Head-Fi.org › Forums › Equipment Forums › Sound Science › The case for eq
New Posts  All Forums:Forum Nav:

The case for eq

post #1 of 5
Thread Starter 


Reading Linkwitz's experience with building an eq circuit for the etymotic 4s: http://www.linkwitzlab.com/reference_earphones.htm I decided to emulate his eq circuit in software.  I inserted notch cuts at 7.5 and 2.5khz in an Izotope mastering plugin and sure enough it smoothed the etymotics response, taming the sharp glare and hardness of the high-mids.  Just like Linkwitz, I noticed improvements in both the low-freq bloom and the high-freq air. The eq'd etymotics are stunningly speaker-like in their response, smooth as silk and beautiful to listen to.  Etymotic describes their design vis-a-vis eq here: http://www.etymotic.com/ephp/er4.html



 



I would caution that some headphones are better than others for using eq.  It's much easier to do harm by boosting freq rather than cutting so headphones with scooped freq responses where you might need to add back would not be great candidates for eq.  For instance, I tried to eq my KRK 8400s and since this headphone scoops some of the mids out, adding back freq was more challenging and didn't yield as successful results.  I already like the KRKs though, even though I acknowledge the smiley-face that is happening there.  The etymotics, their built-in hi-mid boost  nothwithstanding, have ruler flat freq responses elsewhere on the spectrum, although they don't have the top-end extension of some more modern designs.  A quick listen to some of your reference recordings will illustrate this.  There's just something about the mid-range that expresses itself in a way very reminiscent of flat mastering monitors.  And once you cut the artificial peaks at 7.5 and 2.5 it becomes something very special.



 



One last thing, the eq software I used is part of a mastering suite by Izotope called Ozone3: http://www.izotope.com/products/audio/ozone/eq.html



Sharp paragraphic eq with very little phase problems. You can in insert VST plugins in the Fidelia player: http://www.audiofile-engineering.com/fidelia/


Edited by bias - 11/2/12 at 3:23pm
post #2 of 5

Great article. I'm trying this on the er-6 now.

 

Edit: Initial impressions are that it sounds somehow flatter. I expected it to muffle the sound a lot more than it did, but the overall effect takes a bit of high frequency emphasis off of vocals and certain drums without impacting much else. I will experiment with more tracks, but so far it is a very impressive effect imo.


Edited by gnarlsagan - 11/2/12 at 5:15pm
post #3 of 5
Thread Starter 

It's interesting, Linkwitz got very close to nulling etymotic's own 2.5khz boost by using a signal generator and sweeping the spectrum.  I suppose you can do this for any pair of headphones, though, as Linkwitz states it's hard to tell how much a certain freq is boosted or cut and some experimentation in this regard is advised.  It would be great though if someone who owned several of the greatest hits in the headphone world publish the results of their own investigations using a signal sweep to measure the freq response of their cans.  

 

The er4-s is definitely not flat, but that's a virtue of it's design.  The designer wanted to simulate the freq response of hearing at the outer ear, so he built in peaks that would naturally occur as sound waves hit the ear lobe and traveled into the canal.  I think he overdid it though and most people I think would agree.  I hope etymotic will revisit this design to address this issue.  Also, the S and P versions are eq'd further to compensate for the supposedly intrinsic brightness of the then newly introduced compact disc.  The B version, made for listening to binaural perhaps doesn't and so maybe has more top end above 14khz, where the S is deficient.  Why it is lacking top end "air" and "sparkle" is anyone's guess.  And I bet there were several made on this site.  Perhaps the single BA simply doesn't have that wide a freq response.  Perhaps it's an artifact of the time when the headphone was engineered.  Today's playback systems have incredibly wide freq response and recordings are made to take advantage of this.  I've seen other vintage designs similarly lack high end, though maybe not to the same degree.  

 

Having said all that there is something intrinsically "right" about the sonic signature of the design.  Its low mids on down is seemless and smooth.  I use them with custom silicon molds for an airtight seal and the low end bass is natural and extended.  

Also they feel like they're almost free of distortion of any kind, so you can listen for a long time.  And the freq response, problematic as it is, is very smooth.  They just have a very natural presentation and I feel like, aside from the peaks, they are not doing much to the signal.  Diana Krall sounds intimate.  Led Zeppelin sounds loud and furious.  A good reference phone, made even better when tweaked with a good eq plugin.

 

BTW, for anyone deciding to get the silicon molds, make sure you clean your ears before going to the audiologist.  I didn't and the impression process forced wax deep into my canal and deafened my left ear until I went to the doctor.


Edited by bias - 11/2/12 at 8:53pm
post #4 of 5
Quote:
Originally Posted by bias View Post

It's interesting, Linkwitz got very close to nulling etymotic's own 2.5khz boost by using a signal generator and sweeping the spectrum.  I suppose you can do this for any pair of headphones, though, as Linkwitz states it's hard to tell how much a certain freq is boosted or cut and some experimentation in this regard is advised.  It would be great though if someone who owned several of the greatest hits in the headphone world publish the results of their own investigations using a signal sweep to measure the freq response of their cans.  

 

The er4-s is definitely not flat, but that's a virtue of it's design.  The designer wanted to simulate the freq response of hearing at the outer ear, so he built in peaks that would naturally occur as sound waves hit the ear lobe and traveled into the canal.  I think he overdid it though and most people I think would agree.  I hope etymotic will revisit this design to address this issue.  Also, the S and P versions are eq'd further to compensate for the supposedly intrinsic brightness of the then newly introduced compact disc.  The B version, made for listening to binaural perhaps doesn't and so maybe has more top end above 14khz, where the S is deficient.  Why it is lacking top end "air" and "sparkle" is anyone's guess.  And I bet there were several made on this site.  Perhaps the single BA simply doesn't have that wide a freq response.  Perhaps it's an artifact of the time when the headphone was engineered.  Today's playback systems have incredibly wide freq response and recordings are made to take advantage of this.  I've seen other vintage designs similarly lack high end, though maybe not to the same degree.  

 

 

The article states that the recommended eq dips will be different for everyone though. I wonder what kind of variables should be taken into account and how these variables affect sound. For example will larger canals or ears produce a stronger resonance effect? Does the shape of the outer ear affect how much treble attenuation is needed?

 

I'll play around with the eq for a while. I need to find a good signal generator. 

post #5 of 5

Playing around with this using xnor's graphic equalizer for foobar2000. With the set frequencies I have to do the 8db cut at 8khz instead of 7.5khz, but I figure that's close enough. 

 

Excited for this thread!  etysmile.gif

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › The case for eq