Inks traded his Etymotic Research HF5, in order to obtain this IEM, Rockit Sounds R-50, which popped out of nowhere. I've never heard about this manufacturer before, but it seems they reside close to San Francisco, California. I usually don't trust products made by random folks, but hey, I can still give some love to yet another Californian company.
Just like Brainwavz B2, R-50 is equipped with a pair of Knowles TWFK balanced armature drivers, tuned to be slightly warmer with less treble, according to the general public.
PRO: Considering TWFK drivers are used, R-50 is quite well balanced, retaining some resemblance to the diffuse-field reference; This is the first among TWFK-equipped IEMs. And the distortion has been further reduced at the frequency range, of which the human hearing is most sensitive.
CON: Just as expected, R-50's polarity is inverted. There is a harsh peak @ 10 kHz, but it is only audible when the IEM is fully inserted.
ON SECOND THOUGHT #1: Although stock ear sleeves yield a nice, deep insertion, it is still important to observe how different insertion depths help shape the frequency response of R-50 in the ear canal. At approximately 3~4 mm away from the reference plane, the IEM actually gains some amount of the high-frequency extension, while the problematic quarterwavelength resonances above 9 kHz are tamed. Practically, this is more applicable than going deeper, since not many users can insert their IEMs up to where the reference plane is located.
ON SECOND THOUGHT #2: Just like the other IEMs with TWFK drivers equipped, by adding a resistor in series, the mid-range gets boosted, while the treble level decreases. This under-damping scheme can further benefit from a shallow insertion, as shown on the right.
ON SECOND THOUGHT #3: By adding acoustic dampers to R-50's stem, users can further tweak the tonality. And of course, the tendency of acoustic damping is almost same as that of Brainwavz's B2. To my personal preference, "100 ohm resistor + gray dampers + shallow insertion" combination yields the optimal spectral balance.
I've analyzed 3 TWFK-based IEMs from different manufacturers, and Rockit Sounds R-50 is by far the best, due to much consideration put into circumstantial variables; A job well-done by Rockit Sounds.
Quick measurement links
Headphone measurement data by brands:
AKG • Apple • Astrotec • Audeze • Aurisonics • Beyerdynamic • BOSE • Brainwavz • Brookstone • Creative
Earsonics • Etymotic Research • Fostex • Final Audio Design • Grado • Heir Audio • Hidition • HiFiMAN • JAYS • JVC
LEAR • LG • Martin Logan • MEElectronics • Maxell • Monoprice • NuForce • NVX • Ocharaku • Phonak • Radius
RHA • Rockit Sounds • Samsung • Sennheiser • Shure • SONY • STAX • Suyama • TDK • T-PEOS • Ultimate Ears
Wall of Shame (coming soon)
innerfidelity has posted a bit different distortion graph. There is a peak at 5KHz where its not in yours. just curious :P
ReplyDeletehttp://www.innerfidelity.com/images/RockItSoundsR50.pdf
Maybe Tyll's pair was out of tolerance limit, or his THD+N calculation algorithm captures more distortion than my THD calculation. My THD algorithm is fairly simple and straight forward: http://i.imgur.com/7Lgx0Qw.jpg whereas THD+N counts the source's noise level as well: http://i.imgur.com/FxRxH0r.jpg
DeleteOtherwise, his setup clearly has a distortion issue, as his THD figure is definitely more exaggerated than what how it should've been.
I try to stay away from anything that can add the source's characteristic to the result, as some headphone amplifiers may have hard time handling the load's impedance fluctuation. My measurement setup was not only volt-metered, but also precisely calibrated using Etymotic Research's QC record. So I'd say my data is good as it can get:
Test condition: 100 dB SPL @ 1 kHz, by using the transducer with same serial #
Etymotic Research, by the courtesy of Mr.******
THD: 0.326 %
D2: 0.072 %
D3: 0.309 %
D4: 0.037 %
D5: 0.063 %
udauda's measurement setup
THD: 0.339 %
D2: 0.105 %
D3: 0.318 %
D4: 0.037 %
D5: 0.026 %
I see. So you would say only to directly compare results from the same source site not with other people's data correct?
DeleteWhen you compare distortion data from different websites, make sure to check the test condition and what kind of distortion has been measured. Thd at 1mw vs thd+n at 90 dBspl cant be compared against one another..
DeleteYes of course. I didnt realise that the one is THD and the other THD+n. I was basically looking the curve shape rather than the actual values. In most case the have similar curve shapes. For example your Phonak distortion curve is similar to the inner fidelity. This is something that was not in yours as a curve thats why i asked in the first place. Not for the actual values
DeleteYou know what, his phonak thd+n data also has a 5 kHz peak, which corresponds to his R-50 thd+n as well. Such an interesting find!
DeleteJust wanted to tell you that the "Measuring Headphone Total Harmonic Distortion" which would give an answer as to how Tyll performs the distortion measurements is coming soon back from 2011
Deletehttp://www.innerfidelity.com/headphone-measurement-procedures
So i dont think there is much hope we will ever learn the background on this! LOL
By polarity being inverted are you referring to the FK and WBFK being wired 180 degrees out of phase? Or is it the whole unit that's inverted?
ReplyDeleteIt seems the ground and output are soldered inversely.
DeleteWhat's so bad about that? As long 2 pieces are in phase (acoustically) your ears won't detect anything?
DeleteAn absolute polarity inversion is not much of a big deal for general listening, but is the matter of technical integrity.
DeleteHow does this compare with the ER-4?
ReplyDeleteAlso, how would you describe the difference between the 3 ER-4 models in terms of sound?
Thanks