Ultimate Ears UE 900 Discussion and Appreciation Thread - Page 59

Interesting discussion about measurements. I prefer to use my ears to derive ID equalization myself. (equal sine tone loudness)

 

Neither Tyll's nor Udauda's data for Etymotic ER-4S and Phonak 121 matches what I hear in 3-5k range. What sounds flat to me looks like a large dip on their compensated graphs...

 

I stopped looking at compensation a while ago...  I just look at the raws :p  

  Glad Rin settled it, its all in the details as I had expected. 

Part 3 analysis of the UE900 has been posted by Rin one of his best works yet IMHO. A must read for all those interested in these. 

Looking at a raw measuring thinking that's how the ear will perceive it is not going to happen. A flat IEM in raw measurements like the SM3 (in its midhighs) means it has subdued midhighs due to the ears resonance there. Rin uses the ISO DF standard because its the most consistent and backed approach, he has articles on it in his blog.

 

@Astral. No surprise there, ID is a hybrid of Free-Field Equalization hybrid afterall. One of the main reasons Free-field failed even 20 years ago is because of its inconsistency among users, so no surprise theres a deviation there with Tyll. To compare to Rins compensated you would have to use his DF standard. HRTFs dont differ that much, so something is wrong in the calculation if it differs greatly, it should at least be close. 

Edited by Inks - 10/23/12 at 11:33pm

 

It's still data obtained by statistical methods and can't predict anything with certainty for a given individual. Strikes me a bit like complaining about mortality tables because you're still alive at the age of 100. 

Yup, its not always going to match so perfectly deviations are to be expected, just dont expect great ones (like a 10db difference) when the proper matched equalization are done in DF (this is based on what Rin has stated of the consistency among its various studies). FF and ID (to a lesser extent)  can be all over the place at times though, so I can see how things can greatly differ there. 

Edited by Inks - 10/23/12 at 11:42pm

 

I am aware of the fact that the raw won't give the in ear perception.  But raws (IMO) are more comparable than compensated ones from various sources.  So I've learned a little bit of what means what in regards to raw as each compensation looks very different :p

 Oh in that case, yes, I agree if you're going to compare different sources like Tyll or GE to Rin's then looking at the raws is a good idea there.

 

What Rin missed in the UE900 analysis is that 100ohms may be overkill. Something like a 33ohm addition may just be enough to get close enough results hes getting from the 100ohm addition. Its been mentioned and hopefully less resistance can be tested as 100ohms will pretty much demand the addition of an amp.

Edited by Inks - 10/24/12 at 12:14am

  Sorry Rin couldnt answer you (he probably planned to, or missed it), he honestly has had a lot of work going on right now, his backlog is huge and paying attention to every detail has him on his toes and exhausted at times. We did discuss your first question as it was the first thing I personally asked when he first started doing the tip comparisons. Rin has told me that he takes the nature of the IEM/tip in relation to depth, so larger bore tips are naturally going to be inserted a little shallower in his tests (though the nature of the tip depends as well, for ex: a large bore Meelc triflange can go deeper than any other large bore). The only exception were the PFE tip comparison analysis as it was more of a test of the tips rather than the IEMs real-world implications, since it was the first long test to serve as a reference. The Phonaks have  a very shallow insertion by nature, only certain triflanges and double flanges may take it to the reference plane, I think he may update it in the future as he still has one in his possession. 

 

I will ask him about that second question personally and give you an answer when I have it. 

Edited by Inks - 10/24/12 at 12:30am

He likely picked 100 for most linearity. If you go 1/3rd you'll lose low bass while not dropping the midbass which could be worse than not resistance at all. I'm not into resistance loading for 2 reasons. One is the effeciency issue you mentioned and the other is how it affects damping factor. On 'tight' BAs like a 4p it can be designed in but it should do more harm than good on something not originally designed that way. Of course some things are far from perfectly designed anyways.

Nah it was just because it's all he has, hes going ti build the other planned resistors. The damping effect is positive so I don't see why not (unlike the TF10) likely it may have actually been designed with more resistance by the engineer but beta testers may have wanted more bass and volume (also explains the pinhole effect as that may have been an added addition). Happens many times, as you said, things aren't perfectly designed.

No doubt. Could be the case as well. I do think we get too caught up in measurements. More a tool than definitive in my book. Here's Tyll's take on things. http://www.youtube.com/watch?v=HVTSafas1j0 I also tend to share his preferred sig of slightly above flat on the bass 4-6db and a rolled top on headphones to sound like the real thing to me.

Edited by goodvibes - 10/24/12 at 3:11am

 

Thanks!  Me, I tend to just take the tips that fit most comfortably and EQ down the resonances.  Maybe it's not quite the silver bullet though, as I seem to feel the SHE3580 sounding a notch less impressive than before after I swapped up from small to medium stock tips, even after I adjusted the EQ to compensate.  So I look forward to hearing the answer to my second question!

Now i got mine!

They are amazing, at least for me :), great sound and the design is a lot better then the Triple 10, the cable is nice, didnt think i would like it but its good and its not have or stiff in any way.

Feels very good to wear, fits my ears perfect, the absolut opposite to what the 10 did.

 

Thats what i think of them at least

Edited by l1nuxfre4k - 10/24/12 at 8:05am

Ordered one. Delivery might take up to 14 days, though.

UE900 available from dick smith (Australia) for 499, which is $100 more than US suggested retail price and 200 less than Australia suggested retail price!

When i see such price variation I get very suspicious.