New custom IEM company - Custom Art: discussion, review, and impressions thread
Jul 13, 2014 at 12:23 PM Post #1,277 of 4,065
Jul 13, 2014 at 12:27 PM Post #1,278 of 4,065
  Basically yes, didn't feel the magic of 10 drivers..

But, it's 10 drivers...  Something funny going on here.
 
images
 
 
Kidding, did you have a review on it, or what did you think was lacking?
 
Jul 13, 2014 at 2:20 PM Post #1,281 of 4,065
Those UM supplied graphs are raw (no compensation) and, I think, fairly hastily done as a final check to make sure it's fairly accurate to the final sig. I've seen lots of slight variation from user to user on miracle and merlin customs. Probably a combination of each ciem being different and a hurried final test to get them out the door.

Edit: so there will be a peak there but less dramatic in actuality, at least in my experience with the Merlin.
 
Jul 13, 2014 at 2:21 PM Post #1,282 of 4,065
Those UM supplied graphs are raw (no compensation) and, I think, fairly hastily done as a final check to make sure it's fairly accurate to the final sig. I've seen lots of slight variation from user to user on miracle and merlin customs. Probably a combination of each ciem being different and a hurried final test to get them out the door.

That's what I was thinking, probably just raw with no compensation.  I'm predicting V shape if the compensation needs to be applied.
 
Jul 13, 2014 at 3:43 PM Post #1,284 of 4,065
Supposedly the miracle is a very shallow V, so mentor should be slightly more in V shape.. I assume.

Forget it, I think every one of their graphs are raw.  All their graphs have much peakier treble, even more so then the mentor.  This is quite interesting because my 1964 V3's graphs is more like Tyll's graphs where it looks like a semi-circle and the high end looks to gradually roll off.  I guess every company uses have their own compensation methods.  I'm curious which is the more accurate one?
 
Jul 13, 2014 at 5:51 PM Post #1,285 of 4,065
UM's measurements are made with IEC711 coupler so they are pretty accurate. Variations mentioned by shotgunshane are mostly due to IEM placement in the coupler.
Those graphs are raw data, no compensation.
 
Jul 13, 2014 at 5:56 PM Post #1,286 of 4,065
Forget it, I think every one of their graphs are raw.  All their graphs have much peakier treble, even more so then the mentor.  This is quite interesting because my 1964 V3's graphs is more like Tyll's graphs where it looks like a semi-circle and the high end looks to gradually roll off.  I guess every company uses have their own compensation methods.  I'm curious which is the more accurate one?
I haven't seen those graphs, but Tyll provides both uncompensated and compensated. V3 is based on Sonion 2300 for tweeter which have different response than TWFK from Mentor
 
Jul 13, 2014 at 6:17 PM Post #1,287 of 4,065
Well, here are two V3 graphs.  The top one is project's, and the bottom is mine.  As you can see, the graphs looks different.  Project's looks to roll off a bit earlier than mine, and both of ours have barely any extension beyond 10k.   There is just slight dip in the high bass, low mids section, but the drop past 10k is most significant.  The charactersitic is quite smilar to the SE846, and I would say the treble does come close, but the 846 is tiny bit more boosted from what I recall.  The lack of treble extension beyond the 10k I believe creates detail loss.  Oh well.  I believe the Shure has Sonion in it also.  Could that be responsible for early roll-off?  Could it be possible their other models have better extension possibly by using better tweaters?
 

 

 
What Rin measured of the SE846 with B&K 4157 (IEC 60318-4) occluded ear simulator at the reference plane:

 
Jul 13, 2014 at 10:04 PM Post #1,289 of 4,065
UM's measurements are made with IEC711 coupler so they are pretty accurate. Variations mentioned by shotgunshane are mostly due to IEM placement in the coupler.
Those graphs are raw data, no compensation.

I'm not so sure about that. What UM uses is just a cheaper Chinese knockoff of the IEC711. 
 
  Well, here are two V3 graphs.  The top one is project's, and the bottom is mine.  As you can see, the graphs looks different.  Project's looks to roll off a bit earlier than mine, and both of ours have barely any extension beyond 10k.   There is just slight dip in the high bass, low mids section, but the drop past 10k is most significant.  The charactersitic is quite smilar to the SE846, and I would say the treble does come close, but the 846 is tiny bit more boosted from what I recall.  The lack of treble extension beyond the 10k I believe creates detail loss.  Oh well.  I believe the Shure has Sonion in it also.  Could that be responsible for early roll-off?  Could it be possible their other models have better extension possibly by using better tweaters?
 
 
 
 
 
What Rin measured of the SE846 with B&K 4157 (IEC 60318-4) occluded ear simulator at the reference plane:
 

All drivers used in the 846 are made by Sonion. That overly complicated low pass filter is CNC-ed in US then assembled together with the drivers in Sonion's factory in Vietnam. 
 
Early roll off or not depends on the kind of engineering given to said model. I think I posted a 2300 graph that can go beyond 16khz before dropping below nominal level as reply to you in another thread :p
 
Jul 13, 2014 at 10:25 PM Post #1,290 of 4,065
  I'm not so sure about that. What UM uses is just a cheaper Chinese knockoff of the IEC711. 
 
All drivers used in the 846 are made by Sonion. That overly complicated low pass filter is CNC-ed in US then assembled together with the drivers in Sonion's factory in Vietnam. 
 
Early roll off or not depends on the kind of engineering given to said model. I think I posted a 2300 graph that can go beyond 16khz before dropping below nominal level as reply to you in another thread :p

I don't recall it, I know you mentioned the ER4 extending further.  The thing is, the ER4 is one driver, and it's impedance characteristic looks is quite different from multi BA(below is typical impedance curve of one BA). I believe a single can be extend easier. I don't know how it's done on multiple like my NT6(V3 shows it's not easy). I believe those single drivers are Knowels.  Not sure about this, but I saw this with the ER4B, but a capcitor was added to the 100ohm, and it raised the treble.  This was a high pass filter, and possibly the filter only boosted certain frequency beyond, and maybe this practice can be applied to the single driver, but multi would be difficult because of the cross-overs.
 

 

Users who are viewing this thread

Back
Top