Originally Posted by Joe Bloggs
@tinyman How can you use subjective impressions to validate measurements when subjective impressions differ wildly between individuals and even between the same person on different listening sessions? Objective measurements can only be validated by objective fact finding of the existence or lack of confounding factors in the measurement, such as a bad seal, different plug depth, amp output impedance, etc. Going about it any other way and you'd end up concluding that the Earth is flat because you can't see the curvature where you stand.
That said, the main reason you disagree with Rin's measurements is probably that you are used to reading Tyll's graphs and trying to correlate your listening impressions with those. Rin's measurements turn out differently but if you look at the UE900 vs other phones he measured you'd probably see the same relative trends between them as you would with Tyll's graphs. They may not correlate with your relative impressions though because something as simple as tip selection, fit and plug depth changes the sonic differences between IEMs completely.
I do want to see if anyone hears the UE900 similar to Rin's measurements right now, so I'm waiting for people to speak up (more data, the better). I, in the past, have been able to agree to a couple of Rin's compensated findings (my IEM collection doesn't match up too well with what Rin's measured).
I hear them a lot like Tyll's simply because I don't hear that 9 dB dip in the 1k region. Even further, I hear the spike in the 2.5-2.7 kHz region higher than Rin measured it. I will agree there is a bump in the 500 Hz region (both graphs actually show this). As stated earlier, I'm not worried about the bass or treble on Rin's graphs, they look accurate enough for me.
Comparing to Rin's TF10, the depth of the V in the UE 900 is deeper than that on the TF10; something is wildly off as just about every subjective finding has said otherwise (I can't make this comparison myself). Remember, I'm reading the raws, not the compensated stuff.
When measurements don't match up to what people can hear themselves, it raises questions to the validity of the given measurements. This is why I opted to wait for another raw graph and compare it. Tyll's graphs were posted by Inks, which were much closer to what I heard compared to Rin's (specifically the midrange).
Originally Posted by goodvibes
I can't speak to the ue900 but the GR10 I sent Rin sounds a lot like they measured so I doubt it a setup issue. That set were the least top extended pair I've come across and the bass to mid balance was very much as measured. I also feel a somewhat descending top sounds more natural than flat out or peaky one on IEMs but that's a personal view. While other love them and I understand why, TWFKs haven't been my cup of tea. Don't know if that's what's used here but it sure looks like one or at least a derivative.
Many of Rin's raws, I agree with. The Phonak PFE 022 (greys) is a prime example of this. It's almost of a direct carbon copy of what I hear.
The setup definitely isn't the problem. I'm sure Rin measured the 900s multiple times and got similar results. I think the problem lies in the software used to measure, more specifically, the software in combination with the UE 900 together causing some sort of error.
Edited by tinyman392 - 10/22/12 at 9:09pm