[foges]

Hi Sound Science people,

I'm considering picking up a Campfire Audio Andromeda, but after reading reviews and trying to match it with what I've learned from reading this sub-forum, I can't quite get things to line up. In particular, I'd like to drive them from my iPhone 7 (with a lightning adapter), and I've found these pieces of information (fact and opinion)

• The iPhone 7 lightning adapter/DAC is supposed to be very high quality with an output impedance of < 0.4 Ohm.
• The iPhone 6 headphone jack is supposed good, but slightly worse with an output impedance of ~4.5 Ohm
• The CA Andromeda has an input impedance of 12.8 Ohm (according to the manufacturer's website)
  
• People claim that the iPhone 6 does a great job of driving the CA Andromeda, while the iPhone 7 adapter apparently has a noticeable hiss.
• (The CA Andromeda apparently has high sensitivity -- not quite sure how this is measured)

As far as I understand, the 1/8 rule implies that the lightning adapter should be perfectly adequate to drive the CA Andromeda, while the iPhone 6 does not have a high enough output impedance. Why then do people claim it sounds better out of the iPhone 6 (or concretely why is there a hiss with the lightning adapter and not the iPhone 6)?

One thing that might explain it is that the lightning adapter has a higher noise floor than the iPhone 6, but that doesn't really line up with the general findings that there is little variation between different DACs.

I'd really appreciate if someone could help clear up this confusion. Thank you!

-Chris

iPhone 6 & 7 measurements: https://apple.stackexchange.com/que...dapter-dac-compare-to-the-internal-iphone-dac

 

[SilverEars]

Andromeda is a very sensitive iem, and also as shown above, the input impedance drops significantly around the bass region, but also mids are relatively low, but the twin peaks treble area should be the most stable to interactions with various output impedance levels. Andromeda has a slight more or less linear increase in bass toward the subs. So, people sees this as using the impedance value from 2.5-4ohms to level out the bass a bit(to get the lowend toward neutral than warm). In return relative to what was heard before(should have been warmer if output impedance was insignificant close to zero), if the impedance is raised, you'd hear more of the treble since in relatively, the bass has been reduced.

The added impedance reduces the hiss as well(like I've mentioned it's a sensitive iem). Can somebody explain how impedance reduces hiss? Does the output impedance have to do with the gain of the amp, is that the reason?

 

[castleofargh]

Andromeda is a very sensitive iem, and also as shown above, the input impedance drops significantly around the bass region, but also mids are relatively low, but the twin peaks treble area should be the most stable to interactions with various output impedance levels. Andromeda has a slight more or less linear increase in bass toward the subs. So, people sees this as using the impedance value from 2.5-4ohms to level out the bass a bit(to get the lowend toward neutral than warm). In return relative to what was heard before(should have been warmer if output impedance was insignificant close to zero), if the impedance is raised, you'd hear more of the treble since in relatively, the bass has been reduced.

The added impedance reduces the hiss as well(like I've mentioned it's a sensitive iem). Can somebody explain how impedance reduces hiss? Does the output impedance have to do with the gain of the amp, is that the reason?

the easy idea (purely resistive system) is to assimilate a high impedance amp as being a low impedance amp + resistors in series in the cable that will attenuate the signal. from the IEM's point of view it's pretty much the same.
or just remember that power efficiency is best when the source has an impedance close to zero or the load close to infinity. it's not like we need power efficiency, but the rule still applies anyway.

there can be many sources of noises so you can't hope to get away from them all just by attenuating the signal with impedance, and boosting the volume level at the same time. but it will work with some noises. so if the hiss of the DAP happens to be fairly stable noise that won't rise in a linear fashion with the volume level output, then adding impedance at the source is likely to do the trick for hiss(no telling what negative impact it will have elsewhere). and a voltage divider is kind of the same idea but with an added desire to control the impedance seen by the load.
now I'd personally be very careful not to jump to conclusions that high impedance sources in general have lower hiss. there is obviously more to a design than impedance output, and if a DAP hisses more into a very sensitive IEM, then it just means that it has a higher noise floor. could be impedance, could be that the DAP is crap, could be EMI/RFI hitting a op amp particularly sensitive to electromagnetic crap, could be that it has too much gain for our use... hissy stuff are annoying, and I'd leave it at that. ^_^

 

[MindsMirror]

The hiss and output impedance are not really related. The iPhone 6 has a better nose floor and worse output impedance. The iPhone 7 dongle has a worse noise floor and better output impedance. From the pieces of information you listed, it seems like the hiss from the dongle is a worse problem with the Andromeda that the output impedance from the iPhone 6.

There are some caveats in the "most DACs are the same" thing that you mentioned. Most DACs have sufficient dynamic range such that the noise floor will be inaudible in almost any situation. However in this case of using a sensitive IEM without an amp in-between, you can't utilize the full dynamic range of the DAC, the absolute nose floor is more important. So in this comparison, that difference is definitely audible.

 

[SilverEars]

Well there are cases people resolve hissing problems with adding resistors to the output.

I'm trying to figure out what is really going on that reduces the hiss. Even if resistor can be seen as attenuating the signal as a whole, it's still a matter of raising the volume to increase overall signal(audio signal + hiss). Since the audio signal and hiss is going through the resistor, you'd expect both increase or decrease proportionally.

I also understand that there is a difference between gain and volume.

Also, high gain per sensitivity would be too much so that the noise is heard through a sensitive headphone.

How is a resistor suppressing the hiss relative to the signal?

My question was generally regarding reducing noise with added resistance at the output. Now, if you know what goes on, I'd like to know.

 

[71 dB]

How is a resistor suppressing the hiss relative to the signal?

The hiss is independent from volume (happens after gain adjust), so turning volume up increases signal but not hiss.

 

[bigshot]

Of course all of this has more to do with the interaction between the headphone and the amp, than it does the DAC itself. With the proper match of headphone and amp, an amp that has a lot of hiss with one combo will sound fine with another. The problem is almost never the noise floor.

I'm a member of the "less is more" school of thinking. I probably never would buy a set of headphones that required amping. I'm happy with my Oppo PM-1s because I can plug them into anything and they just work.

 

[Strangelove424]

Well there are cases people resolve hissing problems with adding resistors to the output.

I'm trying to figure out what is really going on that reduces the hiss. Even if resistor can be seen as attenuating the signal as a whole, it's still a matter of raising the volume to increase overall signal(audio signal + hiss). Since the audio signal and hiss is going through the resistor, you'd expect both increase or decrease proportionally.

I also understand that there is a difference between gain and volume.

Also, high gain per sensitivity would be too much so that the noise is heard through a sensitive headphone.

How is a resistor suppressing the hiss relative to the signal?

Because the signal power is relative to the resistance and the signal power required to achieve volume gets magnified by a resistor. That also helps with increasing range control for volume. It keeps the noise floor out of your audible threshold, and puts the demand on the amp to increase power level above that threshold for music. Does the noise floor get amplified as well? Yes, but since the resistor put the noise floor outside of the audible threshold, you can't hear it during the quiet parts, and when an actual music signal gets amplified along with the noise you don't hear the noise floor because of masking. Does the noise disappear completely? No, it's just hidden beneath a lower floor. But it's a far better situation than having your ambient voltage level and peak voltage too close to each other.

 

[foges]

Thanks for all the detailed responses!

So to summarize, it seems that one explanation is that there is a constant noise source (independent of signal volume), and this gets drowned out by a higher signal (due to higher resistance) on the iPhone 6, but not with the lightning connector. This is then only really an issue with very sensitive headphones where the you can't make use of the full dynamic range of the DAC/Amp. One could add a pure resistive element in series, but that would increase the source resistance (impedance), getting you closer to the 1/8 limit.

I found this article that seems to be claiming basically the same thing, that there is noise at -86dB, that's audible with sensitive IEMs:

https://www.heise.de/newsticker/mel...dapter-kann-Rauschen-verstaerken-3327892.html

@bigshot: Would you mind expanding on your statement "proper match of headphone and amp ... The problem is almost never the noise floor"? How do you define a proper match?

I guess a follow-up question to all of this is what is an amp that's good for high-sensitivity headphones, while at the same time being very portable?

 

[bigshot]

When you say DAC, you're referring to the chip that converts from digital signals to analogue. Those are usually audibly transparent. No audible noise floor at all. The circuit that is problematic is the one that amplified that signal for the headphone output- the headphone amp. Potentially there may be impedance mismatches between headphones and anything with a headphone jack. And a DAC may have problems as well, but it isn't because of the DAC. It's the headphone amp part of the DAC. If you take line out from the DAC that is causing a problem and run it through an amp that is well matched to your headphones, the DAC itself is fine.

 

[71 dB]

I'm happy with my Oppo PM-1s because I can plug them into anything and they just work.

Technically Oppo PM-1 is a pair of ridiculously expensive 32 Ω resistors which make noise.