Testing for SPEED in IEM's - please participate in my test! [speed-metal / visual-kei / symphonic-metal / instrumental / electric-guitar shredding]
Sep 26, 2010 at 3:02 AM Post #61 of 120


Quote:
mvw2 ->
 
yes I pretty much agree with everything in your last post, but I thought that texture and microdetail go hand in hand with speed, and that in most cases low speed will compromise
microdetail and that high speeds with compromise texture (becoming thin, cold), which is why I found your comment interesting that the CK10 has both exceptional speed and texture in the same package.

 
Not specifically no.  Texture is more the ability for the driver to create variations in the note.  If we think of a sine wave, we have a rise before the peak and a fall after.  A sine wave is smooth, but real audio is ragged and varied.  Texture is more so the ability to portray these variations.  In terms of a drum hit for example an earphone that doesn't have a high amount of texture will just have the hit.  You'll hear the whack or thump and that's it.  A well textured note will show you more.  It will portray the strike, resonation, and roll off of the sound.  An example from another thread I was discussing a cello.  An artist was playing a piece and I was discussing some differences between the DBA-02, RE-ZERO, and Custom 3 in how the information was portrayed.  With a low amount of texture, you would hear the note played.  With more texture, you could start to perceive the string resonate and cello body resonate.  With a lot of texture you could perceive how the bow stroke and released from the string, the resonation and decay of the string vibration, and the resonation across the cello body.  The cello was played in a medium sized room in front of a small audience.  You could gain a sense of how the cello filled the space and the echo in the room.  More texture and micro detail brings out these aspects.  Now speed alone doesn't provide this.  An earphone can be fast and present very little information.  Part of the "BA sound" is the problem of not adequately presenting the variations during the rise and fall of the notes.  The notes in many cases are too short.  The rise and fall happen fast enough where only a small amount of information can be portrayed.  The BA driver can be super fast, but that doesn't mean it will articulate all the information in the audio track.  The SE530 is an example of this.  The note decay is very fast on that earphone.  Notes are very short, very thin, but the drivers are very fast, very clean.  The presentation ends up being a bit hollow/ghostly in a way because the thickness and texture is sucked out of the presentation.  Notes in a way lack the length of time to present a high amount of information.  A slow driver can have articuation.  The MTPG for example is a slightly suggish driver.  The treble is smoothed a good bit and the earphone gets somewhat blurred with complex pieces.  However, it has a thick note with texture.  The note is drawn out enough to allow time to convey those variations and bring through more info, but the driver isn't fast at all.  The CK10 is a unique earphone in that it mixes both high speed and an ability to draw out the note.  BA based earphones like the ER4, Custom 3, and UM3X are like this too where they are BA based but not so much of a "BA sound."
 
What does speed offer then?  Well speed alone is a good measure of how clean or blurred an earphone will be when presented with complex information.  A fast earphone will be able to effortlessly present a lot of information, separately, where everything heard is unique in source.  It's where you can have 10 instruments or 10 singers all going at once and you can pick out each one of them separately.  A lot of earphones aren't good enough to do this.  Part of the need for this is not just speed.  To actually separate out all the sounds and create a self "entity" of each in perception, the earphone needs to be good about a lot of things including texture and dynamic range.  A big part of this is the ability to portray a well sized sound stage that isn't just big but also spacious with both the ability to create proximity and distance.  The ability to create a multi-dimensional perceived space lets the sounds sit in individual spots, separated and unique.  Speed alone doesn't create this but speed does keep this clean and not blurred.  Dyanmic range and articulation/texture create the space, and these along with speed allow an earphone to both bring out and separate a vast amount of information all at once.  One of the best earphones I've used to day for doing this is the UM3X although it's quite forward in presentation (puts the singer right in front of your nose forward, claustrophobic to some).  Earphones like the ER4 and Custom 3 are pretty close.  The Custom 3 is an exceptional example of being able to create a well spaced and linearly spaced sound stage and it does it so well because it has dynamics and texture.  It's speed isn't the best but being a BA it's quite good even though I'd call it a slightly muddy earphone by BA standards.  Then again, I'd call the ER4 muddy too, but relative to how bad dynamics can be, this doesn't really mean much.  The muddiest BA is darn clean and detailed by dynamic standards.
 
So how important is speed?  Well, for me I want enough speed to not have the information presented in a muddy manner.  I want the earphone to be fast enough to deal with complex info and not noticably blur information together.  A lot of the better earphones do this well.  The MTPG might be on the low end of what I might accept, but it's noticeable enough to bug me.  A BA based earphone like the Custom 3 is fast enough not to notice unless you directly compare against a faster earphone like the CK10 that can show you the little details that get blended together.  From blind use, it's really not enough to notice via normal use.  I personally tend to focus more on other things, mainly texture because it so strongly impacts the amount of information presented as well as providing the necessary information to create a perceptible stage space.  There is a lot of information lost when there isn't texture and good articulation of the note.  Second is dyanmics.  I seek both the ability to portray quiet information quietly and loud information with a high amount of energy.  It's oddly hard to find earphones that can go to both extremes well.  Dynamic range helps define articuations better which helps with the sound stage as well as emotion and energy in the music.  For me, speed comes after these things because it has less of an effect.  Speed just keeps the info presented from getting blended together and incoherent.    There will also be some limitation to how much detail can be presented, especially with higher frequency information.  If the driver is too slow, top end detail tends to smooth or blur some and you lose some edge, sparkle, and some of that bite on the top end.
 
In the end, I think it's all about showing as much of the original source information as possible as well as presenting that information as balanced as possible.  There are just a lot of factors other than just speed that get you this.
 
Sep 26, 2010 at 6:32 AM Post #62 of 120
Quote:
Hello Joe the Spider!
 
LOL at your negative grades on the re-zero! that is just what they needed, but could you please check the live recording as well?  the studio recording is too stuffy:
 
http://www.youtube.com/watch?v=cUQi3hV3eM0&fmt=18&t=2m50s
 
 
Jubliee in FLAC as your main critical listening thingy?  That is great!
 
 
p.s. any other earphones you have to review?  will be appreciated!


Hm, I listened to the live version. Better generally, much wider presentation and better separation.
 
For that track, stats for the ZERO are:
Speed - 7 | Again it's not that the ZERO is bad for a dynamic, it's just that compared to BA IEMs it just can't match up. I may be overcompensating in lowering my score since I haven't heard a BA in quite a while.
Separation - 4 | The separation was a lot clearer here. I sometimes think that the ZERO is not dissimilar to having a diminuitive version of the artist in your head - all the detail is there and you can hear where each of the sounds is coming from fairly effecively, but the overall presentation is not quite full enough to reach a realistic timbre.
Reverb - 2.5 | Again here the stage felt wider and airier, pretty good by IEM standards but nowhere near some full size open cans.
 
Now I want to be clear here: the RE-ZERO is an excellent IEM. It's fantastic value, has crisp treble, milky smooth mids and a present but not overshadowing bass, and also impressive separation and soundstage. However, for what you're looking for it seems like a BA phone would be much better suited to your desired sound.
 
Me, I'm aiming to upgrade to a CK10[0]
tongue.gif

 
Sep 26, 2010 at 10:23 AM Post #63 of 120
Are they all guys? I swear it's so hard to tell with the red chick.
 
I tried the track with the CK100 and found it not much different from any other music. I also tried it with the MEe M6, which relative to the CK100, I would score:
Speed: 4
Separation: 2
Reverb: How do I describe this? It's like your 'horizon' sound, but then everything just gets annihilated by a black hole after a few metres. Up to that point, it's ok.
 
Sep 26, 2010 at 10:35 AM Post #64 of 120
ck10
speed 11/11.. i have yet to hear faster iem than this.
is 5/5..
um3x might do slightly better since im able to easily pick up details. but probably just the sound frequency balance that made small highs clearer since mids is recessed, but i think eq'd both can be the same.
 
radius twf11r ddm
speed 9/11
is 4/5..
well this is unburned.. maybe will improve a lot later..
 
crossroad woody2 (1.7k hr)
speed 7/11
is 3/5.. some dynamics, can have quite a good separation after long burn-in.
 
the rest below  is from memory, last one sold was dba-02, somewhat valid;
 
dba-02
speed 9/11.. definitely not as fast as ck10.
is 5/5..
i found my ck10 reveals more than dba-02, hence lesser score here.. others (a lot) may disagree.
 
below are just rough estimates... according to my experience owning them for awhile.
 
hippo vb
speed 7/11
is 3/5
 
sherwood se777
speed 8/11
is 3/5
 
fischer audio eterna
speed 5/11
is 2/5
 
fischer audio enigma
speed 4/11 too much booms... slows down everything.
is 2/5
 
hippo white
speed 6/11
is 2/5
 
hippo groove
speed 7/11
is 2/5
 
shroom EB
speed 8/11
is 3/5
 
i dont really understand reverb..
but if u are talking about huge space in the soundstage, ck10 would be 6/5, dba-02 4/5.
i have no problem going from my amped ATH W10VTG to my (unamped) ck10, but never to the dba-02 (even amped). Currently using Lovely Cube Linear with class-A biased LME49820NA opamp, which imho pushes my vtg very close in comparison to Yamamoto Ha-02.
since ur using sa-5k (i have auditioned it before along with the yamamoto, and compared it to my vtg), dba-02 might be more similar to ur preference, along with ck10, re0.
 
Sep 26, 2010 at 11:50 AM Post #65 of 120
Quote:
Are they all guys? I swear it's so hard to tell with the red chick.


Unfortunately. Hizaki (the red 'chick') has a habit of breaking my mind whenever I look at her/him/it.
 
Sep 26, 2010 at 2:37 PM Post #66 of 120
Sorry to offshoot from the core idea of this thread.  I just feel speed is one, small aspect of the whole picture.  In reality, there are a lot of aspects that come together to create a full, coherent presentation.  Speed is merely one part of this and in some ways not even a dominant part for what you're looking for.  For example, dynamic range and articulation of note are just as important as the speed of the driver (which will typically be quite fast relative to the very slow, low frequency bass notes).  Typically speed only becomes important in higher frequencies where the driver has to be fast enough to keep up and articulate very short bits of information.  Bass in general is very long and sluggish in comparison.  I mean a 100Hz bass note is 100 times longer than a 10,000Hz treble note.  If the driver can produce 10,000Hz sounds, it would have no problem producing a 100Hz sound or even several quick beats of that note.  Comparatively all of the bass stuff will be quite slow.  The problems arise from the driver not being able to flesh out the bass note.  Many SQ earphones focus moderately on mid and treble frequencies and end up being a little thin in note, a little dynamic constrained, a little rolled off in sensitivity, or any mix of these.  This is where a lot of the problems arise.  Some driver simply lack excursion or enough control to produce or keep clean low frequency notes, and information can get muddy.  It's not so much driver speed as it is a design limitation for lack of excursion or control of the driver at higher excursion levels.  The Denon C751 is a good example.  It has a very linear, extended bass response.  It's very dynamic, articulate, and pretty quick.  However the driver is very loosely controlled and under higher volume levels, the design lacks a lot of control and bass gets very floppy, muddy, and incoherent.  At lower volumes it's good though.
 
There are just a lot of factors involved.  I want people to understand this very well.
 
 
Staying more on topic...  Views are from a holistic standpoint, not just bass.  Speed is the relative speed of the driver through the entire frequency range.  Separation is just that, instrument separation, the ability to put sounds in their own, distinct places.  Reverb is largely the thickness of the note, the ability to draw out and fill out the presence, mainly in lower frequencies.  These properties alone really tell very little about how the earphone sounds and presents information.  Even the rankings themselves are subjective and based on a range of criteria that the review decides.  I look at the big picture and have to break down a lot of variations into a single value that represent a slew of aspects that aren't even discussed.  Take the numbers as you will.
 
Custom 3
Speed - 8.5
Separation- 9
Reverb - 8
 
DBA-02
Speed - 9.5
Separation - 8
Reverb - 3
 
RE-ZERO
Speed -  8
Separation - 9
Reverb - 7
 
UM3X
Speed - 10
Separation - 10
Reverb - 9
 
SE530
Speed - 10
Separation - 10
Reverb - 6
 
IE8
Speed - 7
Separation - 6
Reverb - 9
 
Triple.Fi 10
Speed - 8
Separation - 9
Reverb - 9
 
CK10
Speed - 10
Separation - 10
Reverb - 7
 
Ok1
Speed - 9
Separation - 9
Reverb - 8
 
RE252
Speed - 9
Separation - 8
Reverb - 7
 
MTPG
Speed - 6
Separation - 8
Reverb - 8
 
Sep 27, 2010 at 1:18 AM Post #67 of 120
mvw2
I do really like all of your posts
smily_headphones1.gif

 
There are several parameters which will make diffrences in audio playback. There are transport , DAC, and earphones/speakers. Assume that we have an excellent transport; good source and decoder, and really good pair of headphones/earphones. How about the DAC or AMP ? These are also very important part to get most of the information or details from the source.  My Sony Walkman NWZ-X1060 sound differently compared to my iBasso D10, epsecially in the area of  RESOLUTION. In this test, can we assume that SPEED is closely related to RESOLUTION? 
 
If I use different PMPs. DAC. amp,... the scores might be change too.
smily_headphones1.gif

 
So, in this case,  earphones have very important role to deliver the sound information to our ears, BUT, DAC or/and AMP also need to supply a better signal too.  
Imagine how the subwoofer sound like if you use low damping factor or 'weak' amp.
 
Sep 27, 2010 at 1:40 AM Post #68 of 120
This is very cool as it gives a framework for rating other IEMs for those who've heard most or all of these (such as myself 
tongue.gif
). I've not heard the SE530 but the rest look about right and I can use this to say something about a couple of others in relation to the ones mvw2 posted (and using his terminology).
 
Quote:
 
Custom 3
Speed - 8.5
Separation- 9
Reverb - 8
 
DBA-02
Speed - 9.5
Separation - 8
Reverb - 3
 
RE-ZERO
Speed -  8
Separation - 9
Reverb - 7
 
UM3X
Speed - 10
Separation - 10
Reverb - 9
 
SE530
Speed - 10
Separation - 10
Reverb - 6
 
IE8
Speed - 7
Separation - 6
Reverb - 9
 
Triple.Fi 10
Speed - 8
Separation - 9
Reverb - 9
 
CK10
Speed - 10
Separation - 10
Reverb - 7
 
Ok1
Speed - 9
Separation - 9
Reverb - 8
 
RE252
Speed - 9
Separation - 8
Reverb - 7
 
MTPG
Speed - 6
Separation - 8
Reverb - 8


CK100
Speed - 9.5
Separation - 9.5
Reverb - 7
 
W3
Speed - 8.5
Separation - 9
Reverb - 9.5
 
W2
Speed - 9
Separation - 8.5
Reverb - 8
 
SM3
Speed - 10
Separation - 10
Reverb - 9
 
MTPC
Speed - 6
Separation - 8.5
Reverb - 7.5
 
Monster MD
Speed - 5.5
Separation - 8.5
Reverb - 8.5
 
Sep 27, 2010 at 2:51 AM Post #69 of 120
I don't relate speed and resolution.  High speed will typically guarantee you one thing: not getting a muddy sound.  Now speed can help create, edge, bite, aggressiveness but not always.  For example, the DBA-02 is very fast and aggressive.  The SE530 is very fast and has one of the highest dynamic ranges I've used but is often considered a bit smooth, laid back (due to lack of body, note weight).  Speed alone has little to do with the end sound outside of cleanliness of the note and good separation of complex information.  Speed alone does not guarantee an abundance of information nor create high or even moderate resolution. 
 
I see resolution more so in the dynamic range (ability to portray quiet up to loud) and articulation of note (ability to create texture and small variations during the presentation of the notes).  This is where a LOT of the information lies, and if you want to hear more in the music, you need an earphone that can do both well.  I will happily use a slow and muddy driver if it's at least dynamic and articulate.  The MTPG is a good example.  It's arguably a sluggish earphone.  It doesn't do well with complex info and is slow enough to be quite smoothed off on the high end.  However it's decent about dynamic range and articulation, so it actually presents a lot of meaningful information.  I have come to very strongly seek an articulate earphone, one that can create a high amount of texture/variation in the notes.  I have come to also strongly seek earphones that can portray both subtlety (actually kind of rare) and explosiveness (also somewhat rare).  I don't simply mean a quiet or loud earphone.  I mean a range that it can play.  I see dynamics like frequency response.  It's the ability to output very quiet information quietly (many either present none of this or present it too loud) and at the same time the ability to quite literally be explosive (a raw sense of energy, force, presence that comes across almost overwhelming).  A compressed dynamic range is like having an earphone that can play only from 100Hz to 10kHz or only 200Hz to 8kHz.  The extremes are either lost or compressed into the playable spectrum.  Dynamics are the same way.  It's hard to get both extremes in the same earphone.  I look at both articulation and dynamics not only because they help portray more information, but this is also how the sound stage is created.  It's in the texture and dynamic breadth that the sense of space, sparation, location, size, everything is created.  To lack articulation and dynamic range is to lack a sound stage.  It simply won't be produced well or hardly at all.
 
In the end, do I care about speed?  Sure.  I think enough is needed to keep the sound clean.  Enough is needed to let the earphone reproduce complex tracks without blurring stuff together.  It is a need.  However, I value articuation and dynamics more.
 
Why does the CK10 seem to have everything?  Because it does.  I see speed, dynamic range, and articulation as mutually exclusive traits.  They never have to come together.  You can have any one without much of the others.  In the end you want a lot of all 3.  The CK10 is one earphone that has all 3.  I'm a fan of some other earphones because they perform at least a couple of these traits well and isn't too shabby on the 3rd.  The CK10 isn't even my favorite earphone, but from a technical standpoint, it does the most right out of everything I've used.  I very much like a lot of other earphones like the UM3X, RE252, Triple.Fi 10, Custom 3, ER4S, and C751 because they do very well in what I seek and have minimal faults (forgivable/livable faults/limitations).
 
Sep 27, 2010 at 3:52 AM Post #70 of 120
Don't get mad at me, but I'm going to do this on 2 full sized headphones (Audio-Technica ATH-W1000X, Senn HD650) and 1 IEM (Etymotic ER4S.. I don't have my Senn IE8's on me at the momment)
 
W1000X
 
9/11 on speed, I can hear pretty much every individual note played, with the small "air" in between the notes that you mentioned with your little graph. =P
4/ 5 on instrument-seperation. I can differentiate between each instrument aside from "3:27 - getting pretty chaotic here, can you hear the drummer and all three guitarists? I can't..." where I can only hear 2 guitars.
4/5 on reverb
 
HD650
 

5-6/11 on speed, The drum beats on the faster passages seem to start melting together a bit. Same with the guitar that kicks in around 4:11 on the non-live version.
4/ 5 on instrument-seperation. Seems pretty good to me, I can place each instrument in its on section, and can tell each one apart, but again, I only hear 2 guitars at the 3:27 mark.
Not quite sure what to give for the Reverb section of this. It's more muffled and recessed than the W1000X. It's simply not as engaging and exciting to listen to, not sure which located you listed would place where I'm at with these phones. 
 
I'm getting tired and didn't get around to listening to both links with the ER4S. 
 
Sep 27, 2010 at 6:43 AM Post #71 of 120
Update:
 
Thankyou for all your excellent reviews and expert opinions.  Thanks go out to ljokerl, mvw2, JoeTheArachnid, daemontentacle, trysaeder, panges, bakhtiar, sushibowl, rawrster, stevenswall, kensuguro, anaxilus, tehort, rakfunk12.
 
I'm currently in a third-world country in asia and my visa is about to expire, but I spent most of today lying on the couch listening to versailles.  I have to stop being lazy and get to work, so I'll keep this post brief.
 
It seems as though this test is coming along nicely, and evolving.
 
I have concluded I should go back to page 1 and delete the non-live version, the live version isn't only a better recording, I think it's a better performance, it's usually the other way around, but not this time, and I'd like to have a single test track to make this test more concise and indicative "x sound at 4:06" etc., the studio non-live version can remain as a reference.
 
I should also note that when testing, focusing on the lead guitar (as I have done in my notes) is only part of the challenge, I've realised the drums are even more challenging, as well as the second guitarist (the one playing on the left side), so listen closely to the drums and second guitarist, (and the bass guitarist as well).  Also note the piano and string ensemble sounds are fake (played via on-stage speakers at the concert) so they are irrelevant.
 
Rating in either 10/10/10 or 11/5/5 is fine, I'll just adjust the ratings accordingly when I compile all the results into a single list later on (10 / 7 / 7 = 11 / 3.5 / 3.5) or vice-versa.
 
 
 
 
Sep 27, 2010 at 1:56 PM Post #72 of 120


Quote:
 
I see resolution more so in the dynamic range (ability to portray quiet up to loud) and articulation of note (ability to create texture and small variations during the presentation of the notes).  This is where a LOT of the information lies, and if you want to hear more in the music, you need an earphone that can do both well.  I will happily use a slow and muddy driver if it's at least dynamic and articulate.  The MTPG is a good example.  It's arguably a sluggish earphone.  It doesn't do well with complex info and is slow enough to be quite smoothed off on the high end.  However it's decent about dynamic range and articulation, so it actually presents a lot of meaningful information.  I have come to very strongly seek an articulate earphone, one that can create a high amount of texture/variation in the notes.  I have come to also strongly seek earphones that can portray both subtlety (actually kind of rare) and explosiveness (also somewhat rare).  


Totally agree w/ this.  My MDs are very similar to the Golds in speed if not identical I suspect.  But being slow does provide that quality of articulation.  Sounds like I need to hear a CK10 someday.  The treble scares me though.
 
Sep 27, 2010 at 2:16 PM Post #73 of 120
First of all, I think this thread is awesome and I love the idea of testing a bunch of cans / iems at the same time.  The following may sound a bit harsh, but I want to give my 2 cents as a production person..
 
imho I think this test should be taken with a grain of salt..  the results are very subjective.  I agree that some parts of audio cognition is subjective, but as a "test", I think it needs to be a little more methodical and use objective measures and units people can understand..  I'm not sure of the audiophile lingo, but terms like "speed" or "instrument separation" is rather vague and ambiguous.  At least for me, as a mixer and producer, I have no idea what they mean and so those measures are not very useful..  they're just arbitrary units that have no fixed value.
 
Basically what is happening is that the drivers have non-linear response behavior across the spectrum.  This may be due to amplification, driver mechanics, or the canal design of the IEM, (amongst possibilities).  A meaningful way to describe a pair of IEM would be to list its base EQ bias, and its response time overall, and then the response characteristic per freq band. (low, mid, mid-high, 3k-4k, etc)  Sound stage is the phase alignment of the signal, and if that is exaggerated or compressed, then the iem is screwed.  The comparison should be against production gear such as studio monitors or reference cans.
 
I mean, the problem is that I have no idea if the reviewer is able to detect eq curve differences of less than a db or not, or whether the person can discern the difference between compression characteristics.  Like the guitar solo for ex..  is the person not able to pick up all the notes because they are not musically trained?  Or is it really because of the iem?  How many times has the person heard the material before?  How much was the person concentrating on the particular part when listening?  Did he have listening fatigue?
 
The audio measuring units are there, and there are many, many production terms that can be used to precisely describe a signal.  I recommend they be used for reviews so they can become a more objective measure.
 
Sep 27, 2010 at 4:41 PM Post #74 of 120
Unfortunately this hobby is mostly subjective.  Outside of a small amount of real testing, no one really has the money/equipment ot put a lot of these products into real measures.  Headroom is one of the only places that have taken a decent amount of effort to actually measure some of these earphones.  I've run across a couple other sites that have measured data like www.Goldenears.net and sonove.angry.jp/ but overall info is more limited than Headroom.  I'm with you kensuguro in the sense that data should be objective with real, measurable values.  Data is absolute.  Yet, we are mostly stuck with subjective and biased opinion.  For most of us, that's the best we can do.
 
Sep 27, 2010 at 5:18 PM Post #75 of 120
I agree with..  I'd just add that perhaps data isn't limited to measurements from equipment..  like most of the production community has no problem describing outboard gear or software plugins with words.  But there are specific terms, which guessing from your previous posts, you may be familiar with..  like an non aliasing oscillator (no mirroring at the nyquist), a variable knee compressor, fixed ratio compressor, a phasy eq, studio monitors that brings up vocal range too close, 200hz rumble that's 3db too much, etc..  these may sound like ambiguous terms, but they point to very exact things.  Also mixers can pin point sound to a very exact range and volume..  which is necessary for work.  It would definitely help to be able to do that when rating things.  Sounds like 60hz or 100?  Is it 3k or 6k?  Is it 1db difference or 0.1db?  I think the problem is the vocabulary.  Most of sound is a subjective experience, but it can be described using objective terms.
 

Users who are viewing this thread

Back
Top