in digital domain. Then perform ABX test to see if you really can hear a difference. This way it's the controlled test that will give you objective results.

 

[JohnYang1997]

I still notice much more impactful bass response from *stock* er4b compared to *stock* er4s, which if it was only FR should be the other way around, since er4b is brighter. I am using both out of an Apple dongle right now, which has sub 1 ohm impedance.

That's normal. Because your ear compensates for the brightness. Hence you can hear more impact. That's very normal. That's my impression a few years ago. Nothing wrong with that.

 

[Degru]

One last thing is the absence of blind test will almost always lead to the expected outcome. The only way to perform is to record both er4s and er4b and perform eq in digital domain. Then perform ABX test to see if you really can hear a difference. This way it's the controlled test that will give you objective results.

Alright, I will see if my imm6+tube rig at home can give me a usable EQ for my specific pair.

 

[JohnYang1997]

Alright, I will see if my imm6+tube rig at home can give me a usable EQ for my specific pair.

I meant to record music. Then see if you can tell difference between the two music files.

 

[Degru]

That's normal. Because your ear compensates for the brightness. Hence you can hear more impact. That's very normal. That's my impression a few years ago. Nothing wrong with that.

Just got around to inverting the EQ to convert S to B. Comparing EQ'd S to real B, the differences are even more apparent. In this case your EQ is in fact more accurate than mine and tonally the EQ S and B sound basically identical. But, there is none of the dynamics and bass impact that I hear with the real B. It's a really obvious difference. On one the kick drums kick, on the other they just... don't. Playing some extremely dynamic pieces of audio like the erich kunzel/cincinnati pops 1812 overture recording, the cannon blasts have more snap than stock ER4S due to the brighter sound signature, but there is no significant impact and rumble like with the B. "Becoming one of the people.." by James Horner from the Avatar OST has some drums around 3:27 where the B is able to produce a deep satisfying impact and then rumble on the decay, where the EQ S just sounds flat, even though the overall quantity of bass is the same.

With the brighter sound signature, the S's lead in detail resolution is also quite apparent.

Like I said, the differences on highly dynamic music are blatantly obvious when you hear them. It's not purely an FR difference.

 

[manueljenkin]

Well they do. It's actually well established fact in audio science. Fr include almost all details of a device. In signal processing any linear system can be described with a impulse response which is equivalent to fr. In control system, a system can be modeled with a step/bump input signal. In audio research, one can't reliably tell distortion under 0.1% while for all trained listeners 0.2db difference is very easy to tell. To put into perspective 0.1% is 0.008dB. For any given system, it can be seen as a linear system + nonlinearities. Nonlinearity = distortion. Just accept the fact that your eq wasn't perfect. Hence causing issue. Also what kind of eq were you using. You have to use minimum phase eq otherwise will cause excess phase which can be audible.

I've studied control systems. Sorry to break your bubble. Headphones are not LTI systems. Minimum phase etc all comes only after your system is a well known lti. A mechanical device has intertia, the cups have reverb, there's tonnes of nonlinearities. Long story short, lot of sound character lies in transients, and we measure precisely 0% of that.

 

[JohnYang1997]

I've studied control systems. Sorry to break your bubble. Headphones are not LTI systems. Minimum phase etc all comes only after your system is a well known lti. A mechanical device has intertia, the cups have reverb, there's tonnes of nonlinearities. Long story short, lot of sound character lies in transients, and we measure precisely 0% of that.

Like I said any system can be modelled as a linear system + non linearity. The linearity is distortion, thd + imd.
The rest is linear system. Being minimum phase has nothing to do with LTI/LSI systrm. If you want to talk about the result. It comes down to the input dimension and output dimension, if the dimension matches, there should be ways to compensate.
Then we ask the audibility of distortion. If it's below audibility or two devices have similar distortion characteristics then it's out of equation. The rest is just linear issue.
If you read carefully my original post you will see I worded very carefully.

 

[manueljenkin]

Like I said any system can be modelled as a linear system + non linearity. The linearity is distortion, thd + imd.
The rest is linear system. Being minimum phase has nothing to do with LTI/LSI systrm. If you want to talk about the result. It comes down to the input dimension and output dimension, if the dimension matches, there should be ways to compensate.
Then we ask the audibility of distortion. If it's below audibility or two devices have similar distortion characteristics then it's out of equation. The rest is just linear issue.
If you read carefully my original post you will see I worded very carefully.

You're assuming all aberrations are "harmonic distortion". It is not. There's way more stuff than that. There's overshoot (some of which shows up as harmonic distortion, some not), lag (which is typically harmonic inhibition), time to stability, etc. Being bright (lots of treble in bode plot) and being sharp (fast transient attack/slew rate) are not the same thing.

Regardless, I'll excuse myself from this discussion. Meantime, have fun with your "fr=everything meme". I've seen you in youtube claiming nonsense, like x headphone is harmful for ears while y headphone is not without any real science backing. I've been to multiple doctors and there's very little stuff known about what's the impact of anything inside ears.

 

[JohnYang1997]

You're assuming all aberrations are "harmonic distortion". It is not. There's way more stuff than that. There's overshoot (some of which shows up as harmonic distortion, some not), lag (which is typically harmonic inhibition), time to stability, etc. Being bright (lots of treble in bode plot) and being sharp (fast transient attack/slew rate) are not the same thing.

Regardless, I'll excuse myself from this discussion. Meantime, have fun with your "fr=everything meme". I've seen you in youtube claiming nonsense, like x headphone is harmful for ears while y headphone is not without any real science backing. I've been to multiple doctors and there's very little stuff known about what's the impact of anything inside ears.

Overshoot is not frequency response? Why don't you go back and ask your professor about it? Also stability, ringing, natural frequency etc.
Also like I said THD IMD. And decay time can be restored using impulse response. I don't know where did you studied all these. I can tell you really don't know a lot. Let alone er4 has one of the lowest decay time in earphone domain.
About stuff inside ears. The acoustic impedance had been measured since decades ago. And how ear drum, cochlear, hair cell work is very well known. Only what's after the hair cell is very well studied. I never said fr is everything but it's 99%. Do yourself a favor and do some controlled tests.

 

[JohnYang1997]

The most funny thing about it is. When I said about hd800 being harmful to ears. Do you really think a huge peak drilling on your haircell is a good thing? Also there has been studies showed human hearing can compensation up to 4db. People who got used to the sound may stick to certain sound. It's all well known and established.

 

[chinmie]

some earphones /headphones being dangerous to the ears are very real. some just have dangerous spike on them. i by all means know zilch about measurement, but i do know symptoms. if an earphone gives me mild to worse ringing just from using it for some short period of time, i would get rid of it immediately. some of the worst case i had is getting rid of them after just a day of having them.

 

[manueljenkin]

The most funny thing about it is. When I said about hd800 being harmful to ears. Do you really think a huge peak drilling on your haircell is a good thing? Also there has been studies showed human hearing can compensation up to 4db. People who got used to the sound may stick to certain sound. It's all well known and established.

I doubt it's drilling the hair cells more than, say when you turn up the volume of hd600 more to make up for its treble deficit.