Target curves and common points of comparison

Discussion in 'Measurement Techniques Discussion' started by MF_Kitten, Jul 24, 2018.

  1. MF_Kitten

    MF_Kitten Banned per own request

    Banned
    Joined:
    Jan 1, 2016
    Likes Received:
    181
    Trophy Points:
    43
    I feel like the lack of reference points in the world of measurements makes it hard to really imagine what headphones will sound like tonally when looking at measurements. I see Hifiman and Audeze measurements that all look very very similar to pne another, yet people talk about Audeze having a heavy deep low end impact, while Hifiman sounds bright and neutral and light. Even just comparing Hifiman measurements makes it hard tp see that the HE500/560 are more deep sounding headphones.

    But if qe had a target curve applied that would show a mostly flat line if you measured a pair of 650's for example, you could see how different cans would sound relative to the 650's.

    This is just a thought. I'm going to start looking at the Harman curve and see how different headphones sound relative to that. It's at least a good starting point to comoare things.
     
  2. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,624
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Can you be more specific? Which Audeze and which HFM headphones? Which frequency response graphs?

    I'm not seeing the same as you. This is a good topic of conversation.

    P.S. Bass is always relative to the rest of the response, relative to the mids and highs. Frequency is response is also not the sole determiner. In terms of measurements, there is distortion and transient response.

    Note that Harmon curve ONLY works in reference to the Harmon head, ears, and torso simulator. Any thing else, and the results could be bizzare. This is because different measurement systems actually have different response characteristics. I've found that 40% of Germans, the types obsessed with absolute truth, don't understand the relative nature of the Harmon curves.
     
    Last edited: Jul 24, 2018
  3. Lyander

    Lyander Official SBAF Equitable Empathizer

    Pyrate Contributor
    Joined:
    Feb 25, 2017
    Likes Received:
    10,910
    Trophy Points:
    113
    Location:
    Philippines, The
    If memory serves @Hands posted TH-X00 FR measurements using the HD650 as compensation (i.e. Senns measure as a flat line). I thought it was a cool idea to show relative differences in tonality between headphones, though seeing how massive a pain in the pooter it'd be to manually cook up those graphs I can see how it'd be a largely academic exercise.

    @MF_Kitten You also have to factor in that though the graphs they post look the same, that doesn't guarantee that they'll sound even remotely close to one another because, as @purr1n said above, each measurement rig will have its own nitpicky differences, even if using the same compensation curve. Only compare measurements from the same rig etc etc, unless it's A**r's because they just get borked regardless.
     
  4. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,589
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Not sure what you are seeing, but this is what I see:

    Audeze:
    [​IMG]
    [​IMG]

    Chifiman:
    [​IMG]
    [​IMG]

    I do agree that flat plate couplers tend to obscure some of the differences, but even then an HE500 and LCD2r2 on the V1 flat plate coupler still looked noticeably different.
     
    Last edited: Jul 24, 2018
  5. Boops

    Boops Friend

    Pyrate
    Joined:
    Nov 8, 2015
    Likes Received:
    3,175
    Trophy Points:
    93
    Location:
    New York
    I don't remember the @Hands comparison you're talking about @Lyer25, but reading the OP, made me think of something like the below. It's pretty fudged and eyeballed, so don't look *too* carefully, but it would be cool to have a visualization like the last one as an alternative to overlaying FRs.

    sbaf-measurement-mock-01.png

    Edit: updated visuals with axis labels and updated some title text
     
    Last edited: Jul 25, 2018
  6. Lyander

    Lyander Official SBAF Equitable Empathizer

    Pyrate Contributor
    Joined:
    Feb 25, 2017
    Likes Received:
    10,910
    Trophy Points:
    113
    Location:
    Philippines, The
  7. Boops

    Boops Friend

    Pyrate
    Joined:
    Nov 8, 2015
    Likes Received:
    3,175
    Trophy Points:
    93
    Location:
    New York
    I drew them myself by hand in Illustrator. FR in top chart were done by tracing @purr1n measurements in the relevant threads. So they are not 100% accurate but close enough to get the idea ;-)
     
  8. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,624
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Nice. Reworking some visualizations right now. Will incorporate this for comparative FR plots.
     
  9. Boops

    Boops Friend

    Pyrate
    Joined:
    Nov 8, 2015
    Likes Received:
    3,175
    Trophy Points:
    93
    Location:
    New York
    Thanks for posting Hands example. I’ve basically done what he did but with a bit of Edward Tufte special sauce to make the relative differences at the major frequency regions easier to “read” for measurement newbs like me.

    A nice thing about Hands version is the distortion measurements are retained underneath. I’d add that back in if I was doing a second round on these. Would be cool to see distortion also plotted as relative to the HD650 baseline.
     
  10. 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:
    On top of all that, the HD650 is one of the flatter headphones, measurements wise, with my rig. At least, above 500HZ or so. ;)
     
  11. MF_Kitten

    MF_Kitten Banned per own request

    Banned
    Joined:
    Jan 1, 2016
    Likes Received:
    181
    Trophy Points:
    43
    Of course, they look DIFFERENT, but they are all roughly a flat line from super low to 1KHz, where they drop down like a shelf, and keep going in some kind of squiggle from there. The relative levels between low and high are hard to gauge. LCD2's are often talked about as thick and deep, while the HE-400i is often called bright and tizzy. Of course, being a mix engineer and messing with measurements and EQ a lot I know logically what would do that, but just looking through measurements of large diaphragm planar headphones, they have a similar shape to them that hides the reql tonal differences.

    I'm wondering if a very smoothed HD650 curve could be a good target curve. Get the mid bass hump, the mids, and the high end rolloff represented.
     
  12. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,589
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Well, decent planars with a good seal will mostly be a flat line to 500Hz or 1kHz, yes. But after that the differences are pretty stark. In this case the LCD3 has almost 5db less 1kHz than the others, which in itself is a pretty massive difference. The HFMs also have nearly 10db more from 5-10kHz.

    That they "drop down like a shelf" is mainly due to the flat plate coupler and the way it interacts with a typical planar driver and pad arrangement (with the distance to the driver). Ultrabike's SR-009 FR also shows it and I doubt most people think the SR-009 is 10db recessed around 4kHz. But I really I don't want to have this argument for the 4th time. I will say that I prefer like 8-10db less 3kHz than DF (well, for in-ears anyways). I know some people like a DF FR and that's fine too.
    There are some measurements of the LCD-2 and modded HE-6 on this site with a method that shows the FR more like what I hear and those show much less upper midrange drop off.
     
  13. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,624
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I think he's referring to your tweaked Tyll measurements, which were done on a head, not flat plate coupler.

    Yes you do. This is about the 29th time.
     
    Last edited: Jul 24, 2018
  14. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,624
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    For this specific comparison between LCD2C and HE400i, see below. I'm not even going to label which headphone is which. Let's say one headphone is thick and deep, and the other one is bright and tizzy.

    LCD2C vs HE400i.png
     
  15. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,624
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    The devil is in the details and where to look. There is similar shape among planars to a large extent; but the FR measurements between these two headphones look drastically different (at least to me) in the FR plots of the prior post.

    Below is another way of looking at it. Note that the Y-axis has been zoomed in to make differences more discernible. This is a plot of how the HE400i's frequency response differs relative to that of the LCD2C.
    LCD2C vs HE400i difference.png

    Relative to the LCD2C, the HE400i exhibits the following: 1) less deep bass, 2) slightly less middle mids, 3) and significantly more treble. Frequency response measurements don't tell us everything all the time, but in this case, they certainly tell us that one headphone is going to sound leaner, brighter, and a shitload more tizzy (9-10kHz) than the other.
     
    Last edited: Jul 24, 2018
  16. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,248
    Trophy Points:
    113
    This, basically. There can’t be a measurement system invariant graphical representation of a target curve. This is the main reason why Sonarworks didn’t publish their curve. Olive is trying to say that their curve is the optimum response at the eardrum, which eardrum are they referring to? You can’t glue a voicecoil to a person’s eardrum, therefore it’s still an artificial measurement system in the mix.
     
  17. MF_Kitten

    MF_Kitten Banned per own request

    Banned
    Joined:
    Jan 1, 2016
    Likes Received:
    181
    Trophy Points:
    43
    T
    This actually demonstrates why I would like to see a target/difference curve used more often. When you directly compare them, it's hilariously obvious.
     
  18. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,624
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Duly considered. Great questions and anything to make is easier.
     
  19. GTABeancounter

    GTABeancounter Friend

    Pyrate
    Joined:
    Jan 6, 2016
    Likes Received:
    321
    Trophy Points:
    63
    Location:
    Ontario Canada
    Admittedly, I really don't pay much attention to measurements and I'm far from the target audience. BUT I believe this type of curve would be very helpful for those of us who are more concerned about how the new FOTM compares to the benchmarks than say "what mod it might take to flatten out the little peak over here without compromising this nice response here".

    This is a can of worms, but if the experts here could settle on 3 or 4 benchmark headphones that excel at their respective "niche" in terms of signature (Basshead, Fun/V shaped, Warm Neutral, Neutral, Analytical?) that would make for some great comparisons. I'd argue that settling on one headphone that is the epitome of "neutral" is helpful but still falls a bit short.

    It would be really cool if the headphone being measured could be plotted as a straight line at 0DB with 2 benchmark headphones (one neutral and one fun/V shaped) plotted according to their respective differences compared to the primary headphone being reviewed. Having two benchmarks plotted would give the reader a better sense of the magnitude of the differences (ie; scaling is less of an issue)
     
    Last edited: Jul 25, 2018
  20. spoony

    spoony Spooky

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    651
    Trophy Points:
    93
    I've found largely the same effect with some headphones; some of these 'nulls' are not as deep subjectively as they appear on the FR with the flat coupler. Most of these measure very differently with the in-ear rig. I've found the effect varies so greatly that there is no recipe for correction.
     

Share This Page