CHANGSTAR: Audiophile Headphone Reviews and Early 90s Style BBS
Lobby => IEM Measurements => Topic started by: Marvey on August 04, 2012, 07:18:45 AM
-
UERM
ER4S
E2C
Interesting differences with distortion characteristics between driver types. Need more data. To be continued. Will post this on HF. Hopefully some retard won't barf on it.
http://www.head-fi.org/t/621303/differences-in-characteristics-of-driver-types-in-iems-some-interesting-measurements (http://www.head-fi.org/t/621303/differences-in-characteristics-of-driver-types-in-iems-some-interesting-measurements)
-
I've been wondering something for sometime that seems to concern measurements like HD, impulse response and square waves. People seem to judge them in absolute terms but to my ears, if I tried to imagine correlating my experience to visualized data, it seems to me that the overall behavior is just as significant. In other words, even though something appears to measure better absolutely, it may exhibit sporadic or odd/unpredictable behavior even if it performs better below a noise floor or within a specified time domain.
I'll bring over my IEM collection next week to keep you busy.
-
Very nice work, purrin. :)p5
-
3rd, 5th order (odd) harmonic distortion
This tells me the BA drivers produce waves a little bit more square. Maybe partially accounts for how they sound, a little bit faster with artificially higher slope rises.
-
What's the amplifier you've used for the test ? I don't see any obvious variation due to the cross-over in the UERM's graph, and I'm wondering what kind of influence there could be.
-
Just taught myself how to read the graph. Heh, so simple. Still can't fully correlate with other measurement data, such as square waves etc.
Very nice to see these though. Relatively high odd harmonic distortion. You mentioned that it in your experience results in atonality (pitches of notes less distinguished?). Do you feel that's the case with these IEM's?
Also, to teach a newb, how much of a difference is between 0,1% and 1% in dB?
-
What's the amplifier you've used for the test ?
You'd laugh if we told you. :)p17
-
Also, to teach a newb, how much of a difference is between 0,1% and 1% in dB?
20dB
10dB is considered to sound twice as loud, 20dB = 4 times as loud
-
Also, don't just think of distortion impacting music simply as a noise, crackle or his you just pick out and say 'Aha!'. Distortion can also contribute to an odd sense of discontinuity over a respective range in FR. The HE400 is a classic example of this. Many people won't notice the distortion in the mids, but if you know what to listen for, some vocals might sound oddly warbled depending on the track and particular product variant.
-
Is the effect of harmonic distortion which is impactful on sine-waves actually visible in square-waves? It seems to me like square-waves are so convoluted / directly paralleled to FR - which varies wildly in transducers - that actually seeing sine-wave differences in square-wave response is impossible since the universal shape of the square-wave is determined explicitly by the FR, and thus the square-wave response in transducers is unrelated to sine-wave response, which is only 'visible' in harmonic distortion measurements such as these if anywhere, is that actually the case?
-
One thing I've been wondering about: is there a correlation between ringing in a particular frequency, vs measured THD behaviour in that same spot?
-
No strong correlation I see. It does happen sometimes. For example, if there's ringing at 6k, we may see high 2nd harmonic distortion at 3k.
-
How come Sonoves' result is the distortion is cleanly under 0.001% from 2.5kHz to ∞?!
http://sonove.angry.jp/ER4P/ER4P_distortion.gif
http://www.sengpielaudio.com/calculator-thd.htm
-
D2 and D3 are way different, but the rest nearly match... Different levels?
Sonove measured at -20 dB but what is 0 dB?
The numbers from Rin are close, for ER-4B: http://rinchoi.blogspot.com/2012/03/etymotic-research-er-4b.html
And those ER-4S numbers are much closer to Purrin's - it's ER-4S though and you will have to subtract some to get -dB, then convert to %.
http://sonove.angry.jp/AKG_K3003/dist_ER4S.gif
-
Is the effect of harmonic distortion which is impactful on sine-waves actually visible in square-waves?
Square waves would mostly show odd order distortion, if at all. But frequency response would change their shape way more, so they're not really that useful, except maybe for subbass to show the slew rate equivalent of the device and bass sustain. You would see rounded for boosted slow bass, falling for poor sustain and ringy for artificially fast. Best would be of course level without ringing.
I'm talking e.g. 50 Hz square wave. Subbass behavior is not really visible on CSDs and dynamic behavior is not visible on the FR chart.