Some Test Results (SE vs Balanced vs Portable SE)

[luvdunhill]

I will look into the speaker workshop but to be honest I don't see how the various amps would be seeing a different impedance if the response was the same.

simple. you'll have more data variance in the impedance plots and they will be make a much better comparison. For example, one things I use impedance plots for in speaker design is to determine which frequencies the enclosure itself wants to resonate at. This would be interesting data for headphones. I'd recommend looking at D'Appolito's book on speaker testing. It's pretty much the reference on the subject, besides IEEE journal articles.

[Dreadhead]

Yep, I posted those graphs at HF too some time ago. We didn't have an adapter, so we needed to swap the wires at the headphone's end........ snip/QUOTE]

I thought so about the wires. I think you guys did a great measurement I was just wondering if that could be one source of the difference you observed being different that's all.

Nope it's no earthworks M30 but it costs I think about one 40th the price In the end I'm comparing again with relative responses in the two configuration and they are not different at all (at least at the two frequencies I mentioned earlier). That's the joy of this measurement once the setup stays constant and it is accurate to a certiain degree (.5% THD isn't bad) then the rest is moot It's an entirely different story for some other stuff more like what your friend does.

[Dreadhead]

simple. you'll have more data variance in the impedance plots and they will be make a much better comparison. For example, one things I use impedance plots for in speaker design is to determine which frequencies the enclosure itself wants to resonate at. This would be interesting data for headphones. I'd recommend looking at D'Appolito's book on speaker testing. It's pretty much the reference on the subject, besides IEEE journal articles.

I'm not arguing it's interesting I'm saying I don't see the point in an amp to amp comparison. It appears that it's not but I can't see how different amps are going to change the physical properties you're trying to see. I'm not looking to put Headroom out of the headphone testing business anyway.

[Dreadhead]

Oh well at least I figured out why my THD readings were so different than luvdunhill's and that's because as with any good nonlinear effect the THD varies a lot with the amplitude of the forcing. If I decrease my amplitude at 1kHz until the signal has an amplitude of 75db then the THD drops to .05% (-66db) which is much more on the order of what he was reporting. When I volume match at the two though I still don't get much variation between SE and balanced.

[kirkwall]

Oh well at least I figured out why my THD readings were so different than luvdunhill's and that's because as with any good nonlinear effect the THD varies a lot with the amplitude of the forcing. If I decrease my amplitude at 1kHz until the signal has an amplitude of 75db then the THD drops to .05% (-66db) which is much more on the order of what he was reporting. When I volume match at the two though I still don't get much variation between SE and balanced.

Thanks for these -- interesting. I may have missed this in earlier posts, but how would you say that these measurements correlate with your listening experiences of balanced/SE?

best,

k

[luvdunhill]

I'm not arguing it's interesting I'm saying I don't see the point in an amp to amp comparison. It appears that it's not but I can't see how different amps are going to change the physical properties you're trying to see. I'm not looking to put Headroom out of the headphone testing business anyway.

Why does Stereophile post these graphs then? Probably just to fill up space...

[Dreadhead]

Why does Stereophile post these graphs then? Probably just to fill up space...

I had not seen them in their amplifier reviews. The do talk about the output impedance of the amp though. I'm not talking about the headphones here I'm talking about the amp.

[luvdunhill]

I had not seen them in their amplifier reviews. The do talk about the output impedance of the amp though. I'm not talking about the headphones here I'm talking about the amp.

I don't get it. Carry on.

[Filburt]

Try a multitone generation with dynamic amplitude as the reference if you can, as this would best approximate something music-like. With something like weighted averaging you can probably get a decent metric of what the relative distortion is. I actually prefer running an FFT and looking at the harmonic amplitudes, but it gets pretty difficult once you do something like what I just suggested above. If I'm stuck with static amplitudes and only uni or duotones then FFT is pretty much the only thing useful, as THD will tell you almost nothing. There are theoretical and practical reasons why balanced could help but it's also plausible that it doesn't pan out to much practical advantage in the case of headphones due to systemic limitations. I've not tested it myself so I don't know >.<

[hYdrociTy]

>.<

[Dreadhead]

Try a multitone generation with dynamic amplitude as the reference if you can, as this would best approximate something music-like. With something like weighted averaging you can probably get a decent metric of what the relative distortion is. I actually prefer running an FFT and looking at the harmonic amplitudes, but it gets pretty difficult once you do something like what I just suggested above. If I'm stuck with static amplitudes and only uni or duotones then FFT is pretty much the only thing useful, as THD will tell you almost nothing. There are theoretical and practical reasons why balanced could help but it's also plausible that it doesn't pan out to much practical advantage in the case of headphones due to systemic limitations. I've not tested it myself so I don't know >.<

First off the THD numbers I'm pulling are from an FFT based approach as you can see in the plots that I put earlier in the thread. Also far as I'm concerned it shows a lot and others appear to agree since it's a pretty standard test. I think I have the option for a couple other signals with the signal generator and I will see if there are any others of interest.

Actually I was thinking about it this morning and what I will do is find a piece of music (rock? classical?) and and do a volume match recording using the SE and balanced and compare the recording through time after doing some time matching and normalizing the signal. With the software I have this should be pretty easy to do.

I'll give it a swing this afternoon and we'll see what happens but I'm pretty sure the signal will come back through the mic identical to a very high degree.

[Dreadhead]

In other news I just realised why the noise floor was higher on the balanced setups. It is because to volume match the signal both amps have to be working at 6db lower gain which of course raises the noise floor by the same 6dB that I just realised is the difference I saw. So at least that makes sense.

[Filburt]

First off the THD numbers I'm pulling are from an FFT based approach as you can see in the plots that I put earlier in the thread. Also far as I'm concerned it shows a lot and others appear to agree since it's a pretty standard test. I think I have the option for a couple other signals with the signal generator and I will see if there are any others of interest.

Actually I was thinking about it this morning and what I will do is find a piece of music (rock? classical?) and and do a volume match recording using the SE and balanced and compare the recording through time after doing some time matching and normalizing the signal. With the software I have this should be pretty easy to do.

I'll give it a swing this afternoon and we'll see what happens but I'm pretty sure the signal will come back through the mic identical to a very high degree.

I'm not trying to bag on you; I was just trying to help since we've both apparently tried to engage in projects having to do with measuring real-world audio performance via measurements. I've had some success doing so, so I tried to pass on some of what I'd learned through experience to help in this endeavour as I'm interested in the project . I probably should have been a little more specific, though. What I meant by looking at the harmonic amplitudes is analysing the overall distribution against psychoacoustic models so you can get a good idea of what the significance is in terms of overall audio performance.

I've tried the other thing you're suggesting before. It does work; it's just harder to quantify the results enough to make qualitative predictions. However, if you're persistent, and do a lot of tests to figure out what the audible significance is of each component of the result (or can figure it out from some psy models), then it can be helpful.

[Dreadhead]

I'm not trying to bag on you; snip

Sorry if I came of combative I'm a bit of an arguer . I just was surprised at you saying THD was unimportant (which I think I misunderstood). I actually have first hand experience as part of this measurement binge where the THD was actually the killer for the use of a particular pair of headphones (The SA5000s had to be retired). It pretty much made them un-listenable when I corrected the dips in response at some frequencies. The D5000s and HD650s don't have the same issue though.

All I'm planning on doing is looking at the DB difference between the two signals vs time. I'm not a big "psycho" anything kind of person You may find my other thread about turning one headphone into another (which I also have working).

[Filburt]

Sorry if I came of combative I'm a bit of an arguer . I just was surprised at you saying THD was unimportant (which I think I misunderstood). I actually have first hand experience as part of this measurement binge where the THD was actually the killer for the use of a particular pair of headphones (The SA5000s had to be retired). It pretty much made them un-listenable when I corrected the dips in response at some frequencies. The D5000s and HD650s don't have the same issue though.

All I'm planning on doing is looking at the DB difference between the two signals vs time. I'm not a big "psycho" anything kind of person You may find my other thread about turning one headphone into another (which I also have working).

I don't mean to say distortion means nothing. What I mean is that THD isn't very helpful other than as like a preliminary "something is wrong" kind of measurement. Harmonic amplitudes compared against psychoacoustic modeling is a lot more helpful in terms of discriminating between 'good' and 'bad' performance. Measurements are tremendously useful; I used them in my work on my DAC over the summer. It's just evaluating overall audio performance appears to (at least to me) require a finer measurement than what THD offers . One thing to watch out for in DACs, for example, is you can take some standard uni and duotone THD measurements and think yourself distortionless, then you run a complex multitone config to try to sim actual music and watch your harmonic distribution outright stink on top of having some aharmonics or overall indications of DIM problems.

Anyway, like I said, I'm just trying to help out. Good luck on your project

[Dreadhead]

Ok well I did a quick and dirty version of the recorded music:

Radiohead, In Rainbows, 15 Step (about the first 45 sec)

I took two recordings of the song using the balanced signal so that I could compare two recordings that are "identical" to the error of sampling frequency (44.1 kHz) and then I did a SE recording of approximately the same time.

I then loaded the two balanced files matched the start time using a search algorithm to find the first peak over a certain amplitude and assumed they matched (fairly inaccurate but I did not want to interpolate). I then normalized the next two seconds of audio and calculated the decibel difference between the two: -45 dB (not bad considering the "dirtiness" of my method)

I then compared one of the balanced recordings to the SE recording using the same exact method and guess what -43dB.

That means that the two recordings have almost identical behavior. 2dB difference between SE and Balanced is VERY low on the noise floor. I think even more proof that balanced doesn't seem to make any real difference once you volume match.

[Dreadhead]

Here is a longer 10 second sample and a comparsion of the difference between the signals:

Blue: dBFS difference between SE and balanced

Green: dBFS difference between balanced and balanced

As I said not much difference.

[Dreadhead]

And the original balanced signal (just to show it's fairly complex):

[Upstateguy]

Hey Dread

The arguments against your findings are starting to sound like cable believer stuff.

May I ask you something?

Using your best objectivity, do you hear any difference between balanced and SE ?

USG

[Dusty Chalk]

I'm not sure I understand -- what do you mean by, "you normalized the audio"? Is that the volume matching that you described, or do you mean normalization in the digital compression sense?

[Dreadhead]

I'm not sure I understand -- what do you mean by, "you normalized the audio"? Is that the volume matching that you described, or do you mean normalization in the digital compression sense?

Dusty,

I did volume match the two signals (physically) but to get the full scale maximum to be 0db (1.0000) then you need to divide the signal by the maximum amplitude observed to make sure that the signal varies between -1 and 1 so that's what I mean. Sorry if you know this but I figured I'd be detailed so that other could understand too.

I'm not entirely sure what they do in digital compression but I would think it would be similar but for a different purpose.

[Dreadhead]

Hey Dread

The arguments against your findings are starting to sound like cable believer stuff.

May I ask you something?

Using your best objectivity, do you hear any difference between balanced and SE ?

USG

USG (and the other person that asked before),

I honestly thought I did. I was fairly gutted with these findings.

Through this process I've discovered that my DAC3 makes it very easy to do A/B on the signal because all you have to do is increase the digital volume control 6db when you go balanced to SE or -6db when you go the other way. I have confirmed that this is exactly what happens too by measuring the actual headphone output with my measurement mic setup.

I'm spending some time with the HD650s doing this A/B between the two doing the process described above and I have become fairly convinced I don't hear anything but I'd say that I haven't reached a full conclusion yet. I certainly would be stunned to but I really want hear a difference. If I do manage to find one then I'll have to devise a test to find that effect before I'll buy it the other way.

So for now I sit pretty much in limbo with a fairly strong leaning towards no effect (with HD650s too). For example yesterday I was 30 minutes into a session and was convinced I'd found something and that I was listening to SE and then I looked down to discover I was in balanced.

The problem is that unless you do a very careful volume match with and SPL meter and without moving the headphones doing an A/B is close to impossible and balanced is naturally 6dB louder than SE and louder nearly always wins in audio stuff (all else being equal).

That said I'm not selling my GS-X so don't PM me I have no plans to do A/B on GS-X vs AE-2 but as the measurements show I don't think they'd be conclusive either, in the end they are both amazing amps. I am consistently amazed by the AE-2 and I am convinced it is better than the Total Bithead I had before (and will again once it returns from repair).

[Dreadhead]

I don't mean to say distortion means nothing. What I mean is that THD isn't very helpful other than as like a preliminary "something is wrong" kind of measurement. Harmonic amplitudes compared against psychoacoustic modeling is a lot more helpful in terms of discriminating between 'good' and 'bad' performance. Measurements are tremendously useful; I used them in my work on my DAC over the summer. It's just evaluating overall audio performance appears to (at least to me) require a finer measurement than what THD offers . One thing to watch out for in DACs, for example, is you can take some standard uni and duotone THD measurements and think yourself distortionless, then you run a complex multitone config to try to sim actual music and watch your harmonic distribution outright stink on top of having some aharmonics or overall indications of DIM problems.

Anyway, like I said, I'm just trying to help out. Good luck on your project

Thanks for the insight on the DAC issue. The problem is that if you have a nonlinear system (which it shouldn't be but is) and put in multiple frequencies then you end up with first the nonlinear responses to the frequencies themselves (harmonic distortion), and difference frequencies (aharmonics). Nonlinearity is a PITA as the saying goes. I wonder if you're phasing all the signals relative to each other and causing actual distortion due to the beat pattern that actually comes out the other end clips digitally. Unless you introduce random phasing to your audio signals that's almost certainly going to cause serious issues that would appear to be other things.

I'm not entirely sure what you mean by DIM. Intermodulation?

Thanks for the advice.

[Dusty Chalk]

Dusty,

I did volume match the two signals (physically) but to get the full scale maximum to be 0db (1.0000) then you need to divide the signal by the maximum amplitude observed to make sure that the signal varies between -1 and 1 so that's what I mean. Sorry if you know this but I figured I'd be detailed so that other could understand too.

I'm not entirely sure what they do in digital compression but I would think it would be similar but for a different purpose.

Okay, so you just multiplied everything by a constant -- that's not the same as digital compression normalization.

[Dreadhead]

Well Headroom is redoing their headphone response curves:

Products - HeadRoom - Right Between Your Ears

And they have some balanced ones:

=263&graphID[]=303]http://graphs.headphone.com/graphCompare.php?graphType=0&graphID[]=263&graphID[]=303

=283&graphID[]=313]http://graphs.headphone.com/graphCompare.php?graphType=0&graphID[]=283&graphID[]=313

It looks like they find a lot more difference between the two that I do. Interesting. We'll see what comes out with the 650s.

One thing to note is their measurements are at the ear-drum mine are just using a coupler and are supposed to represent at the entrance to the ear.