Headphones Frequency Response: Challenges & Solutions

Discussion in 'Measurement Techniques Discussion' started by samvafaei, Jun 19, 2017.

  1. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    The Harmon curve is where I would vehemently disagree! It doesn't look right to me, nor does it sound right to me. The mixing / mastering / screening rooms I've been to, at the listening position, will measure flat out to a few hundred hertz (some to 100hHz, others to 300Hz, or others just slope down straight from 20Hz) and then slope down slowly -6db maybe more, by 20kHz. In fact, monitors that measure flat on-axis will tend to have curve more similar to the B&K 1974 curve more than the Harmon.

    I would agree that HD800 / HD800S sounds just about right in the bass from 500Hz below, slowing rising, hitting 90Hz perfectly, with the exception being the slow roll-off from that point toward 20Hz, which isn't as noticable because most musical content doesn't have much sub bass information.

    The HD800S does not have a massive bass boost of the Harmon curve: one that slowly starts below 200Hz, and hitsa peak boost of +5db near 50Hz. This is basically Dr Dre. Beats bass curve to me!
     
    Last edited: Jun 22, 2017
  2. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I can't agree with the supposition that older recordings did not have enough bass/mids. Perhaps older recordings had rolloff at the very extremes, 20Hz and 20kHz, but reproduction of bass/mids was not an issue, even as early as 1950s, or even earlier than that. Look up Altec VOTT. I have a miniature version of that in my living room. Same technology. Actually vintage type Altec horns and compression drivers!

    In terms of "older recordings", you need to be more specifc. I own a lot of records from the 1950s, some in mono. I can assure you that that these recordings do not lack bass/mids! Some argue that the Capitol Records recordings of the late 50s have never been surpassed in terms of sound engineering techniques, even to this day. The notion that high-fidelity (20Hz to 20kHz, under 0.1% THD, etc.) only really existed in recent times is totally unfounded. HiFi has been around for a long long time, even during the days of tubes. And no, the tube gear of that day didn't like sound colored and muffed like some modern interpretations of such gear. Much of the "tube" sound is the result of cheap output transformers with high distortion in the bass. Much of the art and science of transformer winding has been lost because of the introduction of transistors.
     
    Last edited: Jun 22, 2017
  3. 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:
    The Harman/Olive/Welti/McMuffin curve seems to make some sense on some measurement heads if you ignore the bass. The whole bass boost thing is crazy! It's definitely consumer/noob friendly, I guess, but it's definitely not accurate. However, there may be a FEW headphones where they need a bit of extra bass to give them the slam and power their driver might inherently lack (Looking at you, planars!). But even then, the extent of that curve usually even makes those cases too bassy.

    At the end of the day, the best thing you can do is listen to decent speakers, plus sub if needed, with known-good measurements for a while, and decently setup in an OK room if you can. I mean, I haven't heard the best speakers around, but it's obvious to me how wrong most headphones sound based on that (i.e. I'm no speaker expert!). Just ignore headphones and especially headphone measurements for a while. Then go back to the headphones once you feel very confident with what you hear with speakers, listen, think about how they sound relative to the speakers, and then take a look at if your measurements make any sense. This is very general and non-targeted advice.
     
    Last edited: Jun 22, 2017
  4. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    This! I think it's important to identify who exactly are these listeners in these studies. My reference is people I know who work in the TV and film industries, and people here. I know for a fact that if a studio engineering intern / apprentice calibrated speakers to use the Harmon curve, they could kiss goodbye to being retained as an employee.

    I'm in the middle of developing standards for content security in the film industry. On this panel, there will be NO randos, untrained people, scientists, or academics. We only want subject matter experts and people who are actually digging the ditches.
     
    Last edited: Jun 22, 2017
  5. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,276
    Trophy Points:
    113
    The rooms we've toured in LA and elsewhere usually don't have this kind of sloping. I have a hunch that Katz might be responsible for it. I know that Neumann, Unity Audio, Genelec and others generally shoot for a flat anechoic response. Usually in untreated rooms you have a pretty good response north of Schroeder frequency (if it's not a total bathroom acoustically) and down south it's room mode city.

    In our speaker calibration software our default setting is flat around the sweet spot. It's also what most of our clients use daily. Now, I won't say that there is one correct response for recreative listening. In terms of headphones my experience is that once you give folks flat, the response they have immediately is mostly dictated by the uncorrected headphone sound. Meaning - closed can guys find the bass okay, open can guys find it excessive. Give them three days and it evens out.
     
  6. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I don't know which rooms in which studios that you are talking about. All the screening rooms, production bays, mixing rooms, etc. at the places I've visited around the world in AUS, UK, HK, SK, NY, LA all sounded the same more or less to me. Some of the monitors (Adam, KRK, JBL, Genelec are common) may be slightly bassier than others, some less so.

    All the rooms I've been in are fully treated. Even the smaller mom and pop facilities, they are treated. One possible difference between us is that I only visit film, TV, mixed music facilities. Mostly post work, but some recording. In two channel mode, the measurements at the listening position will generally be a slight downward slope on an RTA. Keep in mind the microphone will oriented straight forward, which will be bit off-axis slightly from the monitors, and the monitors may not be toed or only slightly toed in.
     
    Last edited: Jun 22, 2017
  7. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,276
    Trophy Points:
    113
    Okay, maybe I see why there's a difference. Are you using a 0 degree mic calibration? Probably yes. Our software needs a 30 degree calibration file, which we provide together with the 0 degree one. Our measurement is a "smart" average (closer to sweetspot gets more weight) of 24 measurements, with mic always pointing towards the centre of phantom stage.

    So yeah, different systems.
     
  8. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    No calibration on mic. Just point straight ahead. And again, fully treated rooms.
     
  9. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    This discussion is stupid. To end the fight once and for all I'm introducing the Serious-Special-Sauce-Target for Tyll's graphs. One target to rule them all:

    Serious-TargetV0.63.png
    Kidding aside, I attached some graphs that compare this to a few headphones. IMO it's generally not too bad, but Tyll's coupler might not always give the absolute best results. Target could also use some work. Still not shit.


    EDIT:
    Realized that this target was shit and it's now long obsolete.
    New thread here: http://www.superbestaudiofriends.org/index.php?threads/innerfidelity-fr-target.5560/
     

    Attached Files:

    Last edited: Dec 21, 2017
  10. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Pretty much. Can you process MDR-Z1R?

    The follow up, the method to calibrate has been pretty routine for like forever. Point microphone ahead. Play pink noise. Look at RTA. None of this stuff is terribly complex. There are variations on B&K 1974 curve: some like to have a flat midrange, steeper or more shallow slopes, just a straight slope from bottom to top, a slight mid-bass bump (already inherent, but could also think of this as a slight sub bass rolloff), etc. But I assure you, no place I've been to has the Olive +5db bass bump with wide Q of 0.7. That's the kind of shit that makes sound engineers look at you funny.

    Seriously, all this work has been done before and is inherently known and expected of by good sound engineers. I don't even know why people are reversing the Olive curve to obtain a Olive speaker curve. This is way way overthinking it by people who've never been behind a console.
     
    Last edited: Jun 23, 2017
  11. samvafaei

    samvafaei New

    Joined:
    Jul 15, 2016
    Likes Received:
    35
    Trophy Points:
    13
    Agreed, Soundstage is a big part. I just wrote something about Soundstage and Imaging on another forum that am going to copy/paste here in case you want to share your opinions on it:

    "We have our own ideas about localization effects at rtings. I haven't written a full article on what we call Soundstage and what we call Imaging yet. But if you click on the blue question marks in our reviews, you get a short description of what each test is.

    Soundstage: Soundstage qualities are not inherent to the audio content, the headphones have to 'create' them rather than 'reproduce' them. They determine whether the sound is perceived to be coming from inside or in front of the head, how open and spacious the soundstage is, how much the headphones acoustically interact with the environment, and how strong the phantom center is.

    So basically the combination of HRTF, room/reverb effects and crosstalk is what I put under Soundstage. In other words: the difference between a pair of loudspeakers in a room and a pair of headphones. We have 4 tests for Soundstage: PRTF, Openess, Acoustic Space Excitation, and Correlated Crosstalk. You can read more about them on the website.

    Imaging: Imaging qualities are inherent to the audio content, the headphones have to 'reproduce' them rather than 'create' them. They determine how accurately the objects are positioned in the stereo image, and how transparent the imaging is.

    In other words, what I call imaging are the localization cues that are already in the music. That is the amplitude and phase qualities in the music which result in panning and depth effects (Look up ITD and IID on Wikipedia under the Sound Localization article). Unlike Soundstage, there's not much difference between Imaging capabilities of speakers and headphones.

    Since our Gold Standard for headphones sound is the "perfect loudspeaker, in the perfect room" (whatever perfect means!), then headphones that have a speaker-like Soundstage would get full score from our Soundstage test. It's a bit more complicated than that though!"

    Agreed. B&K has more Mid range which to my ears muddies up the sound, especially the 250-500Hz bump. But if you like the B&K sound, that's OK too. I put it down to personal preference.

    I personally don't like the Sennheiser's lower-mid bump and to my ears muddies up the sound (and am not alone here). Most well-liked closed-backs have little or none of this low-mid bump, and instead have more low-bass. This suggest to me that since open-backs can't produced a lot of low-bass, they try to compensate by adding low-mid/high-bass. And since that has been done for a long time, some have developed a taste for it.

    I agree that using a TV frequency response to support my claim was not the best idea. But I have listened to a lot of TVs recently, since I'm developing the sound section of our TV reviews as well and this is a very good sounding unit, for a TV of course. And its In-room measurement is very close to Harman's In-room measurement. Another TV with great sound which measured even closer to Harman's was the Sony X930c (with from facing speakers, from a couple of years ago). My point was that more and more manufactures are building sound units based on Floyd Toole's research and to my ears, they sound better than other designs.

    As I mention in the video, we have locked this curve down for 2017, but will check if there is room for improvement for 2018.

    The microphones were in the same position on each head. At the pinna opening. The curves are compensated, so don't be distracted by the flattness of the HD 800 S. I check when I get back to the office on Monday to see if there was an issue with the compensation and will report back. But as I mentioned, that issue is not relevant to the point I was making. My point was the relative difference. The amount of difference between 600 and 800 S at 2KHz on me, is not the same on other heads. I also did multiple reseats to make sure it wasn't a placement issue.

    No problem. We can agree to disagree on the Harman curve. But here is the HD 800 S vs Beats Executive. Not quite the same amount of bass ;-)

    http://www.rtings.com/headphones/graph#250/323/290

    Agreed. As I mentioned it before, that idea of mine was quite a stretch, but the only one I could personally come up with to explain the bump in the B&K curve.

    Tyll uses the (inverted?) Independent of Direction curve that comes with the HMS as his compensation curve, which makes everything look bass-heavy.

    Not to toot my own horn here, but just as a way of introducing myself and to give a bit of background. I have a bachelor's in music composition and a master's in sound recording (the same program that Sean Olive did at McGill, but 20 or so years apart!). And I have studied under the likes of George Massenburg, Richard King, Steve Epstein, Phil Ramone, and Jeff Wolpert. So I'm not a total noob when it comes to these kind of stuff. :)
     
  12. briskly

    briskly Friend

    Pyrate
    Joined:
    Oct 3, 2015
    Likes Received:
    59
    Trophy Points:
    18
    Location:
    New York
    For the Harman 2013 study, 7 or so listeners who had completed the Harman listener training program. The other "untrained" listeners included a bass player and several junior recording engineers working in film.

    The variation observed between the different listeners makes it hard for me see this curve applied so broadly. The 7 to the left are the trained, the 4 to the right side are not.
    Harman listeners.png
     
  13. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Throw the 4 on the right side away, give them a Sony MDR-Z1R has a consolation prize. Throw #53's results away on the basis that this person is the one farthest from the average of the seven trained.
     
    Last edited: Jun 23, 2017
  14. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Then I have one question. Take away Olive's work. Take away B&K's work. Let's throw all that stuff away.

    Would you find on a subjective basis, a frequency response that consisted of a +5db bump at 50Hz and Q = 0.7, acceptable as reference for speakers to mix or master music or a soundtrack? We are talking about using your ears: how you would feel about this bass response for performing actual work, say mixing or mastering? Pretend you are a sound engineer working on a project for Warner Bros., Ed Sheeran, or Justice League.
     
    Last edited: Jun 23, 2017
  15. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,276
    Trophy Points:
    113
    Some clowns are actually trying to use the Olive-Welti curve for speakers? That's dumb. Sure, there is the B&K 1974 which is basically some curve that many people seem to like. Heck, turn the volume 6dB up and everyone will like the louder sound, provided there are no AFR nasties. Do people like extra bass? Yeah, they also like U curve and Grado.

    One can play with preferential tunings as long as they make sense to all kinds of music. It's like asking - what kind of spice works for every food out there?
     
  16. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    TBC, I just like to refer to B&K because there was an actual study around it. It's really much less formal than that in practice. More like "some bass with a slight slope down" or "yeah, that looks good".
     
    Last edited: Jun 23, 2017
  17. samvafaei

    samvafaei New

    Joined:
    Jul 15, 2016
    Likes Received:
    35
    Trophy Points:
    13
    I'm guessing maybe there was some misunderstanding here? The Olive-Welti curve is a headphones target, and it would be obviously wrong to tune a room to that curve! So I'm assuming Marvey was referring to the in-room response of the Harman Listening Room, which I included in my second post. Which is based on Floyd Toole's speaker research and Sean Olive's research on in-room preferred target curves. All of these data are available on Sean's blog.

    Marvey, to your question. Since I currently don't have access to a room with a similar response to Harman's, I'll answer your question the best I can, which is a bit indirectly. Our measurement of the JBL Everest Elite 700 showed that our 2017 target curve is in agreement with Harman's target curve up to 1KHz. The JBL is based on the Olive-Welti target , which is basically the as same Harman's measured in-room response of a GRAS 45CA. So I can take it as my reference. When I listen to my reference tracks on the JBL, it has just the right amount of bass and mid.

    Also, this is my go-to simple and basic EQ for the HD 800 S when I'm listening with them for pleasure, which basically follows our current target curve:

    [​IMG]

    If we still don't agree on this, we can exchange some reference tracks I guess, but I'm not sure if this debate will get anywhere. It seems to be just a matter of taste at this point. I'm OK with agreeing to disagree, and am open to be proven wrong in the future. Maybe our 2018 curve will look different. These are uncharted territories as far as I'm concerned. Thanks for your inputs so far. :)
     
  18. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I would be curious as to frequency response the measurements of the JBL Everest as set up in your listening room (it doesn't need to be fully treated). By measurements, I mean a straight microphone setup pointed forward, and an RTA with pink noise playing. Would also like to know dimensions of the room and how far away from the walls the JBLs are positioned.

    EDIT: Forget it. That JBL you mentioned is a headphone, I thought you meant the speakers.

    I only own a HD800 which has physical modifications on it. Can you provide an EQ using your target for an HD650?

    The problem we are having is not having common references. It's possible the differences are much smaller than I suspect.
     
    Last edited: Jun 23, 2017
  19. samvafaei

    samvafaei New

    Joined:
    Jul 15, 2016
    Likes Received:
    35
    Trophy Points:
    13
    Maybe I wasn't clear enough, the JBLs that I mentioned are headphones. That's why I could take them as a reference.

    For HD650, if you invert this curve, from 20Hz to 1KHz, it'll be very close to what I would come up with.

    http://www.rtings.com/headphones/graph#245/323
     
  20. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    From that HD650 plot, I would have to disagree with you then and just leave it at that.
     

Share This Page