Hifiman IEM's: RE-400 and RE-600
May 15, 2013 at 2:24 AM Post #811 of 3,507
Quote:
I made some comparisons to RE-272 earlier in this thread. <snip>


Well, RE-400 are closest to RE-ZERO in tuning. What they did sacrifice for the (yet bit more than RE-ZERO) boosted bass is the highs extension of RE-272. However, they are less peaky than either RE-ZERO, RE0, RE-262 or RE-272. RE-272 indeed had a notable 4k peak and some midrange shout which could be annoying with some music. RE-ZERO had a smaller one at 3k. RE-400 is very linear (diffuse field) up to 5k if fitted deep.
If you're interested, udauda has measured them and all the other ones mentioned too. Note that this coupler has exaggerated 3k boost, it's the same GRAS Etymotic uses and we know how ER4S are "accurate" yet have audible 3k boost. The author knows about it. Also 5-6k cut is personal - some people (like me) need a dip there, some don't.
 
Whoever said GR07 or BA200 are superior to RE-400 hasn't listened to them. At all. I hope so, otherwise they're just plain wrong. It's not possible.
GR07 aren't even more linear or extended plus they sound slightly "plasticky" - and the shell tends to prevent deep fit; BA200 are more linear but have some midrange boost like Etys, are a bit bit more midbassy; both have some distortion - in GR07, annoying 6k+12k ringing; in BA200, noticeable odd order harmonics in midrange.
BA200 is tunable via insertion depth, but only for worse, while GR07 is almost not tunable because the shell prevents deeper insertion and only the peaks move around. You could mod BA200 (damper, impedance) to make them more neutral sounding, but they will still sound like a balanced armature IEM.
 
And yes, distortion even below 1% is audible. 1% is -40 dB. You need to go less than 0.1% (-60 dB) in the midrange, preferably much less. The low distortion gives a "character" to the sound - either "sharpened" if it's high order, especially odd, like in balanced armatures; or over smoothed if it's low order, like in some cheap dynamic drivers. In bass, odd order can give the sound of fake speed, while even order will again smooth it out, but make it sound "stronger".
 
The trick is, RE-400 drivers are highly tunable via the shell, much more so than RE-272. If you toss in major damping inside, you get a flat (as in non diffuse field) response that's yet more refined. Funny thing is that the shallow fit then brings them back to DF with less equalization. (with the large tips) And makes them actually airy, more than RE0 or RE272.
The bass can be controlled by reducing venting making them more balanced, which is what I'll try as the next mod.
 
 
RE-272 could be improved the same way, cutting the slight shout whiile trimming the 4k and smoothing out peaks.
I think they should revisit RE-272 drivers and maybe shells (except not in the mediocre plastics and smaller) and add the new coating to linearize them, while slightly modifying the tuning to keep the highs and maybe bump bass. That could be RE-600. I hope it will be.
 
May 15, 2013 at 3:08 AM Post #812 of 3,507
Rin doesnt use a GRAS nor a 2cc coupler, he uses a simulator similar to Tylls. The 3k boost is there because of compensation, it's where a big ear resonance resides so a peak is needed there to compensate, it's been proven many times. I do think that a flat diffuse field response are a bit too bright, but it works on a downslope. Thus ER4S 3k is a bit much but not because of the coupler, a W2 and W4 have a peak near there (df tuned)but end at much tamer results, more flat in that region ime because its a downsloped diffuse field

Saying BA200 is worse deeper is just plain wrong, but its a VERY special iem for tip selection. Er6-18 is the only one that works for me, in general tips have to be short and match its nozzle bore diameteter.

I do get the portion about the harmonics though, definitely there ime as well
 
May 15, 2013 at 5:02 AM Post #813 of 3,507
RE-400 owners, how is the build quality on the cable? I have the RE-0 and really enjoy the sound quality, and therefore am really interested in upgrading to the RE-400 (or 600...), but I just want to make sure the cable won't break quickly! I know the RE-0 is notorious for its build quality (or lack thereof) and as such has already started to break apart near the strain relief, but if a RE-400 owner could share how well theirs is holding up, that'd be swell!

I've only used my RE-0 a few handful of times and keep it hanging on my makeshift headphone "rack" when not in use, which has been pretty often this past month.
 
May 15, 2013 at 8:03 AM Post #814 of 3,507
Quote:
RE-400 owners, how is the build quality on the cable? I have the RE-0 and really enjoy the sound quality, and therefore am really interested in upgrading to the RE-400 (or 600...), but I just want to make sure the cable won't break quickly! I know the RE-0 is notorious for its build quality (or lack thereof) and as such has already started to break apart near the strain relief, but if a RE-400 owner could share how well theirs is holding up, that'd be swell!

I've only used my RE-0 a few handful of times and keep it hanging on my makeshift headphone "rack" when not in use, which has been pretty often this past month.

I've been using my RE 400 for one month and IMO you don't need to worry about the cable's build quality, it's slightly above average I think
biggrin.gif

 
May 15, 2013 at 9:27 AM Post #815 of 3,507
Quote:
And yes, distortion even below 1% is audible. 1% is -40 dB. You need to go less than 0.1% (-60 dB) in the midrange, preferably much less. The low distortion gives a "character" to the sound - either "sharpened" if it's high order, especially odd, like in balanced armatures; or over smoothed if it's low order, like in some cheap dynamic drivers. In bass, odd order can give the sound of fake speed, while even order will again smooth it out, but make it sound "stronger".

 
 
 
 
THD on the BA200 is 0.5%. have you checked the numbers? -46db hello?
 
i did the listening test
 
http://www.klippel.de/listeningtest/lt/
 
and couldnt go below -27db on the music track. hell i reached -39db on the test tones. thats less than -46db
 
maybe you have super hearing or something but this is personal. you cannot assert a subjective point as an objective one!
 
 
 
have you done an ABX test on your claims? or is it just a feeling?
 
 
 
Quote:
http://www.axiomaudio.com/distortion
 
The noise test tones had to reach 8,000 Hz and above before 1% distortion became audible, such is the masking effect of music

 
May 15, 2013 at 12:58 PM Post #817 of 3,507
Quote:
Originally Posted by FarCry /img/forum/go_quote.gif
 
THD on the BA200 is 0.5%. have you checked the numbers? -46db hello?
 
i did the listening test
 
http://www.klippel.de/listeningtest/lt/
 
and couldnt go below -27db on the music track. hell i reached -39db on the test tones. thats less than -46db
 
maybe you have super hearing or something but this is personal. you cannot assert a subjective point as an objective one!
 
have you done an ABX test on your claims? or is it just a feeling?

Unless you give me Silverlight for Linux that works (no, Moonlight doesn't), I'll have to skip that page for the time. Maybe later.
 
Sole 3rd order harmonic distortion is audible to me on some complex music at -30 dB. (synth or orchestra with lots of strings and trumpets) And that was an ABX test. Tones I can do to -60 dB, but that one is real hard. -40 dB is quite easy on the other hand.
 
However, single harmonic is much harder to hear than the mixture served by real drivers, especially balanced armatures which have relatively slowly decreasing distortion vs order. Remember that THD is a geometric sum, the scale is logarithmic and that the added harmonics intermodulate with the rest of the signal.
 
I can't do an ABX with real BA200 vs something else that sounds exactly like it. (I don't even own them - I had a few listens and they were quite ok.) Instead, I'd have to create the right harmonic distortion mixture. Rin's page doesn't have enough information - the chart is cut off. Not to mention creating the right filter to add it will be quite a task.
 
Bottom line is, TDK BA200 sounds like a midbassy balanced armature IEM. They do have this specific sound and but few sound different in this regard. It is a good IEM if you like that kind of sound and can deal with the quite high amount of midbass.
 
I'm quite sure the compensation used by both Tyll's, Rin's and Golden Ears is flawed or at least doesn't match (within 6 dB) what I hear, specifically in 3-7k range. (overemphasizes around 3k, is missing a 6k notch) Purrin's handcrafted one does nearly perfectly on the other hand.
 
--
The other quote is irrelevant - flat noise is (suprisingly?) harder to hear than harmonic distortion. Harmonics tend to change color of the sound, noise doesn't. Human hearing is geared towards rejecting random noise.
 
--
Re: Cables: mine broke near the plug this time, but it had witnessed some rough handling. (Rough enough that it broke the fabric sheath in one place.) They didn't harden at least.
 
May 15, 2013 at 4:29 PM Post #818 of 3,507
I know AstralStorm already commented on the non-hardening cables, but can anyone else share their experience with the build quality? My RE-ZERO is now dead (hardened cables / dead strain relief, channel imbalance), and I'm looking all over for a side/upgrade to the RE-ZERO with bigger soundstage and similar sound signature. Does the RE-400 have as sparkly treble as the RE-ZERO? 
 
May 15, 2013 at 5:40 PM Post #819 of 3,507
Quote:
Bottom line is, TDK BA200 sounds like a midbassy balanced armature IEM. They do have this specific sound and but few sound different in this regard. It is a good IEM if you like that kind of sound and can deal with the quite high amount of midbass.

 
 
 
yes i know about that. its all about the FR not the distortion!
 
 
http://1.bp.blogspot.com/-QjJAh8yPpmI/UVr_Y1NXP_I/AAAAAAAAFlU/hojfgGAaEY8/s1600/comp.png
 
 
and you said ABX at -30db
 
 
apart from that the third order distortion is 0.4% -48db where according to the textbook rin gave me privately (Springer Handbook of Acoustics) says that the audible threshold for third harmonics can be up to -50db
 
 
give me a break if you can tell the -2db difference. my 2 cents
 
May 15, 2013 at 10:42 PM Post #820 of 3,507
Interesting, different compensation curve? They definitely aren't linear in bass, what the heck? Although the higher midrange looks spot on like what I heard.
But yes, that's more or less how it sounds in comparison with Triple.fi 10. (assuming you can seal with those, they're way too fiddly with my ears)
 
Now then, I said it sounds like a BA and that I don't care for this kind of "sharpened" quality of sound. On the other hand, its bass is pretty clean in comparison to some other ones, e.g. TWFK.
RE-400 does not sound like anything, it's quite clean. Although I'd toss the damper and either use very deep fit, or take them apart, insert damping material e.g. cotton, also remove the filter and use the very short Hifiman large biflange.
Former will get you the smooth rolled off sound, latter gives a flatter presentation with way more air. (Do not use it without extra damping, or 6k notch gets real annoying.) Very similar to what RE272 was capable of and in fact with the same 4k bump, except with more bass.
 
If you switch to the long biflange (from RE272) instead with the damping, the sound will become yet more linear with a midrange 1 - 2k dip (3 dB) and slight 4k bump (2 dB), while improving slightly in the highest end extension. Yes, it becomes slightly V shaped, but the 6k peak as well as most others high end ones disappear completely.
 
If you add the filter back though, the sound becomes linear (bass peak and 4k peak are reduced to 1 dB, midrange dip becomes -2 dB centered at 1k) and keeps the superb extension. (The extension gain is substantial, from 16.5 to 18kHz.) Sensitivity drops, but bass becomes fuller.
I'm still working on the mod, but it's slightly on the back burner. If I manage to slightly decrease 1k dip and bump extreme subbass somehow, it would be nearly perfect for shallow fit. I probably need slightly more of the foam filter Hifiman is using.
 
May 15, 2013 at 11:04 PM Post #821 of 3,507
Interesting, different compensation curve? They definitely aren't linear in bass, what the heck? Although the higher midrange looks spot on like what I heard.


insertion.png


^Exactly as I heard it...

P.S. Posting a link to the original source may be of more use: https://www.google.com/url?sa=t&source=web&cd=1&ved=0CCoQFjAA&url=http%3A%2F%2Frinchoi.blogspot.com%2F2013%2F04%2Ftdk-ba200.html%3Fm%3D1&ei=fEyUUbiLAYrm8gTp8ICgCg&usg=AFQjCNHUBJ-nKQ39w5j-q5cGtiQvORE74Q
 
May 15, 2013 at 11:05 PM Post #822 of 3,507
Quote:
insertion.png


^Exactly as I heard it...

P.S. Posting a link to the original source may be of more use: https://www.google.com/url?sa=t&source=web&cd=1&ved=0CCoQFjAA&url=http%3A%2F%2Frinchoi.blogspot.com%2F2013%2F04%2Ftdk-ba200.html%3Fm%3D1&ei=fEyUUbiLAYrm8gTp8ICgCg&usg=AFQjCNHUBJ-nKQ39w5j-q5cGtiQvORE74Q

 
That's Rin's graph of the BA200 using different insertion depths...  Which insertion depth do you hear it as?  The graph is compensated.
 
May 15, 2013 at 11:10 PM Post #823 of 3,507
That's Rin's graph of the BA200 using different insertion depths...  Which insertion depth do you hear it as?  The graph is compensated.


With regards to bass, I'd describe my personal listening experience most in line with "the reference plane."

EDIT: answered my own question. The graph at the very end is merely the uncompensated FRC.
 
May 15, 2013 at 11:22 PM Post #824 of 3,507
You know the TF is kind of proof that 2-3k boost works, the fact that it lacks it is the main reason the TF10 is heard to have recessed mids. 
 
May 16, 2013 at 1:22 AM Post #825 of 3,507
if i were Fang , i would take you to work for Hifiman - you know your s**t 
tongue.gif

Quote:
Well, RE-400 are closest to RE-ZERO in tuning. What they did sacrifice for the (yet bit more than RE-ZERO) boosted bass is the highs extension of RE-272. However, they are less peaky than either RE-ZERO, RE0, RE-262 or RE-272. RE-272 indeed had a notable 4k peak and some midrange shout which could be annoying with some music. RE-ZERO had a smaller one at 3k. RE-400 is very linear (diffuse field) up to 5k if fitted deep.
If you're interested, udauda has measured them and all the other ones mentioned too. Note that this coupler has exaggerated 3k boost, it's the same GRAS Etymotic uses and we know how ER4S are "accurate" yet have audible 3k boost. The author knows about it. Also 5-6k cut is personal - some people (like me) need a dip there, some don't.
 
Whoever said GR07 or BA200 are superior to RE-400 hasn't listened to them. At all. I hope so, otherwise they're just plain wrong. It's not possible.
GR07 aren't even more linear or extended plus they sound slightly "plasticky" - and the shell tends to prevent deep fit; BA200 are more linear but have some midrange boost like Etys, are a bit bit more midbassy; both have some distortion - in GR07, annoying 6k+12k ringing; in BA200, noticeable odd order harmonics in midrange.
BA200 is tunable via insertion depth, but only for worse, while GR07 is almost not tunable because the shell prevents deeper insertion and only the peaks move around. You could mod BA200 (damper, impedance) to make them more neutral sounding, but they will still sound like a balanced armature IEM.
 
And yes, distortion even below 1% is audible. 1% is -40 dB. You need to go less than 0.1% (-60 dB) in the midrange, preferably much less. The low distortion gives a "character" to the sound - either "sharpened" if it's high order, especially odd, like in balanced armatures; or over smoothed if it's low order, like in some cheap dynamic drivers. In bass, odd order can give the sound of fake speed, while even order will again smooth it out, but make it sound "stronger".
 
The trick is, RE-400 drivers are highly tunable via the shell, much more so than RE-272. If you toss in major damping inside, you get a flat (as in non diffuse field) response that's yet more refined. Funny thing is that the shallow fit then brings them back to DF with less equalization. (with the large tips) And makes them actually airy, more than RE0 or RE272.
The bass can be controlled by reducing venting making them more balanced, which is what I'll try as the next mod.
 
 
RE-272 could be improved the same way, cutting the slight shout whiile trimming the 4k and smoothing out peaks.
I think they should revisit RE-272 drivers and maybe shells (except not in the mediocre plastics and smaller) and add the new coating to linearize them, while slightly modifying the tuning to keep the highs and maybe bump bass. That could be RE-600. I hope it will be.

 

Users who are viewing this thread

Back
Top