[Lunatique]

I have put the W4 through its paces, and after the first round of extensive testing, here are my impressions.
 
My main reason for getting the W4 is to hopefully replace the SE535 and W3, since I find them too colored or fatiguing/bright without EQ. Ideally, I wanted a pair of IEM's that I can listen to without any EQ. In a way, the W4 gives me that, but not in the way I had hoped.
 
First of all, the W4 is definitely much better than the W3, since it's obvious that Westone listened to all the complaints about W3 and tuned the W4 according to those complaints, while they also tried to keep some of what people loved about the W3. So the result is like a compromise between the W3 and what I consider a neutral and accurate sonic signature--in other words, the W4 is still colored, but just not as severe as the W3. To give you an idea of the kind of EQ curve it takes to get the W4 close to the LCD-2's sonic signature, here's a screenshot:
 

 
I can probably refine this EQ curve even more and get it even closer, but for now, this already gets pretty close (though missing a bit of clarity/resolution/punchiness overall across all frequency ranges, and I think it may be because of the differences in the inherent physical characteristics of the two driver technologies). For those of you who own both the LCD-2 and W4, give this EQ curve a try, if you want your W4 to sound more like the LCD-2 (which IMO is definitely a superior headphone in all ways possible).
 
Compared to the W3, the W4's bass has a similar shape, but only half as prominent. That doesn't mean its bass is neutral/flat/accurate though, since W4's bass is still significantly emphasized, but just not to the degree of W3 (about halfway between W3 and neutral). If you listen to any music where the accuracy of the bass response is critical, it becomes immediately obvious the bass is still too prominent. Sparse jazz arrangements that contains a double-bass makes this very obvious. The W4 renders double-bass with too much bloom/mud compared to a much more accurate headphone like the LCD-2. I also gave the W4 a boost in the lowest sub-bass, because I found it lacking some body down there (though the W3 didn't seem to need that boost). The cut in the 125Hz range is exactly half the amount of cut I use for the W3, which tells me that the W4's bass emphasis is exactly half of the W3, but still 4 dB too prominent. Compared to the SE535, the W4's bass is definitely not as accurate/neutral. SE535's bass is actually one of its strongest features, since it's very close to being neutral.
 
The mids pretty good for the most part, although that pesky 7KHz ear canal resonance is there, just like with most IEM's. Part of me wishes that IEM engineers simply just design around that ear canal resonance, but I understand why they don't--it's because we all have different shaped ear canals. If you use my EQ curve and turn it on and off, it's very obvious how of a difference a steep narrow cut in the 7KHz region makes. Without that cut, all cymbals, hi-hats, shakers, tamborines...etc sound really congested and distorted. This may not be obvious if you have never listened to the W4 with my EQ curve, but once you do, you can't not notice it--it's as plain as day. Compared to the W3, W4's mids are not nearly as fatiguing/bright in the 7KHz range. Compared to the SE535, the W4 is less colored, since I have always felt that the SE535's mids are too emphasized, to the point of being fatiguing and overly bright.
 
The treble needs about the same amount of boost as W3 in the 13KHz range for a bit more air, so they are actually kind of similar. Compared to the SE535, the W4's treble is a bit more clear/prominent, since the SE535 would require a bigger boost in the 13KHz range to sound more neutral.
 
With all that said, I think the W4 is the first IEM I have owned that I could listen to without using any EQ and not have a frown on my face. While it is indeed colored without EQ, it is colored in a way that is fairly pleasing and acceptable for a lot of music. But most importantly, it does not break my first rule of audio devices, which is to first and foremost, do no harm. The W4 is not fatiguing/overly bright, and although on some bright material the level of sibilance is noticeably higher than better headphones like the LCD-2, it's not to the point of being painful like many other headphones. The bass emphasis while noticeable, actually is quite acceptable for most musical material, and it's only when you get into really refined music like jazz and classical do you notice it and want to turn the EQ on to smooth out the bass response so it's not overpowering what shouldn't be overshadowed.
 
I think it's very likely I'll sell my SE535 and W3, since the W4 is the only one among the three I could actually listen to without any EQ and not feel annoyed by the excessive brightness. While the W4 isn't perfect, and in general I'm disappointed that it still requires significant EQ'ing to get close to sounding neutral, but at least it is colored in a way that is acceptable to me. I think my IEM journey ends here, until one day I hear something much closer to my idea of neutral, but I'm not going out of my way to audition any more IEM's, since I really don't use IEM's that much anyway.
 

 

[phntmsmshr]

nice review, Agree with the EQ setting of bass emphasis around the 31hz point, disagree with the big dip at 8khz - makes them sound too tinny to my ears. Here's one I made for them:

 
I tried to keep the mids relatively neutral since they tend to dominate at default. Maybe I like highs to sparkle more than some, but it does make them a little sibilant at times on music prone to it. 

 

[Confispect]

Congrats sounds like the Atrio MG7 no EQ whatsoever. 

 

[i2ehan]

Thank you for posting your impression. My W4 arrives tomorrow, and I can't wait to put it toe to toe against my CK10's and IE8's. Based on your impression, and others alike, it seems to me that there isn't a clear cut indication that the W4 delivers a solid "improvement" over the W3 or even UM3X. It's another "try before you buy" experience. Though you preferred the W4's to the W3 and Shures, based on the un-EQ'ed sound preference, I'm somewhat dissapointed as well, as I really thought the W4 would bring a wow factor, much like the UM3X had upon its release, yet it clearly hans't. Whether one believes it or not, I'd really like to see how the W4's deliver post burn-in, if any change is present at all. Nonetheless, it seems to me that those with the W3 need not RUN for the W4, as both are solid performers in their own.

 

[phntmsmshr]

Quote:

Congrats sounds like the Atrio MG7 no EQ whatsoever. 

detail and separation. 
 

 

[Spyro]

It's still all personal preference.  Many find the W4 almost too neutral and bass lite which is 100% opposite of what Lunatique is hearing or looking for in an IEM.  Quite frankly I would find W4 unlistenable with EQ off.  Bright is the last adjective I would use to describe W4.  But we all hear differently.

 

[lucozade]

Nice review , interesting that some find the w4 to be sibilant , i have not heard any trace of sibilance on them in the 25 hrs or so that i have been listening to them , along with the sm3 i find the w4 treble to be the safest and easiest on the ears of all my iems .

 

[phntmsmshr]

Quote:

Nice review , interesting that some find the w4 to be sibilant , i have not heard any trace of sibilance on them in the 25 hrs or so that i have been listening to them , along with the sm3 i find the w4 treble to be the safest and easiest on the ears of all my iems .

In my case the sibilance is definitely the choice of music rather than the phones themselves, since others I have are worse. X10 probably the least sibilant but that's the rolled off highs at work.
 
 

 

[Danz03]

12dB cut at 7k!!! Seems a bit drastic, that's more than what one needs to cut the HD800's 6k spike and the W4s are nowhere near as bright or sibilant as the HD800!, are we really talking about the Westone 4s here??? I do have both the LCD-2s and the Westone 4s, W4s may sound a tad brighter than the LCD-2s but 12dB at 7k? I would have returned them right away. :-?

 

[Lunatique]

 

Quote:

nice review, Agree with the EQ setting of bass emphasis around the 31hz point, disagree with the big dip at 8khz - makes them sound too tinny to my ears. Here's one I made for them:

 
I tried to keep the mids relatively neutral since they tend to dominate at default. Maybe I like highs to sparkle more than some, but it does make them a little sibilant at times on music prone to it. 

 
The 7KHz dip is necessary because of the peak resonance in the ear canal. If you don't dip it at 7KHz, you will hear a severe sharp resonance. I don't understand how it's possible to get a tinny sound by dipping a narrow band at 7KHz, when in fact, at that frequency range is where the sibilance lives (4KHz to 8KHz). In order to take the "meat" of out the sound, you'd have to cut the lower mids, and that will definitely make everything sound tinny. 
 
Have you actually tried to reproduce my EQ curve exactly the same? Don't just do one or two bands--replicate the entire curve and then listen carefully, then compare it to the bypassed version. Try it with a range of different styles of music.


Quote:

It's still all personal preference.  Many find the W4 almost too neutral and bass lite which is 100% opposite of what Lunatique is hearing or looking for in an IEM.  Quite frankly I would find W4 unlistenable with EQ off.  Bright is the last adjective I would use to describe W4.  But we all hear differently.

W4's sub-bass maybe not be the most authoritative, but it's upper/mid bass is definitely too prominent. My point of reference is professional audio standard, where the signal chain must be as neutral and accurate as possible, without any significant dips or peaks in the entire frequency range. I know a lot of hobbyists audiophiles/enthusiasts prefer a "fun" sonic signature, where the certain frequencies are boosted for a more lively sound, but as a composer, I cannot allow any such subjective coloration in my signal chain, since the music I create must be played back on a very wide range of audio devices, so my starting point must be the most neutral possible. When using that standard to judge, the W4's bass is about 4 dB too prominent in the 125Hz range (wide band).
 
The W4 is not bright. I said it has a peak resonance in the 7KHz region, and this is very common in IEM since it's an inherent flaw of the technology and how it interacts with our biology. The W3 is definitely brighter than the W4 in the sibilance region. Compared to the LCD-2, HD650, 007MK2...etc, the W4 has more prominent sibilance, but not unbearably so like the W3.
 
Quote:

Nice review , interesting that some find the w4 to be sibilant , i have not heard any trace of sibilance on them in the 25 hrs or so that i have been listening to them , along with the sm3 i find the w4 treble to be the safest and easiest on the ears of all my iems .

Sibilance has nothing to do with treble. Technically, sibilance resides in the sibilance region from 4KHz to 8KHz and would be consider mid to upper-mids, and frequencies above that would be called air--what people refer to as treble. They are two distinct regions and characteristics. Many people seem to think treble contributes to brightness, but that's not the case at all. You can have rolled off treble and still have a very bright sound if the sibilance region is too prominent.
 
What your music collection consists of also make a difference. I have specific tracks I use to test sibilance, as they contain passages where a particular phrase would have notes that just pierces the ears, and if you don't have songs like that in your collection, then it's very likely you can listen for many hours and not notice it. 

Quote:

12dB cut at 7k!!! Seems a bit drastic, that's more than what one needs to cut the HD800's 6k spike and the W4s are nowhere near as bright or sibilant as the HD800!, are we really talking about the Westone 4s here??? I do have both the LCD-2s and the Westone 4s, W4s may sound a tad brighter than the LCD-2s but 12dB at 7k? I would have returned them right away. :-?

 
Try to replicate my EQ curve and listen for yourself. Don't just look at the numbers because often the numbers only tell part of the story. Also, the HD800 isn't an IEM where the ear canal peak resonance is the main culprit. When EQ'ing often you do surgical cuts and boosts to attain a specific sound, and while on paper it looks severe, but depending on how wide or narrow the band is and where you are doing it, the result will be exactly what you want. That's why you have to trust your ears, not just your eyes. 
 
Quote:

 Based on your impression, and others alike, it seems to me that there isn't a clear cut indication that the W4 delivers a solid "improvement" over the W3 or even UM3X. 

Actually, the W4 is a significant improvement over W3, since its bass isn't nearly as out of control, and it's also not as piercing in the sibilance region. I have only heard the UM3X briefly in the past, and it sounded like a more neutral version of the W3, and if I remember correctly, the W4 is more colored than the UM3X, but I can't say for sure unless I have the UM3X in my studio for some comprehensive testing. 

 

[shotgunshane]

If this FR graph can be believed, the spike appears higher http://sonove.angry.jp/Westone4/FR_Westone4_vs_ER4P.gif

However I tried your eq and it sounds pretty good with the couple of songs I listened to. Bass sounds excellent and it removed the mid/treble resonance that I never noticed before. Cymbals actually sound clearer. I'm not sure the dip needs to be 12 db's but I like how it sounds. That being said, it is very tough to recreate your eq in EQu on the iPhone. I have to add a few more points here ad there to get the overall shape. If you happen to have EQu and can send me the exact coordinates, I'd appreciate it.

It will be interesting if the um56 tips change my impressions when they arrive.

 

[Lunatique]

Quote:

If this FR graph can be believed, the spike appears higher http://sonove.angry.jp/Westone4/FR_Westone4_vs_ER4P.gif

However I tried your eq and it sounds pretty good with the couple of songs I listened to. Bass sounds excellent and it removed the mid/treble resonance that I never noticed before. Cymbals actually sound clearer. I'm not sure the dip needs to be 12 db's but I like how it sounds. That being said, it is very tough to recreate your eq in EQu on the iPhone. I have to add a few more points here ad there to get the overall shape. If you happen to have EQu and can send me the exact coordinates, I'd appreciate it.

It will be interesting if the um56 tips change my impressions when they arrive.

 
Can that FR graph be trusted? What is the source? That hump in the bass confirms my impression--it's almost exactly the same amount of dB as my cut. Now that I have a FR graph, I can tweak the W4 EQ curve to be even more refined (such as changing the bass cut bandwidth to a gently curve to match the FR graph--assuming it can be trusted).
 
The cut at 7KHz has more to do with the peak resonance in our ear canals, and less to do with the IEM itself. I suspect different people will need to EQ that peak resonance based on their own ear canal shapes. When I made the cut, I simply went as far as it took to get rid of the unpleasant peak resonance, and then experimented with the bandwidth of the cut until it sounded the least colored. I don't care that it's a 12 dB cut, because if that's what it takes to get rid of the peak resonance of my ear canal shapes, then that's what it takes. I would never make such a cut on a full-sized headphone, but with IEM's, it's a different set of rules.
 
I do have EQu, but I far prefer Equalizer over it. Equalizer is just a much easier app to use in terms of precision. I highly recommend it over EQu. You can replicate my EQ curve almost perfectly with Equalizer (and I can give you my Equalizer settings to duplicate).
 
 
 
 
 

 

[shotgunshane]

That would be great. I haven't messed with Equilizer much, so this will give me an excuse. You can post them here or pm them but others will probably be interested.

Here is the thread I found the graph. Its at the bottom of the page. Not sure how reliable it is.
http://www.head-fi.org/forum/thread/552328/westone-w4-vs-audio-technica-ath-ck10/45#post_7451898

 

[Danz03]

I tried your EQ setting, although it does tighten up the bass and make it more LCD-2 like, it also makes the HF less LCD-2 like and it completely takes away the nice soundstage of the Westone 4, making it sound a bit dry. Like how you mentioned we all have different shape ears, your setting probably works perfectly for you and you only. Another thing you haven't mentioned are the tips you used, what I find is that the tips do make a lot of difference to the sound character of IEMs, more so than EQ settings sometimes.
 
Quote:

Try to replicate my EQ curve and listen for yourself. Don't just look at the numbers because often the numbers only tell part of the story. Also, the HD800 isn't an IEM where the ear canal peak resonance is the main culprit. When EQ'ing often you do surgical cuts and boosts to attain a specific sound, and while on paper it looks severe, but depending on how wide or narrow the band is and where you are doing it, the result will be exactly what you want. That's why you have to trust your ears, not just your eyes. 

 

 

[Lunatique]

Quote:

I tried your EQ setting, although it does tighten up the bass and make it more LCD-2 like, it also makes the HF less LCD-2 like and it completely takes away the nice soundstage of the Westone 4, making it sound a bit dry. Like how you mentioned we all have different shape ears, your setting probably works perfectly for you and you only. Another thing you haven't mentioned are the tips you used, what I find is that the tips do make a lot of difference to the sound character of IEMs, more so than EQ settings sometimes.
 

 
I favor the triple-flange with the smaller flange cut off, since it provides the most isolation while I don't have to wait for any kind of foam to expand in order to seal properly. I have never found tips to change the sound, since a seal is a seal. I guess if the tips vary in how deeply they are inserted, then they might alter the length of the ear canal.
 
If you raise the volume a bit when using my EQ curve, I think it'll fill out the entire spectrum nicely, and it'll also restore the lushness to the soundstage. This probably has to do with the fact that in its natural state, the W4's bass and upper-mids are prominent, so let's say at volume 40%, everything already sounds full and lush because those prominent regions are already peaking your comfortable listening level threshold (in other words, the bass and upper-mids are actually hitting 70% due to being more prominent, while the sub-bass, mids, and treble remains at 40%). Once you use my EQ curve, the frequency response flattens out and the average level now drops to 40%, and you'd need to bring it back to 70% by raising the overall volume, so everything sounds full again (and also because of the Fletcher-Munson curve).