Some Test Results (SE vs Balanced vs Portable SE)

[Dreadhead]

Well looking through the headroom results it's interesting to see that the physical properties of the headphones are changed as well. The isolation changes with balancing:

As does the noise floor of the measurement:

Extremely interesting.

[kirkwall]

Very interesting for sure. My experience with RS1s and HD600/650s via HeadRoom electronics, balanced and SE Max, is that they're different but I couldn't choose one as an absolute preference. With the HD650, and level-matched, balanced seems to deliver a quieter, more solid sound with better LF extension and a subjectively wider soundstage; the MAX SE more focus and speed, so that lyrics and dialogue in particular are easier to follow. Perhaps with with really high-end balanced gear the benefits of balanced are more pronounced? Has anyone DB-tested identical configs in the two configurations and had a consistent marked preference?

best,

k

[Torpedo]

probably just a test artifact, especially regarding the isolation test. the cups can't be positioned exactly the same each time. i'm not terribly surprised that balancing phones lowers the noise floor.

Yep, balanced operation is used mainly to reduce noise, so no surprises here.

[Dreadhead]

probably just a test artifact, especially regarding the isolation test. the cups can't be positioned exactly the same each time. i'm not terribly surprised that balancing phones lowers the noise floor.

That's why I made a point of using an adapter to go SE. Moving the heaphones can change the curves A LOT (not just the isolation). That's what I was saying.

As far as the noise floor goes, that's the noise floor of the measurement not of the phones. The phones are being forced with a sinusoid and will put out at that frequency and harmonics of of the frequency but not a wide banded noise like that. The fact that the noise floor changes almost certainly points to the measurement being carried out at a different volume or the SE amp has severe wide banded hiss problems. In my setup the noise floor was identical between SE and balanced when the signal amplitude was the same.

[Dreadhead]

Yep, balanced operation is used mainly to reduce noise, so no surprises here.

Balanced reduces noise by common mode rejection but to do that you have to convert the signal back to SE which we don't do in most balanced SS headphone designs.

[Torpedo]

Mmmm trying to imagine the picture of one leg pushing at the driver while the other is pulling and having a noise spike with the same polarity on both legs, my take is that it's cancelling itself. Wouldn't it?

[Dreadhead]

Mmmm trying to imagine the picture of one leg pushing at the driver while the other is pulling and having a noise spike with the same polarity on both legs, my take is that it's cancelling itself. Wouldn't it?

You know what I think you're right Sorry about that! I had been convinced (by myself) of what you were saying and then convinced the other way (again by myself) but with your mental picture I think you've settled it in my head Sorry again that was dumb of me :palm:.

I still think it's a measurement artifact though since the difference is on the order of the 6dB that occurs when you go from balanced to SE. That and when I did the tests it didn't show up. The floor was actually a little higher for balanced because each amp was working at a lower gain and hence the noise floor of the amp was at a higher level.

[slwiser]

Well Headroom is redoing their headphone response curves:

Products - HeadRoom - Right Between Your Ears

And they have some balanced ones:

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

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.

I think Headroom's new charts are going to be very much discussed since their AT measurements sure don't reflect the sound that I hear.

[Dreadhead]

damnit dan ^

[Torpedo]

You know what I think you're right Sorry about that! I had been convinced (by myself) of what you were saying and then convinced the other way (again by myself) but with your mental picture I think you've settled it in my head Sorry again that was dumb of me :palm:.

I still think it's a measurement artifact though since the difference is on the order of the 6dB that occurs when you go from balanced to SE. That and when I did the tests it didn't show up. The floor was actually a little higher for balanced because each amp was working at a lower gain and hence the noise floor of the amp was at a higher level.

hehehe, when you made your comment I thought you were right, but then I tried to imagine the picture of the driver working fed by a balanced signal, and I realized that what it was doing is actually converting the balanced signal in SE motion, so the common mode noise is probably rejected.

I think it's very difficult to make valid conclusions from those measurements. You could be also measuring the noise difference from balanced to SE into the own amplifier or even the source if they're using SE or balanced outputs from the generator. Your take about the noise floor being lower for having a doubled voltage signal seems quite valid as far as the noise voltage doesn't get doubled too. If it's common mode noise, I'd expect it to be lower nonetheless.

How is output impedance on the amp affected by using SE or balanced outs? My guess is that it's higher, so you'd be lowering the damping factor, which could have audible and measurable effects. Funnily enough on our measurements on the HD600 we didn't notice any change on impulse response nor in the bass response.

[Dreadhead]

hehehe, when you made your comment I thought you were right, but then I tried to imagine the picture of the driver working fed by a balanced signal, and I realized that what it was doing is actually converting the balanced signal in SE motion, so the common mode noise is probably rejected.

I think it's very difficult to make valid conclusions from those measurements. You could be also measuring the noise difference from balanced to SE into the own amplifier or even the source if they're using SE or balanced outputs from the generator. Your take about the noise floor being lower for having a doubled voltage signal seems quite valid as far as the noise voltage doesn't get doubled too. If it's common mode noise, I'd expect it to be lower nonetheless.

How is output impedance on the amp affected by using SE or balanced outs? My guess is that it's higher, so you'd be lowering the damping factor, which could have audible and measurable effects. Funnily enough on our measurements on the HD600 we didn't notice any change on impulse response nor in the bass response.

Thanks for being generous about my idiocy. Your point about the noise being possibly common mode is valid but again I highly doubt it unless they have some very exotic electronics nearby because to be wide banded like that it would have to be running at a lot of different frequencies. Most of the noise we are usually trying to get rid of is 60Hz and the harmonics of it.

I is my understanding that the impedance effects come in two flavors:

1. If the output has high impedance this can cause a boost at some frequencies in many phones where the impedance curve is not flat. For example this happens when you plug balanced 650s into the back of a DAC1 and causes a bass boost and some roll off and is likely why some people feel the bass impact is improved by doing this.
   
2. The damping factor which could be part of it but resonances are usually very narrow banded (and if they aren't they are well damped) so I really doubt that's the difference.
   

In the end I think that their measurements seem to show enough inconstancy that I don't think they are of any use for any conclusions.