This is yet another IEM received from Inks. Once again, thnx for giving me this opportunity, my friend. Please keep in mind that the listed reference response has been obtained with
gray filters attached.
PRO: RIGHT ON THE REFERENCE TARGET! Considering its MSRP of $129, the price / performance ratio PFE 022 potentially has is very high. You can even tweak the response with different filters too. Great value for the performance.
CON: There is a lot of harmonic distortion, especially the third harmonics, in the mid-range, at which the human hearing becomes the most sensitive. This is definitely audible, and seems due to the over-equalization done by the manufacturer to forcibly meet the target response. Well, I guess it is a reasonable compromise to take for the low price.
ON SECOND THOUGHT: Although PFE 022's accuracy is right on the target, its practical insertion depth is little too shallow due to the stem being too short. The owner, Inks, also notes that its sound is slightly harsh for him. The magenta-coded plot on the right shows the response with PFE 022 inserted shallow. No wonder it was harsh.
ON SECOND THOUGHT #2: The response of PFE 022 can be tamed, by using filters with various acoustic impedance value. If I was the owner, I'll just stick with gray filters, since they yield the least peaky response.
I think that those harmonics are derived from a characterictics of a BA transducer used in it. It can be found in the measurement result of Etymotic Research ER-4B, also.
ReplyDeletehttp://rinchoi.blogspot.kr/2012/03/etymotic-research-er-4b.html
:-)
Ah! I was wrong. I missed the value of THD in the resulf of ER-4B. ER-4B has a THD below 1%, but PFE 022 doen't.
DeleteDoes a serial resistor make any difference?
ReplyDeleteI had not built an impedance adapter back then, but you can expect a boost in the treble as source impedance rises. I hope I could do a comprehensive analysis.. just need to get a hand on PFE with all the filters.
DeleteWould the resistor also lower the distortion?
Deletealthough further analysis is required, I expect a slight, about 3~6dB decrease in distortion.
DeleteRin, I use a 32ohm output and find it improves the sound depth and brightness in a good way. I have a pfe112 with all filters you can measure...
DeleteHey Rin i reached a practical block. Could you please help me out?
ReplyDeleteI have a debate with a head-fier telling him that i am worried about Phonak's distortion as seen in your graph. He said that he hasnt noticed it. Your comment is that the 1.5% and 1.7% distortion is definitely audible as the human threshold is at 1%
Could you please tell me a way in order to assess such statement?
Would you think this site does a good job in this area?
http://www.klippel.de/listeningtest/lt/
If you have any other proposal please tell me. When you assess two earphones and the high distortion is a deal breaker next step is to see if it is actually audible. That would be preferable to other head-fiers in order to make better subjective reviews apart from the scientific part. I asked about this on the phonak thread but still no answer.
Thanks man
LOL, I won't take credit for claiming that harmonic distortion lower than 1% (-40dB) is negligible in acoustics: http://www.r-type.org/articles/art-143.htm
DeleteThe thing is, even-order harmonics are tend to be less annoying, or perhaps pleasing, than odd-order components. The link you provided is a good starting point, but please note that the test result is only applicable to your own specific setup & test condition.
Based on the link i gave previously i did a test with my CK7 11mm dynamic drivers, a rather old model from 2007. Anyway i tried really hard with this setup
DeleteResults for "full range speaker" playing "two-tone 70Hz 800Hz"
and the lowest i could get was -39db which translates to 1.12% but with a song, this in particular
Results for "full range speaker" playing "Music T. Chapman"
i got up to -27db which translates to 4.46%
and this song is particular selected as stated from the site
Property of the test signal:
It is a critical and realistic test.
The mix of low (bass) and high (vocal+guitar) frequencies causes a lot of intermodulation products.
4.46% for a realistic test is a lot. Am i doing something wrong? Is is something wrong with my hearing system?
Also i forgot to mention results from these tests have shown that there are people that reached -54db in the chapman song and -60+ for the two-tone
Deletewtf? how is this possible? could you try it yourself and see? although the category selection at the beginning of the test is for speakers not for earphones and maybe that is the reason someone can discern 0.2% THD !!
Ah, such test can easily be cheated on the users' end, or there was really an expert listener with his golden ears.. :P
DeleteNope, there is absolutely nothing wrong with your hearing system. Non-expert listeners do have hard time picking up distortion with musical samples, that is all. Your performance rating might improve by using a better pair of loudspeakers/headphones, but I doubt it would make much difference..
Although there is not much of scientific significance, you can get a general idea on the non-expert listeners' audibility threshold on harmonic distortion here: http://www.axiomaudio.com/distortion
Thanks again. The reason for asking all of these is the following question. Should this high distortion value be a deal breaker or not? thats what i am trying to find out. I got a lot of opinions between purchasing the BA200s and the phonaks 112 for the same price and i got favourable views for the phonaks 112. but this distortion thing puts me off. and i dont want to get phonak and realise i did a mistake :P
DeleteWhat would be your personal view between the two based only on SQ?
IMHO, BA200 pwnz PFE 112 in every aspects, period.
DeleteRegarding distortion, that may have been me. Using eq to flatten the pfe to the level of the er4s isn't as hard as one would think, and they sound fantastic. However, now that i've eq'd them to be almost identical to the er4s in response, I am starting to hear the effects the distortion is probably having. I don't really "hear" anything out of the ordinary when I listen to music, but when I eq them to the same response as the er4s, they don't sound as smooth. After a lot of eq experimentation, I think this is the distortion. I simply have reached the best possible quality with the pfe, and they simply don't get any clearer or smoother because of the distortion. So while it isn't really "audible" in the sense that you can easily "hear" it, I think I am starting to be able to hear the "effect" it has on the music. If I take this into account, then I would say yes, I was wrong before when I told you it isn't audible. It is very audible when comparing them to something like the er4s. A fast a/b switch shows the er4s is twice as smooth as the pfe with the same frequency response. So I've kept my pfe as a "quick grab" earphone that sounds awesome, but for any sort of serious listening I never want to use them because the er4s is so smooth with the same if not better overall quality. The er4s has comfort issues, but if you get used to them I've heard nothing better. The ba200 was pretty good though. I kind of wish I had kept them longer to try different tips and insertion depths, as I read the treble improves when deeper.
Delete