1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

How much power is enough?

Discussion in 'Sound Science' started by ToTo Man, Aug 1, 2018.
1 2 3 4 5
7 8
  1. castleofargh Contributor
    instead of listing flat headphones, it's safer(and more realistic) to assume that none are because we are physically different people expecting a different neutral. also http://seanolive.blogspot.com/2009/10/audios-circle-of-confusion.html

    in any case, you'd want to measure stuff yourself to try and rule out unknown parameters once you're down that rabbit hole. for a 300ohm headphone it's not really important, because small mistakes, neglecting the extra cabling used to measure, and the amp's own impedance curve, aren't usually going to make any big difference compared to 300ohm and more of the headphone. but when trying to match theory and measurements on lower impedance headphones/IEMs, even my own low-fi measurements fail to be accurate enough to account for everything and give me pretty result strictly aligning to electrical laws. not that there is more to audio signal than electrical laws!!!!!!!! just the like with everything else, it can be hard to remove or control all variables in our circuit.
    another point which is only my anecdotal experience fooling around with resistors, an EQ, and a mic, along with something I read long ago somewhere that may or may not be true(check that endorsement!^_^): changing the damping ratio for headphones and IEMs, isn't IMO, as significant as it can be on some big fat speakers(FR non withstanding!!!!!!!). if you make sure to compensate the frequency response, the changes aren't all that, subjectively. IMO, it's yet another case of being misguided by how an impulse response may look. if you look at the impulse response out of the headphone, you'll wish to get critically damped stuff all year long, but I'm of the opinion that impressions of differences will come from many other places and shouldn't be assumed to all be the headphone and all be electrical damping.

    also in the case of your own amp, there are a few out of the ordinary things in it(starting by all those impedance settings), so you can't rule out that the amp itself may change sound on its own in some audible ways, when those impedance settings are switched.
    ToTo Man likes this.
  2. bigshot
    That may sound nice, but it isn't at all like the soundstage from my speaker system. Not even close. The thing that makes it sound good is the space around the sound. You can't have soundstage without physical distance. Headphones sound like the space is packed inside the head, not a dozen feet in front of you with the sound blooming out into the room. It's the same with near field speakers. If you sit too close, the soundstage shrinks to nothing.

    I think the reason that the term is used incorrectly so often is because most people today don't own full scale speaker system. All they know is headphones, so they apply terms to headphones that don't apply.

    Flat is a calibration target. The response curve in mixing stages is carefully monitored and kept calibrated. So if you keep your system the same way, you will hear what was intended by the people making the music. Transducers are not naturally flat. They have errors to one degree or another. So you correct to get as close to the target as you can. The limiting factors are the transducers and room, but the closer the better. Hopefully you reach the point where human ears can't discern an imbalance any more. But perfection isn't required. Anything in that direction is an improvement. Once you get as close as you can, you can certainly add coloration to taste. A flat response isn't necessarily the end destination. But it's definitely the best place to start.
    Last edited: Oct 1, 2018
  3. 71 dB
    Yeah, those calculations aren't "easy" if you haven't studied electric engineering. Yes, you need to find measurements somewhere or measure yourself. An example is a long series of calculations (I have an excel table for them) and now I need to sleep…
  4. kkl10
    Random, tangentially related question to those in the know:

    Is it possible to speculate about what happens to the electrical properties (impedance curve, sensitivity, etc) of a headphone when you let an arbitrarily powerful magnet stick to the back metal plate of the driver (conventional dynamic driver)?
  5. 71 dB
    1. Aren't you a bit self-centered here with your speaker system? No speaker system is perfect, not even yours so why would anyone's speaker system be the ultimate goal of anything?

    2. Our hearing doesn't measure space and distances. You don't have measuring tapes in your ears, have you? Instead our hearing analysis spatial information in the pressure changes that our ears are exposed to. It doesn't matter what created spatial information to the sound. Now, the reason why you find your speaker system have so good soundstage is partly because the spatial information added by the acoustics of your room is 100 % real. So even mixed with the spatial information of the music itself, it sound realistic and there is feel of space. With headphones we rely on the spatial information of the recording, because there isn't any listening room creating additional spatial information. That's why the quality of spatial information of the recording is so important with headphones. The problem is that headphones on our ears aren't totally spatial-information free. It has spatial information and that information says the sound source is at our ears. This spatial information is mixed (convoluted) with the spatial information of the recording. This causes miniature soundstage in a good case or sound packed inside head in a bad case. Often the bad cases can be modified to become good cases using crossfeed, because unnatural ILD exposes brutally that you are using headphones. Your room does that too, acoustically removes unnatural ILD. Headphones give me a miniature soundstage of 1-3 feet in size with normal good recordings and about 6 feet with the best recordings. Recordings with massive reverberation such as church music tend to work best in this regard. Even modern electronic music has soundstage with headphones thanks to sophisticated tools to create spatial effects, so we are not limited to ancient church music.

    Interestingly, when I listen to music outdoors with my Sennheiser PX-200 portable headphones, I sometimes get a HUGE soundstage. I do pre-crossfeed my portable mp3-music so there are no issues with excessive ILD and the environmental noises seems to "augment" the spatiality of the recording, fool my hearing to think the music is environmental too. Again, this works with the recordings having longer reverberation (J.S. Bach's organ music / cantatas for example). the soundstage can be 20-30 feet, bigger than what I get with my speaker system home!

    3. Well, I have a multichannel speaker system and before 2012 when I discovered crossfeeding and became a headphone nut I mostly listened to my speaker system. So I know what speaker soundstage sounds like and how it differs from the soundstage of headphones. Why don't I listen to my speaker system that much? I live in a small rented appartment house so blasting music out of speakers disturbs my neighbours (limiting when I can play louder) and also I can't build a proper listening room with fine-tuned acoustics so headphone sound is superior, uncolored.
    AutumnCrown likes this.
  6. 71 dB
    The process is quite complex, even for me, but I can try:

    Headphones+amp can be modeled with this simple circuit somewhat accurately. On the left we have the output impedance of the amp (Rout). In the middle we have the voice coil (Rc + Lc). The right side is the diaphragm, or how it looks on the electric side.

    The impedance of this circuit without Rout (only headphone) is Rc + jωLc + (1 / (1 / jωL + R + jωC)), where ω = 2πf


    Below is the impedance curve of an fictional headphone and we see how we can get some values from it. The peak value (at 100 Hz here) is basically R + Rc, because voice coil inductance Lc kick in at 10 kHz or so, is neglectable at ~2 kHz here the minimum value (Rc) is seen. Another way to get Rc is to measure the DC resistance of the headhone (impedance at 0 Hz). Here we have Rc = 120 Ω, so R must be 122 Ω. We also have fo = 100 Hz. The next step is to iterate (for example in excel) the other values. Here L = 204 mH and C = 12.4 µF give the correct shape for the impedance bump around 100 Ω. Same with Lc.


    What is the frequency response error, if we feed this headphone from an Rout = 100 Ω amp?

    Lmax = 20*log(242 / (242 + 100)) = -3.00 dB
    Lmin = 20*log(120 / (120 + 100)) = -5.26 dB

    The error is Lmax-Lmin = 2.26 dB. Sometimes the impedance maximum is at 20 kHz (and even more beyond that), but most of the time the maximum is around 100 Hz due to the resonance frequency of the diaphragm.

    What is the damping ratio of this headphone (without amp)? L = 204 mH, C = 12.4 µF and R = 122 Ω

    Damping ratio = sqrt(L/C)/(2R) = sqrt(0.204/0.0000124)/(2*122) = 0.53

    This means the headphone itself is underdamped and needs electrical damping. Is the 100 Ω amp providing enough electrical damping? Now, R becomes R* = R*(Rout+Rc)/(R+Rout+Rc) = 122*(100+120)/(122+100+120) = 78.5 Ω and we have:

    Damping ratio = sqrt(L/C)/(2R*) = sqrt(0.204/0.0000124)/(2*78.5) = 0.82.

    Still underdamped. How low output impedance do we need to have critical damping?

    Zout_critical_damping = (sqrt(L / C)*(R + Rc) - 2*R*Rc) / (2*R - sqrt(L / C)) = 15 Ω
    VNandor and ToTo Man like this.
  7. bigshot
    My speaker system isn't perfect. I have concessions I have to make for the room. But it keeps getting better! That's what counts. I'd like to get a miniDSP. I could use tighter control over the EQ than the one built into my AVR.

    But a speaker system doesn't have to be perfect to sound head and shoulders better than headphones. Even modest multichannel speaker systems sound better than headphones.
  8. 71 dB
    You sound like a vinyl nut claiming vinyl is better than CD. I don't say either way about speakers and headphones. They have their own pros and cons and anyone with the insight into audio you have should know that.
  9. castleofargh Contributor
    you mean to use the headphone while an extra magnet is near the driver? that would alter the magnetic field around the coil in some way, so I would expect a change in efficiency, and perhaps some distortions caused by a less uniform magnetic field throughout the course of the coil while playing music. but all that would really depend on the magnet and where it is. I wouldn't recommend trying off axis placement as it might push the coil and diaphragm in a direction they weren't made to go. so with that off the way, the rest would probably have a more limited impact.
  10. bigshot
    Or perhaps someone who says that cross feed is better?

    By the way, CD is better than vinyl on just about every measure you can compare. That is self evident if you look at the specs and listen.

    It's also pretty self evident that the best speaker systems trounce the best headphones. If you don't have a room and can't bother people around you, headphones are a good compromise. But music is recorded, mixed and mastered to be listened to on speakers, not headphones.
    Last edited: Oct 2, 2018
  11. kkl10
    Could this affect the headphone impedance curve?
  12. 71 dB
    1. I think I have explained enough why crossfeed is beneficial for headphone listening (see [3.] below).
    2. I totally agree.
    3. How many percent of people do you think can afford "the best speaker system"? Are a pair of Genelec 8351A good enough? Personally I am unwilling to invest more than $300 on headphones so I use "the best bang for the buck" -stuff. You are right, music is mainly recorded, mixed and mastered to be listened to on speakers. That is actually my main justification for the use of crossfeed: It makes headphones more speaker-like simulating acoustic crossfeed with electric crossfeed.
  13. 71 dB
    Yes. A magnet could change the value of force factor Bℓ which changes the impedance curve because mechanical impedance Zm of the diaphragm is seen on the electrical side as electrical impedance Ze:

    Ze = (Bℓ)²/Zm.​

    This means that if Bℓ (magnetic field B) increases/decreases just 10 %, Ze will decrease/increase 21 %. Increasing Bℓ 10 % increases damping ratio (heaphone alone) 21 %.
    kkl10 likes this.
  14. bigshot
    Oh! Headphones are definitely cheaper and more convenient than speakers. They're easier to set up too. And they are kinder to the neighbors late at night. They just don't generally sound better. Sound expanding out in a room sounds better than sound squirted into each ear. And crossfeed doesn't even come close to what a room adds to the sound of speakers. All of this is as obvious as the fact that LPs aren't as good as CDs.
  15. Whazzzup
    How about this much

1 2 3 4 5
7 8

Share This Page