Campfire Audio Andromeda Review: Holy Cow, This is a Dream, Awesomesauce!

Discussion in 'IEMs and Portable Gear' started by purr1n, May 13, 2016.

  1. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I disagree with almost everything you've said.
    1. There a bit of stridency up top with the Andros, but IME, the location of the 8-9kHz peak changes depending upon the length of the tube / coupler used. It is a bit hot in the mid-treble (why I use the Comply tips), but the FR plots are misleading in this case because it's coupler related. Also note that Ken's plots don't show a sharp peak.
    2. Many people here feel that the "Diffuse Field" or Etymotic's proposed curve is a bit too pronounced. I myself am one of them. Another consideration is that many speaker builders will at times build in a 3kHz "BBC dip". In essence, we've decided that the DF or Ety curve is nonsense, a fad from several years ago, and are glad that people don't follow this anymore. Honestly, I couldn't give a rat's ass about authorities such as ISO or Ety, especially if they are wrong.
    3. Many people here prefer the Andromeda with higher output Z so the bass is less pronounced. The effects of this are shown in very early measurements I provided. This was emphasized post after post, and subjectively confirmed by other people. The point was roll your own bass boost with your source. I don't know how you missed this. I think out of everyone here, I only know one person who prefers the Andromeda from a 0 Zout source.
    4. My reference is speakers (I have three speaker listening rooms) using the B&K curve (at listening position) based on their AES paper from the 70s. The Andromeda is close to this with the exception of slightly hotter mid-treble. The Etymotic IEMs are nice too, but IMO, too lean with slightly too much energy in the 2-3-kHz area. The Etys sound nothing like any of my speaker projects. FWIW, I build most of my speakers and calibrate their FRs per above. I also evaluate FR based on a wide set of recordings, not all of them good, for the purpose of finding issues. Marginal recordings, some a tad too strident, too shrill, or too bassy, are useful this way.
     
    Last edited: May 16, 2017
  2. k.e.

    k.e. Winning combo: narcissistic and stupid

    Joined:
    Oct 29, 2016
    Likes Received:
    286
    Trophy Points:
    63
    Location:
    Germany
    Home Page:
    1. Yes they do. Well, they only show +10 dB at 9 kHz, but Campfire personally actually applauded and confirmed bartzky's measurements shown here.
    2. I don't know what you've decided or who "we" is, but actually many IEMs still follow that target, including the ones used for reference often.
    3. Ok. I was informed that I should make sure to use the lowest output possible. Not quite a consensus here then. Do know that 8-9 kHz are boosted the most with higher output impedance. What is the ideal OI then? When do you think sibilance is getting annoying?
    4. Etymotic don't perfectly hit their own target. The ER4 has slightly more midrange than their target aims for.
     
  3. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    "We" meaning established long time members here. I really don't give a shit that many IEMs makers follow that target. Even more DO NOT. I actually see a trend against it, mainly because people do want to hear lean shrill sounding shit.

    I don't know exactly who here informed you of such. There is 90%+ consensus HERE ON SBAF that Andromeda's work best with Z out from 2 to 5 ohms. Whether it is 2 or 5, that is a different story and depends upon personal preferences. Keep in mind that SBAF members do not give a crap about what HF or prominent or fake reviewers have to say.

    Please go to relevant measurements. Too lazy to provide you with links. We know EXACTLY how any transducer will behave in regards to FR and ZOut if we know its impedance curve. It comes down to Vout = Ziem / (Zout + Ziem) * Vin where Z = impedance (FR). No calculus, differential equations, or laplace transforms involved. Just addition, multiplication, and division.

    Pointless argument. Ety hits it close enough. Oh god, I would be worried if Ety hit those targets. It's be not only lean, but also lean and slightly bright.

    BTW, do you actually own or have even listened to the Andromeda? Or speakers calibrated to the B&K (or similar target curve used in mastering studios)? I could buy the argument that the Andromeda are slightly V shaped. But if you made such an argument, you'd also have to say that the Ety is slightly lean and shrill.
     
    Last edited: May 16, 2017
  4. k.e.

    k.e. Winning combo: narcissistic and stupid

    Joined:
    Oct 29, 2016
    Likes Received:
    286
    Trophy Points:
    63
    Location:
    Germany
    Home Page:
    I quote from my review as published:

    "I find the upper midrange to be slightly boosted... This tuning quickly unveils a little harshness in recordings... this adds a bit too much crunch for my taste as a consumer."

    "From my listening experience, the [ER4] SR have too little bass and noticeably roll off in the low-end. Bass information is simply missing at low volumes. I am absolutely sure that some will disagree with me, but I am not a person that likes to turn up the volume high... Mids can shout even a bit more than with the XR, making this a very mid-centric monitor. Treble is a tiny bit less splashy"

    Is this what you were looking for?
     
  5. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Yes. You are hereby awarded with this medal:

    [​IMG]
     
  6. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Looks like Marvey already ended the pointless measurement and FR debate.
    Anyway...

    I haven't heard those, but almost everything that measures somewhat close to the DF target sounds way too forward in the upper mids to me. Even the Apple BA IEM. Not what I would call natural.
    I like the UERM/DBA-02 midrange signature. The Andromeda was reasonably close, but for overall midrange FR I'd still take the UERM. I do prefer my modded HD800 over the UERM in terms of tonality. It's a bit more forward around 1-2kHz, probably slightly too much. My speakers I would say are the closest to neutral - slightly more laid-back. I basically listen to my speakers exclusively now (not a matter of tonality btw).

    Essentially (on most rigs) what I want to see is flat without compensation. The UERM falls into this category (on some couplers). I also kind of liked the STAX in-ear. With the innerfidelity measurements I look at the uncompensated measurements for in-ears.

    The thing is db/W (efficiency) is what I'm more concerned about than db/V (sensitivity), even though we obviously use voltage output. The Andromedas simply have a very low (and wildly varying) impedance that gives them a very high sensitivity, but the efficiency is not all that special. The specs might not be completely accurate, but it's probably similar to the UERM and B2.
    Sensitivity might be more meaningful in a daily use case, i.e. which headphone has to be turned up more, or which one emphasizes hiss the most (which is what you seemed to be talking about), but I have a feeling that efficiency at least has some correlation to sound (which I don't feel sensitivity does - otherwise the 32 Ohm Beyers would sound better than the 600 Ohm Beyers. When in reality it's probably the other way around - the 600 Ohm ones have a slightly higher efficiency and less resonances due to lower moving mass.)

    It might seem logical at first, but this is where I disagree based on my own experience. At least the measurements from three different speaker systems I took at the ear-canal opening looked nothing like headphones measured the same way. And (as you mentioned), despite the speakers measuring very similarly with an omni mic, they measured very differently on the head (differences up to 10db).
    One of the measurements (in the most reverbant room) showed 10db ear gain at 4kHz, while an HD600 measured the same way didn't even show a 5db bump at 3-4kHz. The HD600 already sounds slightly forward at 3-4kHz so a +5db EQ in this region doesn't make it more similar to speakers at all.

    Yes and this is another reason why I don't like the targets based on speakers measurements like the Harman curve.

    Here's an innerfidelity article that showed what I mean:
    http://www.innerfidelity.com/content/aes-headphone-technology-conference-highlight-paper

    "Sound-pressure Level Divergence (SLD) - Okay, if that was weird, this is weirder. It turns out that the mind perceives loudness differently depending on the nature of the sound field. Basically, for the same perceived loudness in free-field, diffuse field, and headphone listening, different levels will be measured in the ear canal."

    People seem to be comfortable looking at just FR graphs, but they're not even close to telling the whole story. I think the main reason why the speaker-based targets don't work for headphones is because the sound-field is so different. A speaker in a room will have 300ms or more reverb time, while a headphone has no reverb at all. With the speakers the first wavefront hitting the eardrum will have a very different FR from the diffuse room reverb. I think it'd be a better idea to use the first few ms as a target. Then again that is closer to the free-field target, which sounds like shit. Probably because the free field target is based on a perfectly flat measuring mono speaker playing right in front of the head.
     
    Last edited: May 16, 2017
  7. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX

    Easy to answer this. IEMs bypass pinna and also interact with the auditory canal.
     
  8. bartzky

    bartzky Acquaintance

    Joined:
    Dec 22, 2016
    Likes Received:
    62
    Trophy Points:
    18
    Location:
    Germany
    Home Page:
    An IEC 60318 coupler has the same acoustical transfer impedance as an average human ear up to 10 kHz. Hence every peak up to 10 kHz that occurs with this coupler will also occur in an average human ear. At least if the measurement was taken correctly, but I like to assume Ken did that.

    I do agree that a pure Diffuse Field (DF) is not a good target for headphones or IEMs, when not listening to binaural material. Mr Killion already considered the fact that "normal" music is mastered for loudspeakers and therefore will sound to bright with a DF-calibrated IEM. So he reduced 10 kHz by 5dB for his own Target. IMO his adjustment did not reach far enough. But just saying "they are wrong" is way to easy - Ety does at least has a very elaborated standpoint. I'd like to know why you think that they are all wrong and even more: what is correct in your opinion?

    Because IEMs bypass pinna there need to be some kind of pinna effect compensated into the FRs of IEMs. This is more an argument for some kind 3kHz bump than against.


    Everyone has got their own preferences and I totally respect that. But do you think a flat line on a coupler really is neutral from an objective point of view?

    Alright, I did not quite get what you wanted to say at first :D
    This might be true for single dynamic drivers. Regarding BA-based earphones the impedance and efficiency is also highly influenced by the crossover. Just think of a small resistor in parallel to a BA: the efficiency will drop and the impedance will too. Think of a resistor in series: the efficiency will drop too while the impedance will rise. It's still the same driver, but numbers for efficiency and impedance are totally different. I assume the sound quality will still be the same (the FR might differ a lot).

    Very interesting. Did you use a manikin or your own head and ears?
    The HD600 should show a 3.5kHz bump of about 10dB... at least it does that in Tyll's measurement.

    That's a good point. Wrong timing may be the thing we have to accept when listening to headphones as long as we don't DSP the hell out of it.

    Regarding different soundfields: IMO the directivity is an important aspect. Loudspeaker listening never is pure DF or FF. With low frequencies there will be a lot of reverb and therefore an indirect and diffuse soundfield. As the frequency rises the sound will become more direct so the free field will become more important. The FR of IEMs and headphones should take this into account and look more like DF in the lows with a rising amount of FF to the high frequencies. Mix that with established room curves and you've got the best target you can get, IMO. Though timing will ever be "wrong" (= not like speakers).
     
  9. gixxerwimp

    gixxerwimp Professional tricycle rider

    Pyrate
    Joined:
    Sep 30, 2015
    Likes Received:
    5,773
    Trophy Points:
    113
    Location:
    small island claimed by China
    I like my Andros. They make me feel happy when I listen to them |\/|
     
  10. GoodEnoughGear

    GoodEnoughGear Evil Dr. Shultz‎

    Pyrate
    Joined:
    Oct 25, 2015
    Likes Received:
    3,070
    Trophy Points:
    113
    Location:
    Cape Town, South Africa
    <noise>
    It's nice to see a relatively polite disagreement, for a change. Kudos to you both.
    </noise>
     
  11. Spl4tt

    Spl4tt New

    Joined:
    Jul 1, 2016
    Likes Received:
    13
    Trophy Points:
    8
    Location:
    Switzerland
    Whatever is going on here, I ordered my Andromeda yesterday. My UM Merlin broke, so I need something portable until they are repaired (Which will probably take a long time), and with all the praise the Andros get I'm excited to hear them.
    I hope it won't take too long until I have them, really looking forward to them
     
  12. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Hard to say. It's not just the coupler, it's the compensations too.

    Not against 3kHz bump. This discussion has been covered before with others, and noted in the comments with my own IEM measurements. See point #5 here: http://www.superbestaudiofriends.org/index.php?threads/note-on-iem-measurements-changelog.2153/ Just think Ety does it too much on the 3kHz. Way too much.

    On how normal music is mastered for loudspeakers and not DF IEMs: lolololol, this is the exact reason why DF is dumb and wrong. What is the point exactly of DF if it does not work for how all recordings are mastered? Heck, strict DF even sounds wrong on the binaural recordings I have. (Can - Flow Motion, Glenn Gould - Bach's Goldberg Variations reperformance, Art Tatum - Piano Starts Here - Live At The Shrine reperformance)?

    The reason I am believe that I am more correct is because of the references I use and my own experience. I've worked behind mixing consoles in a earlier life, I visit mastering studios related to my job), and I've built dozens of speakers which measure flat (1m) or similar to the B&K curve (at listening position). And also because of the measurements and work I've done in the past five years. Feel free to peruse the DIY speaker project discussions and measurements on CS and here. Also, IME, putting speakers in a very reverberant room (I have three speaker rooms, some more "live" than others) does not result in +20db at 2-3kHz, or even +10db. Try +2-4db at most, and why many speaker builders incorporate a BBC dip.

    This depends on the coupler and compensation used. The target curve for my own measurements is close the aforementioned B&K curve, but with a small bump 3kHz. Not a flat line.

    It comes down to compensation curves and target curves. You will never arrive at truth as long as you don't get this. And you can throw it all out the window if you don't have well trained ears and have calibrated loudspeakers. You have much to learn padawan.

    Getting back on topic and bottom line: the Andromeda from a 2-3 Zout source is very close to neutral loudspeakers (just a bit hot in the mid-treble). For reference, the UERM (leaner and with slightly more pronounced 3kHz, with DF tuning approaching, not still not quite to extent of Etys) sort of sounds like Yamaha NS10 monitors. The NS10 measures with early bass rolloff (won't be as significant in a real room on table or near a way) and a bump at 2kHz!

    Finally, an axiom of SBAF is that strict DF does not work. The topic used to get brought up every five months or so, and finally settled about a year ago. Don't even bother bringing this stuff up. This site is not about science. It's about what sounds right to people with trained / good ears. HF is a much better place to rehash DF, FF, HRTF, with other people who also have no idea what sounds right.

    [​IMG]
     
    Last edited: May 18, 2017
  13. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    The IEC 60318 coupler "average human ear" model may not be very representative of every other random human ear, because the variance might be significant. Specially when we are dealing with the depth and width of the external auditory meatus and auditory tube.

    Differences in volume with ear canals may result in differences in acoustic impedance (which affects 10 kHz and below). This is probably more important on IEMs than full size headphones. The reason is that the volume between the headphone cup and the head may be more constant and significant than the ear canal volume for full size headphones. So the volume driven by a full size headphone may be somewhat more constant across different individuals than when dealing with IEMs.

    As it turns out, I in particular have very deep and wide ear canals. I learned this after impressions of it were taken by UERM folk. It turns out IEMs have to drive a larger volume in my case than with the "average" dude. So as far as IEMs is concerned, what for some sounds too bassy, to me sound just right. What to some sounds just right, to me sounds limped dick.

    This all means to me that IEC 60318 "average human ear" acoustical transfer impedance is pointless given the significant variance in acoustical transfer impedance from individual to individual.

    Note I'm not even going through internal ear hair which throws another monkey wrench to the equation on to blind following of the IEC 60318 deal.

    DF sounds like shit with most music. So in general, it sound like shit. So it's pointless and somewhat of a failed experiment.

    For me in particular, correct is something that sound great with widely available music material. Incorrect, and wrong, is something that sound like shit with any random music release.

    I'm confident @Marvey is well aware of the pinna gain. Which as it turns out, is different from individual to individual. In my opinion, a standard IEM compensation curve is somewhat pointless because of the variance. I feel full size cans have a better shot at having a standard measurement target.

    I know what you are getting at. Do you think any "standard" non-flat line curve on coupler is going to be neutral to every individual when dealing with IEMs? Not by a long shot bro.

    Impedance interactions between a driver and an amplifier are going to be problematic for dynamic and BA-based drivers. Regardless of efficiency and driver type, the frequency response is going to change if the interactions are significant.

    I respect Tyll, but I'm personally convinced that using a manikin to measure full size headphones is pointless. You get a gain due to the pinna or auricle or whatever, that you are only going to take out after the fact through compensation. And that compensation better match the gain of the manikin external ear gain. Why not just measuring w/o the pinna or auricle or whatever? The driven volume should not change much.

    If a headphone or IEM does not sound like shit one does not need to DSP the hell out of it.

    I would just forget about this "soundfield" talk. They never worked for me on full size headphones. And for IEMs the "average" target is a crap target for any random dude because of the variance in ear characteristics across random individuals. It's just pointless IMO.
     
    Last edited: May 18, 2017
  14. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    1. Good / trained ears > IEC, DF, FF, HRTF, ISO, Olive Curve, etc.
    2. Having good loudspeaker references > IEC, DF, FF, HRTF, ISO, Olive Curve, etc.
     
  15. k.e.

    k.e. Winning combo: narcissistic and stupid

    Joined:
    Oct 29, 2016
    Likes Received:
    286
    Trophy Points:
    63
    Location:
    Germany
    Home Page:
    I will try to end this in peace. It was me who started this debate and I want to apologize for the escalation and noise it brought.

    I will absolutely agree that there is no universally applicable target response. Variables are too high. I do think that pinnae effects should be considered, but also room reflections, bass compensation, loudness - and in case of the Andromeda - source. Psycho acoustics play an important role and thus every measurement can be deceiving.

    My perception has changed a lot since I started this hobby and it will continue to change. No way do I have the right to claim that I have better trained ears than others. I was of the impression that the Andromeda could sound even better if they had a more forward midrange and a better refined treble. I did not emphasize enough that these are my impressions. I like my vocals upfront. That's what I am used to from speakers as well.

    Peace!
     
  16. Bill-P

    Bill-P Level 42 Mad Wizard

    Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    4,801
    Trophy Points:
    113
    Location:
    San Jose, CA
    Damn. You guys sure like to argue.

    Anyway, for me, Andro is it! End-game portable setup for sure.

    [​IMG]

    Right now I'm using medium-sized "punctured" Spiral Dot tips (more on this later when I have more free time to write up on it next week). That and AK120 has ~3.5 ohm of OI, so the combo hits the sweet spot for me. Any more OI than that and treble becomes too much to bear, and any less causes too much bass. No comment on tip mods until I can confirm that it's not really placebo.

    But now I finally have essentially an HD580 in portable form, and I'm just super freaking happy. No other IEM has ever come this close.

    There is indeed a bit too much 8-10KHz region (or somewhere within there bothers me a bit, but it's definitely higher than the 4KHz and 6KHz peaks of HD600 and HD800 respectively). But... I'm not bothered by it as much as with full-size headphones, probably because it decays quite fast and it's quite high in frequency. At worst I'll just EQ it down. AK120 has waaaaaaaaaaay too much headroom with Andro anyway. I'm barely using up to 2/3 volume (I typically listen at 80dB to give an idea), and there's no hiss in 99% of music if I don't have to push volume knob that high.
     
  17. HeartOfSky

    HeartOfSky Acquaintance

    Joined:
    Apr 16, 2017
    Likes Received:
    44
    Trophy Points:
    18
    Location:
    Seattle
    Is it safe to speculate that you "punctured" your tips in order to avoid the sound loss that happens when negative pressure occurs from too tight a seal?

    If that's the case, it never crossed my mind as an option. I've been restricted to the black silicones that came with my Vega, because they're the only ones that have the smallest of that effect.

    Is there a "right" way to puncture tips?

    Thanks!
     
  18. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    The IEC60318 is pretty neat in how it contains three different volumes in parallel to simulate an acoustic impedance curve that is supposedly similar to that of a human ear. The ridiculous thing of course is how I can get similar results (using my own developed compensation curves) from a two cent silicone tube from the hardware store.

    Ultimately, the problem with the IEC couplers is that they look nothing at all like the inside of the human ear. I strongly doubt the various peaks I've seen with various IEM measurements are actually there, at least to those extents as measured. The surfaces of the volumes inside those IEC metal couplers reflect much more acoustic energy than the insides of the human ear - thus creating the potential for standing waves (peaks and maybe nulls). The fact that changing the length of the coupler that I use also changes the frequency of the peak is evidence of these peaks being the result of standing waves.

    The human ear has Eustachian tubes, an eardrum and related middle ear boney parts (that absorb energy), tiny little hairs, ear wax, and skin and other fleshy surfaces. These are avenues for energy to be absorbed or dissipated. One other possibility is that our brains have calibrated to the acoustic effects of the ear tube, and hence is able to filter out its effects. The eardrum works in the time domain and inner ear mechanisms are RTA like, but do not work like FFT bins of fixed length.

    The only evidence I offer is that if the treble peak measurements of say the UERM and Andromeda were actually there, then I would probably have torn these two IEMs off my ears, poured gasoline of them, set them on fire, stomped on them, and then floated their melted charred remains down the LA River. So would 90% of the other UERM and Andromeda owners.
     
    Last edited: May 18, 2017
  19. Hands

    Hands Overzealous Auto Flusher - Measurbator

    Staff Member Pyrate MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    12,285
    Trophy Points:
    113
    Location:
    Colorado
    Home Page:
    I finally heard these from one of the LH Labs devices from the LH GO2PRO, I think it was. Too bright for me. Didn't care for overall timbre. Sounded kind of U-shaped tonally.
     
  20. jexby

    jexby Posole Prince

    Staff Member Pyrate Contributor
    Joined:
    Sep 28, 2015
    Likes Received:
    8,145
    Trophy Points:
    113
    Location:
    Exit stage left....
    and that folks, is why @Hands is among us. providing alternative hearing and balanced perspective to the SBAF universe.
    those were my Andros, some random dual flange tips and GO2Pro (Special Edition) with low OI (not iEMatch dongle IIRC) fed via iPad + Tidal.
    let the Opeth roll.
    :D
     

Share This Page