Head-Fi.org › Forums › Equipment Forums › Computer Audio › Measuring headphone output response
New Posts  All Forums:Forum Nav:

Measuring headphone output response

post #1 of 11
Thread Starter 

I'm writing a couple of fairly serious reviews on audio players I've bought recently. I thought I'd include a graph of the response of each of the unit's headphone outputs. But I've never done this before.

 

Can someone here explain to me what I need in order to do this and, if possible, roughly how it's done? I'm assuming I'll need to load the players with a test tone and then play that back over the headphone output and then run that signal to a line input or similar, on my soundcard. And then measure that sh*t somehow.

 

Any advice?

 

Thanks.

post #2 of 11
post #3 of 11

You may also want to get a cheap digital multimeter, and a few resistors (e.g. 2x33 and 2x100 Ohm) for testing. The maximum output voltage of the device is an important parameter, and you cannot measure it with RMAA. The output impedance is also important, you can test it by measuring the output voltage unloaded, and then loaded e.g. with a 100 Ohm resistor, and the impedance can be calculated from the voltage drop. Check if the bass response changes with a low-ish resistor load - if it gets rolled off, then the output is capacitor coupled, which is also useful to know about. In addition to RMAA, you can also use other software, such as the freely available Visual Analyser for real time FFT and waveform display. Here is a FLAC file that includes some types of signals that RMAA does not test (warning: it contains some sounds that are nasty when played with headphones), you can record and analyze it with a good wave editor.

Read this for some information on potential problems you can run into with sound card and RMAA based testing. Note that some cheap DMMs may only measure AC voltage reliably in a narrow frequency range.

 

post #4 of 11
Thread Starter 

 

Thanks for the replies, I guess I'll go with Rightmark. Thanks for the audio file also. What about the actual hookup then? If I run the amplified signal via a regular 3,5mm cable to my interface, should I hook that straight into a line jack? It's amplified, I mean. Not sure what input is right for that kind of signal. 

 

And then I just measure the frequency deviations from the original file, right? With Rightmark.

post #5 of 11

In the best case, you only need a cable like this, with the male jacks connected to the player and the ADC/sound card, and the headphones connected to the female jack. This should work as long as the output voltage of the device to be tested is safely below that of the maximum input voltage of the ADC (this is often 2 Vrms, like for the Xonar sound cards, but some devices may allow more, while onboard audio is often more limited). For most portable players, the 2 Vrms is enough, but not for portable amplifiers like the FiiO E11. To avoid an excessive signal level, use lower volume (start low, and keep it below the clipping level), or build a simple voltage divider that reduces the voltage to a level that works with your ADC. For a simple FR measurement, keeping the volume low enough should be sufficient.

If you only intend to test with RMAA, no sound file is needed, RMAA will generate its set of test signals.

I do recommend measuring the maximum output voltage and the output impedance, if possible, since these parameters are important for determining how well the device will drive a particular headphone.

 


Edited by stv014 - 2/15/12 at 5:47am
post #6 of 11
Thread Starter 

Alright, thanks. What's the reason I need a split y cable like that, and not just a straight 3,5mm-3,5mm? Is it for playback monitoring only?

post #7 of 11
Quote:
Originally Posted by petsy View Post

Alright, thanks. What's the reason I need a split y cable like that, and not just a straight 3,5mm-3,5mm? Is it for playback monitoring only?


It is for testing with a real load. Many sources perform much worse then than when driving only a line input.

 

post #8 of 11

If you are measuring the player, why use the headphones and mic at all?

Connect up using a 3.5 to 3.5 and keep the extra layer out of the equation.

Unless I am missing something here.

post #9 of 11
Quote:
Originally Posted by ROBSCIX View Post

If you are measuring the player, why use the headphones and mic at all?

Connect up using a 3.5 to 3.5 and keep the extra layer out of the equation.

Unless I am missing something here.


I did not suggest using any microphone. The headphone (or at least resistors that simulate it to a limited extent) is recommended so that the player is tested under realistic conditions. With a simple 1/8" to 1/8" cable and no load at all, the player may perform very well even if it has audible flaws when driving a real (especially low impedance) headphone.

 

post #10 of 11
Quote:
Originally Posted by stv014 View Post


I did not suggest using any microphone. The headphone (or at least resistors that simulate it to a limited extent) is recommended so that the player is tested under realistic conditions. With a simple 1/8" to 1/8" cable and no load at all, the player may perform very well even if it has audible flaws when driving a real (especially low impedance) headphone.

 


Ah, I see what you mean, loading the output.  I thought you were talking about using a mic to measure the output of a player through headphones.

Makes more sense now:)

 


Edited by ROBSCIX - 2/15/12 at 7:42am
post #11 of 11
Thread Starter 

Ah, I see. That makes sense. Guess I'll go out and get me a split y cable then. And I'll try and make sure I don't over load the inputs. Thanks.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Computer Audio
Head-Fi.org › Forums › Equipment Forums › Computer Audio › Measuring headphone output response