Separate names with a comma.
Exactly why I don't use Complys.
Just noticed Sennheiser Middle East is in the same building as my office.
Would it be strange if I knock on their door and ask for a loaner to test out the IE800S?
I didn't hear anything suggesting a 10 kHz peak, so we measured the IE800S today. Here's a preliminary frequency response measurement:
First impression: The Sennheiser IE800S improves on the IE800 with a greater sense of refinement -- the IE800S is less "fun" than the IE800 (in a good way), more precise (in a good way), with a more linear tonal balance. We're working on the Head-Fi Buying Guide update now, so I'll post a more thorough review (and probably more measurements) in there.
Here is the measurement system we used for the measurement above:
GRAS 45BB-12 KEMAR with anthropometric pinnae for low-noise earphone and headphone testing
Audio Precision APx555 Audio Analyzer
Rupert Neve Designs RNHP headphone amplifier (set to unity gain)
Herzan custom acoustic enclosure
This gives us good insight on what yo expect when it comes to the response frequency of ie800s!
Thank you, Jude
Thanks Jude....quite a contrast from the reported graphs on the original ie800.
any chance on (re) measuring the originals on your set-up?
But not so different from this one here.
Jude is really showing us that there are a lot of garbage measurements around. While I'm sure there is always further to go and he will refine his measurements with experience, already we can see that this is miles ahead of what's been posted.
If Jude still has his hands on some original IE800 is like to see some measurements on the same rigfrom that too.
When you wear a headphone and I wear that same headphone, our individual transfer functions can vary rather substantially, and these differences (independent of preference) can significantly impact our individual experience with a given headphone. (This is an interesting discussion that we'll be having more of soon.) That said, I think if the graphs above reasonably represent an average human experience of the IE800, then, in a sweep through that range, most of us should hear the drops both of those graphs show at 3 kHz (and thereabouts). In the first graph, relative to 1 kHz, the drop to 3 kHz is steep and approaches 14 or 15 decibels, with an even deeper drop through 5 kHz. In the second graph, it's slightly less steep (but still steep) and is around 12 or 13 decibels.
Put your IE800 on, and play the following video at a moderate volume level:
Do you hear a drop in level of 13 to 15 dBSPL as the sweep transitions from 1 kHz to 3 kHz? Some may, but I think most will not. I definitely do not.
Sure. Right now we're working on the Head-Fi Buying Guide, and there's a long way to go. I'll do that after, as we're running new measures of some of the new gear in the Guide (like the IE800S). That said, I'd expect our IE800 measurement will look a lot more like the IE800S plot I posted above than the IE800 plots you posted.
I just listened to this with my IE800S and I did hear the dips clearly I also heard well up to 19khz and then after it tailed off very very quietly.
The limit of human hearing is 20khz? so that sounds like it’s normal then.
The audio of that video has so much very apparent aliasing artifacts it's not really possible to use it for any purpose IMO.
Yes, and the artifacts are more apparent at higher frequencies -- it's a YouTube video, and it's obviously not as clean and pure as the tone generator from our audio analyzer. That said, you should still be able to use it for this purpose. Again, if you have an IE800, do you hear a big drop in level (13 to 15 dBSPL) as the sweep transitions from 1 kHz to 3 kHz? Some may, but I think most will not. Again, I definitely do not.
I'm quite sure the drop isn't that large on the original ie800.
When I'm EQ'ing it, I add 3-5 dB of 1-3 kHz or so, but not more
What Jude showed us is how the measurement looks in a properly calibrated professional measurement setup. Plus, as he also mentioned, the angle/insertion of these IEMs will vary between different users and will affect the sound perception, the same way how the depth and the angle of IEM insertion into a coupler can affect FR measurement. At the end of the day, I think FR measurements are there for a reference since it doesn't guarantee that everybody will hear it exactly as they "see" it on the graph. That's why it's always a good idea to get as many data points, including FR captures and actual sound descriptions as others hear it. Then, it's up to you to filter it out, to toss out those which use word salad like "resolving tonality"